Article Text


Sex selection and regulated hatred
  1. John Harris
  1. Sir David Alliance Professor of Bioethics, Institute of Medicine, Law and Bioethics, University of Manchester, Manchester, UK
  1. Correspondence to:
 John Harris
 Institute of Medicine, Law and Bioethics, School of Law, Williamson Building, University of Manchester, Oxford Road, Manchester M13 9PL;


This paper argues that the HFEA’s recent report on sex selection abdicates its responsibility to give its own authentic advice on the matters within its remit, that it accepts arguments and conclusions that are implausible on the face of it and where they depend on empirical claims, produces no empirical evidence whatsoever, but relies on reckless speculation as to what the “facts” are likely to be. Finally, having committed itself to what I call the “democratic presumption”, that human freedom will not be constrained unless very good and powerful reasons can be produced to justify such infringement of liberty, the HFEA simply reformulates the democratic presumption as saying the opposite—namely that freedom may only be exercised if powerful justifications are produced for any exercise of liberty.

Statistics from

Sex selection: options for regulation. A report on the HFEA’s 2002–03 review of sex selection including a discussion of legislative and regulatory options. A critique

I have titled this paper discussing sex selection and the recent report by the Human Fertilisation and Embryology Authority (HFEA)1 “Sex selection and regulated hatred” after the famous essay on Jane Austen by D W Harding.2 In that essay Harding argues that Austen’s work is best understood as subtly expressing Austen’s hatred for the small-minded and petty bourgeois world, full of prejudices and conceits she describes, a hatred regulated by her ironic and deft prose. The HFEA report embodies a sort of mirror image of Harding’s insight. In the report, the opinions emerging from the consultation exercise, which, unsupported by evidence or valid arguments are impossible to distinguish from prejudices, are given formal approval and proposed regulation by a government appointed body set up with the responsibility to provide expert leadership. The leadership in fact exemplified in this report owes much to the legendary Duke of Plaza Toro, so faithfully and tellingly recounted by W S Gilbert.3

In her introduction to the HFEA’s report Suzi Leather, Chair of the HFEA, remarks:

I consider that our conclusions and the advice contained in this report represent an informed, balanced and proportionate response to the very complex issues raised by sex selection and I hope it will stand as a principle point of reference for all those—Government, professionals and the interested public—who will be involved in taking the debate forward.

I suppose that by stretching a point one might call this report “informed”; balanced and proportionate it isn’t. However, while undoubtedly informed, it is informed largely by the results of a public consultation, the “hostility” of which to sex selection is manifest and even explicitly acknowledged in those terms by the HFEA, and is accepted at face value. But more importantly the HFEA’s report is hopelessly inconsistent and, in the rare cases in which arguments appear, very poorly argued. The form of the document is also revealing. The bulk of it—that is, Chapters 1–4 (of six chapters) concerns background, overview of the reportage of commissioned research, and public attitudes. Virtually the whole of the burden of establishing the report’s conclusions is permitted to, and clearly does, emerge from the public consultation. Whether or not the public consultation was informative of the debate it was in effect determinative of the conclusions that the HFEA reached. After these first four chapters the report moves swiftly to its conclusions without any demonstration of balancing or considering the relevant arguments. The ethics literature review, for example, is highly selective failing to refer to many key texts.


As is now well known, the HFEA has come out strongly against all but strictly therapeutic uses of sex selection; and by “therapeutic” is meant uses which prevent the passing on of sex linked disorders. Let us start with the inconsistency involved in this exception. The first thing to note is that the HFEA was very cautious. It ruled out one method of sex selection, namely flow cytometry, because “It is not possible to discount a theoretical risk to health with the use of this technique”. Of course one will never discount risks if this absurdly high standard of caution is employed. Such a standard would rule out the benefits of almost all medical procedures, because there is always a theoretical risk, if not an actual risk, however slight. It may be that this method of sex selection should not be used on safety grounds, not because there is “a theoretical risk to health”, but surely because there is a real and significant risk.

