Research Paper 23 2000-01

Is it Medically Legitimate to Provide Assisted Reproductive Treatments to Fertile Lesbians and Single Women?

Maurice Rickard
Social Policy Group
27 February 2001

Contents

Major Issues

Introduction

The Notion of a 'Medically Legitimate Intervention'

Proposition 1: An intervention is medically appropriate only when it is the only effective option

Proposition 2: The domains of medical responsibility and of personal responsibility are distinct

Proposition 3: The purpose of medical interventions is to repair conditions

Proposition 4: Medical treatments are for medical conditions

Conclusion

Appendix 1: The distinction between medically relevant factors and other relevant factors

Endnotes

References

Major Issues

In the current public debate, there have been two key points of dispute in the question of who should have access to medically assisted reproductive treatments like IVF and artificial insemination: the issue of marital status, and the issue of whether a woman (or a couple) should have to be medically infertile to access these treatments. The former issue has generally had greater prominence. The latter issue is nonetheless crucial to the fate of a significant range of lesbian and single heterosexual women who are childless not because of a medical or physiological impediment to pregnancy, but because of the personal or social circumstances of their lives.

There have been a number of objections in the public debate to allowing 'socially infertile' women access to assisted reproduction. The objections have focused on the quality of lesbian couple and single-mother parenting, and the interests and welfare of the unborn child.⁽¹⁾ However, another issue, one which is absolutely fundamental, is the question of whether it is ever medically legitimate to provide medical treatments like assisted reproduction to women who do not have a medical condition. Whatever the implications of any of these other arguments about parenting or the rights of the child, or rights to reproduce, or rights against discrimination, they are all academic if it turns out that it is simply medically illegitimate from the start to provide such treatments to such women. It is a question of some importance, therefore.

The question of what is medically legitimate is tied inextricably to the idea of what the real purposes of medicine are, and this is inevitably subject to ethical justification and rational defence. Arguably, all of the objections that have plausibly and credibly been advanced to show that assisted reproduction for socially infertile women is not medically justified will be variations on four key propositions about medical purposes. If any are valid, this would serve to exclude assisted reproduction as a legitimate intervention for socially infertile women. The propositions present arguments based on:

1. the availability of other non-medical alternatives for socially infertile women
2. the responsibility that socially infertile women have for their own childless condition
3. the inappropriateness of enhancing a lifestyle through medicine, rather than repairing an abnormality, and
4. the claim that medical treatments are for medical conditions (not social ones).

Even if these propositions and arguments are not always advanced in exactly the way they are stated here, the suppositions they make will often inform people's attitudes, as well as the public debate. Moreover, they tend to do this largely unnoticed, and unrecognised as the suppositions they are. Perhaps one reason for this is the fact that they are inevitably abstract and philosophical in nature, thus difficult to discern sometimes, let alone get a grip on. Notwithstanding this, it is crucial that these deep drivers of the debate are brought to light and examined. Despite their philosophical and socio-ethical nature, they do have public health policy impacts, particularly in the area of public funding. Such impacts are not always easy to trace back to their philosophical source. The challenge is to lay bare these sources and critically examine them.

When the various claims, suppositions and arguments underlying the four propositions above are pinned down and critically examined, they appear to turn out to be much less compelling than they might have seemed initially. The conclusion that emerges is that there is insufficient reason to exclude socially infertile women from assisted reproduction on medical grounds. This is not a blanket justification for allowing socially infertile women access. There may still be other non-medical grounds for excluding them (grounds which will not be entered into here). But it is nonetheless an important and substantial conclusion, particularly since it has direct bearing on the question of whether assisted reproductive services for socially infertile women should be publicly subsidised through Medicare.

Introduction

The recent legal challenge to the Victorian Infertility Treatment Act 1995,⁽²⁾ and the Federal government's subsequent Sex Discrimination Amendment Bill (No. 1) 2000 have given rise to extended public debate about who ought and ought not to gain access to medically assisted reproductive treatments (ART), such as IVF and artificial insemination.⁽³⁾ There have been two key points of dispute in the debate: the first relating to marital status as a condition of eligibility for ART, and the second relating to medical infertility as a requirement.