The HFEA then goes on to consider so called “gradient methods” of sex selection and notes that “there is no reason to suspect that gradient methods pose a significant risk to the health of offspring”. We should note in passing, of course, that even this certificate of risk free health given by the HFEA would fail its previously employed rigorous test, namely that “it is not possible to discount a theoretical risk to health”. Since it is never possible to discount such a risk, if this is an objection it applies to gradient methods as powerfully as it does to flow cytometry … however, let that pass. What is clear is that the HFEA regards even minimal risks to the health of resulting children which may flow from risks inherent in the methods of sex selection (rather than any social or psychological outcomes) as decisive in rejecting sex selection except where it would be used to rule out the inheritance of sex linked disorders. Now this is a startling conclusion, since at paragraph 129 the HFEA states: “The risk of passing on a serious sex-linked genetic condition is a good and, other things being equal, sufficient reason for prospective parents to be offered the options of sex selection”. Note that the risks inherent in the procedures of sex selection will still be present where the purpose is to eliminate sex linked disorders, and so the HFEA’s argument seems to be that it is reasonable for parents to expose their future children to risks to their health because the alternative for these parents is to expose different children4 to greater risks. The HFEA avoids consideration of a third alternative in this context, which is that parents do not need to expose children to significant risks at all. They can use embryo selection or abstain from reproduction. However, since the avoidance of greater risks to different children is not something that can benefit the particular children who will be born as a result of the sex selection, it surely cannot justify exposing these children to risk if the risks are unacceptably high. To talk of greater risks to different children is slightly problematic. If no sex selection is used either boys or girls may result. Sex selection to eliminate sex linked disorders usually tries to eliminate males since they are more likely to be affected. The interesting question is who benefits? The class of children saved from “risk” by sex selection are at risk of being born with a serious disease. But in some cases such an existence may still be preferable to non-existence. So they may be saved from the risk of disease at a greater (almost 100%) risk of non-existence. In such a case it might be argued that only the parents and society benefit from this risk avoidance strategy and that the welfare of the child to be born has no place in the calculation.

Compare using one child as a bone marrow or even a kidney donor for a sibling. This would equally be a case in which a child would be required to run risks rather than expose a different child to greater dangers, but I doubt that would be thought obviously consistent with arguments requiring the welfare of the child concerned to be taken into account.5 I assume that the welfare of “the child who may be born” refers to the welfare of the particular child calculated to be the product of the combination of choices and technology used. The only possible alternative understanding of the meaning of the phrase “welfare of the child to be born”, sees it as requiring a eugenic programme for reproduction aimed at producing the best of all possible children in the circumstances. This would be an altogether different project, and one, I would suggest, that is even further from the project of the framers of the Human Fertilisation and Embryology Act or those who voted for it.

The only way to sustain the idea that seems to be in the heads of the members of the HFEA is to argue that although exposed to greater risk, such risks are in the interests of the child exposed to them because it is that child’s only chance of existence.

It is as if the future children have been offered a bargain: “Here’s the deal, you have a chance of coming into existence but only if you accept greater than normal risks—take it or leave it!”. A rational embryo or would-be embryo would take the deal, because the alternative is non-existence. This is the only appeal which makes sense in terms of the interests of “the children who will be born”; but I doubt if the HFEA would wish to endorse it because then it would have to do so in the simple sex selection case.6

We should note that this way of thinking of things does not involve the attribution of interests to non-existent beings (although I see nothing in principle wrong with such an attribution). We may translate the hypothetical deal I have described as if it were put post facto to existing children. We say to them: “Ok you exist, but at greater risk than would have been required for other kids to exist. Was it worth it? Was it a good deal?”. I imagine they would answer “Yes” unless life for them was not worth living.

Now of course there is a sense in which the HFEA is quite right. The Authority wants to say that for parents at risk of producing a child with a sex linked disorder there is an important therapeutic advantage in sex selection. Given that they are going to procreate it enables them to have a child with less risk of malformation or disease than available alternatives. For this choice to be ethical we have to judge the risk involved in the sex selection procedure so small as to justify it in terms of dangers to the resulting child, unless we appeal to the argument that asks what a rational embryo would choose—the so called “non-identity” argument. This argument, invented by Derek Parfit, shows that reproductive choices which select the child to be born cannot harm that child or do other than promote the child’s welfare unless they create a child with a life not worth living.7 Remember in the case of a genetic link with sex linked disorder the parents get a child without having to risk having a child with a sex linked disorder, which by hypothesis they do not want. But that is the same in the simple sex preference case. In that case too sex selection gives the parents a chance to have a child, who is free of a condition (male of female), which the parents do not want. Of course the parents in one case have a more pressing or serious justification according to some. But this too is a matter of judgement of a considerably problematic nature, for in neither case do the parents have to procreate. They can abstain. The alleged case of “necessity” is predicated upon the procreative imperative. But a lot more argument is needed to show that the imperative involves mere procreation as opposed to chosen procreation. Which brings us to the HFEA’s other arguments concerning choice and to reproductive liberty and the democratic presumption.


At paragraph 132, the HFEA sets out and commits itself to what may be called the democratic or the liberal presumption; the Authority expresses it thus:

The main argument against prohibiting sex selection for non medical reasons is that it concerns that most intimate aspect of family life, the decision to have children. This is an area of private life in which people are generally best left to make their own choices and in which the State should intervene only to prevent the occurrence of serious harms, and only where this intervention is non-intrusive and likely to be effective.

This is a firm and consistent statement of one of the presumptions of liberal democracies; that the freedom of citizens should not be interfered with unless good and sufficient justifications can be produced for so doing. The presumption is that citizens should be free to make their own choices in the light of their own values, whether or not these choices and values are acceptable to the majority. In this report, however, the HFEA simply surrenders to the hostility to sex selection of a majority (not of citizens, but of respondents to a consultation which necessarily samples a tiny fraction of the population) and gives, in the end, no weight to this important liberal principle and presumption underlying all democratic societies. We will examine the rather impoverished reasons why the HFEA has abandoned democratic principles but before doing so we need to consider substance of the idea so often now referred to as “reproductive liberty” or “procreative autonomy”.


When people express their choices about procreation they are claiming an ancient, if only recently firmly established, example of what may be termed a “fundamental right”. This right or entitlement is found in all the principle conventions or declarations of human rights. Sometimes it is expressed as the right to marry and found a family, sometimes as the right to privacy and the right to respect for family life (see the United Nations Universal Declaration of Human Rights Article 16, 1978, the European Convention on Human Rights, Article 8 and Article 12, 1953, and the International Covenant of Civil and Political Rights Article 23, 1976). This right or entitlement is often discussed in terms of “reproductive liberty” or “procreative autonomy”.

The right or entitlement to reproductive liberty has a number of different sources and justifications. Some see it as derived from the right to reproduce per se, others as derivative of other important rights or freedoms. Certainly there is no widespread agreement as to the nature and scope of this right; however, it is clear that it must apply to more than conventional sexual reproduction and that it includes a range of the values and liberties which normal sexual reproduction embodies or subserves. For example, John Robertson outlining his understanding of this right suggests8: “The moral right to reproduce is respected because of the centrality of reproduction to personal identity, meaning and dignity. This importance makes the liberty to procreate an important moral right, both for an ethic of individual autonomy and for the ethics of community or family that view the purpose of marriage and sexual union as the reproduction and rearing of offspring. Because of this importance the right to reproduce is widely recognised as a prima facie moral right that cannot be limited except for very good reason.”

Ronald Dworkin has defined reproductive liberty or procreative autonomy as “a right to control their own role in procreation unless the state has a compelling reason for denying them that control”. “The right of procreative autonomy has an important place … in Western political culture more generally. The most important feature of that culture is a belief in individual human dignity: that people have the moral right—and the moral responsibility—to confront the most fundamental questions about the meaning and value of their own lives for themselves, answering to their own consciences and convictions ... The principle of procreative autonomy, in a broad sense, is embedded in any genuinely democratic culture.”9

Julian Savulescu gives a classical twist to arguments about reproductive liberty suggesting that the core idea derives from an element in John Stuart Mill’s defence of liberty, which highlights the crucial role played by the freedom to experiment. Savulescu sets the idea out thus10: “Reproduction should be about having children who have the best prospects. But to discover what are the best prospects we must give individual couples the freedom to act on their own value judgement of what life constitutes a life of prospect. ‘Experiments in reproduction’ are as important as ‘experiments in living’ as long as they don’t harm the children who are produced. For this reason, reproductive freedom is important. It is easy to grant people the freedom to do what is agreeable to us; freedom is important only when it is the freedom for people to do what is disagreeable to others.”

Arguably Robertson’s, Dworkin’s, and Savulescu’s accounts all centre on what I believe to be the key idea of reproductive liberty, namely respect for autonomy and for the values which underlie the importance attached to procreation. These values see procreation and founding a family as involving the freedom to choose one’s own lifestyle and express, through actions as well as through words, the deeply held beliefs and the morality which families share and seek to pass on to future generations.11

Given that the freedom to pass on one’s genes is widely perceived to be an important value, it is natural to see this freedom as a plausible dimension of reproductive liberty, not least because so many people and agencies have been attracted by the idea of the special nature of genes and have linked the procreative imperative to the genetic imperative. Whether or not this suggestion is ultimately persuasive, it is surely not possible to dismiss the choices about reproduction and access to the relevant technologies which constitute the point of claiming reproductive liberty as a simple and idle exercise of preference. Reproductive choices, whether or not they prove to be protected by a right to procreative liberty or autonomy, have without doubt a claim to be taken seriously as moral claims. As such they may not simply be dismissed wherever and whenever a voting majority can be assembled against them. Those who seek to deny the moral claims of others (as opposed, possibly, to the exercise of their idle preferences) must show good and sufficient cause. We do not, to paraphrase Ronald Dworkin, allow the majority to determine what religion others are permitted to follow because we rightly regard such matters as issues of personal liberty and fundamental rights.


In the paragraphs following paragraph 132 the HFEA rehearses many of the considerations that were adduced in the public consultation and it is not clear always whether it is endorsing these or merely repeating them. The HFEA rightly dismisses considerations that sex selection may produce a gender imbalance and relies in effect on the following few, and I believe totally inadequate, considerations.

The first is set out in paragraph 139 where it states “In our view the most persuasive arguments for restricting access to sex selection technologies, beside the potential health risks involved, are related to the welfare of the children and families concerned”. The HFEA then glosses this concern for children by noting that, “Children selected for their sex alone may be in some way psychologically damaged by the knowledge that they had been selected in this way as embryos”. This is a very tendentious and unwarranted way of putting the point. The HFEA produces no evidence, nor indeed could it produce any evidence, that children would be selected for their sex alone. This is the point which derives from Kantian ethics, that individuals must be treated as ends in themselves and not as mere means. However, it is very difficult to find evidence or even persuasive anecdotes that if people are treated as means they are treated as mere means or exclusively as means. It is very unlikely that children selected for their sex would be selected solely for their sex. Indeed it is difficult to understand what that might mean. Also the idea that even if this were the case, they would be so unloved and treated so unacceptably badly that it would cause psychological damage is a piece of reckless speculation for which no evidence is produced and indeed no evidence could be produced. The HFEA then repeats (or is it endorses?) two other considerations, namely “That such children would be treated prejudicially by their parents and that parents would try to mould them to fulfil their (the parents) expectations. Others saw a potential for existing children in the family to be neglected by their parents at the expense of sex selected children.” Well, these so called dangers may be theoretically possible, but they are hardly realistic. Suffice it to say that for these highly speculative and fanciful dangers (for which no evidence is produced and indeed for which so far as I am aware no evidence exists) to count against the powerful formulation of the liberal imperative would be effectively to deny that imperative any weight or role at all; and indeed this is precisely what the HFEA has done, because in paragraph 147, which is the final statement of the HFEA’s justifications for rejecting sex selection and indeed for proposing legislation against it, it states the following (my italics):

In reaching a decision we have been particularly influenced by the considerations set out above relating to the possible effects of sex selection for non medical reasons on the welfare of children born as a result, and by the quantity of strength of views from the representative sample polled by MORI and the force of opinions expressed by response to our consultation these show that there is very wide-spread hostility to the use of sex selection for non medical reasons. By itself this finding is not decisive; the fact that a proposed policy is widely held to be unacceptable does not show that it is wrong. But there would need to be substantial demonstrable benefits of such a policy if the State were to challenge the public consensus on this issue.

Thus the powerful statement of the democratic presumption at paragraph 132 that “the State should intervene only to prevent the occurrence of serious harms” has been converted into the requirement that “there would need to be substantial demonstrable benefits”. Here not only has the democratic presumption been turned on its head, but the burden of proof has entirely shifted from the requirement that the State show that its interference is necessary to prevent the occurrence of serious harms to the rather feeble requirement that those who wish to exercise liberty must qualify for this freedom by showing that its exercise provides substantial demonstrable benefits. If this is to be the case liberty is meaningless and the presumption of liberal democracies is overthrown.

This is effectively what the HFEA report recommends. The illiberalism of this conclusion and the poverty of the arguments produced to defend and sustain it make it imperative that this report is not only rejected but that its conclusions and recommendations be recognised for what they are, namely an attempt to formalise the tyranny of the majority and to institutionalise contempt for the principles of liberal democracy.

Sex selection: options for regulation. A report on the HFEA’s 2002–03 review of sex selection including a discussion of legislative and regulatory options. A critique


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.