This paper focuses on the latter concern, specifically, whether access should be granted to women who strongly desire a child, and who are not medically infertile,⁽⁴⁾ but who are very unlikely to conceive through heterosexual intercourse either because:

• their sexual identity or preferences preclude this (i.e. fertile lesbians), or
• they are unlikely to find or meet an acceptable partner for intercourse or conception, or the fathering of their child before their safe reproductive years end (i.e. fertile single heterosexual women).

For these two groups of women, their childlessness is not due to a physiological or medical condition, but to personal or lifestyle preferences or life circumstances. They have consequently come to be referred to as 'socially infertile' women.

It is fairly apparent how lesbians and lesbian couples might count as socially infertile. But there may be a number of different circumstances in which single heterosexual women might be socially infertile. Often, they will be women who are older and experiencing a degree of urgency associated with their limited remaining reproductive years. Some may be childless at this stage simply because they have not experienced relationships that present the potential for conception, despite actively seeking or being open to these relationships. Others may have had such opportunities, but have judged that none of the relationship options open to them have, in the end, been acceptable for conception. Others still may be childless because the life they have chosen to lead has precluded or minimised their opportunities for relationships (e.g. women involved in their occupation or career). Some single women may even be childless through the unexpected death of their partner, or through their partner's actions. David Molloy, a Brisbane obstetrician and Chairman of the IVF Directors Group, notes the example of:

Both lesbian and single heterosexual socially infertile women seek ART in Australia. Because they are physiologically fertile, the relevant treatment or technique they will seek will simply be medically assisted insemination, either with anonymous donor sperm or with that of a male they know. The exact extent to which socially infertile women seek, and are given access to assisted insemination in Australia is difficult to determine. No nationwide data is systematically collected, though informed estimates have placed the figure at 150 lesbians per year,⁽⁶⁾ and more for single heterosexual women.⁽⁷⁾ One recent Australian survey, however, has indicated that one in five lesbian respondents intend to become mothers in the next five years, and 16 per cent of these through medically assisted reproduction.⁽⁸⁾ In the United States, it has been estimated that 10 000 children conceived through medically assisted insemination have been born to lesbian mothers,⁽⁹⁾ and that an estimated 30 000 single women undergo donor insemination every year.⁽¹⁰⁾

Currently, in some Australian states (Victoria, Western Australia and South Australia) it is illegal to provide ART to women who are not infertile or at risk of transmitting a genetic disorder (unless their husbands or defacto partners are infertile). It is illegal in those states to provide ART to socially infertile women. In New South Wales, Queensland, ACT and Tasmania, however, there is no legislative restriction relating to the provision of fertility treatments,⁽¹¹⁾ only guidelines provided either by the National Health and Medical Research Council (NHMRC), or by state bodies or by individual clinics. The NHMRC guidelines are the most widely adopted, and they are currently silent on the issue of access by fertile single women and lesbians.⁽¹²⁾ Socially infertile women seeking ART are likely to seek it in these states, engaging in what has been described from the similar European experience as 'reproductive tourism'.⁽¹³⁾ Public attitudes to the idea of lesbians and single women accessing ART appear to be divided, with a substantial minority approving of it, but the majority is resistant to the idea.⁽¹⁴⁾ There is also ambivalence among sperm donors as to the acceptability of lesbian and single women recipients.⁽¹⁵⁾ The current public debate on this issue has no doubt been sustained by these divisions of opinion.

In the current debate a number of arguments have been advanced against allowing socially infertile women access to ART. In large part, those arguments have either questioned the quality of single and same-sex parenting, or else relied on the purported right of the child to be parented by a male. For instance, the Australian Catholic Bishops' application in October 2000 to challenge the McBain v State of Victoria decision is avowedly motivated by a desire to give unborn children a voice.⁽¹⁶⁾ Of course, there are also blanket arguments against providing medically assisted reproduction to anyone, physiologically infertile or not.⁽¹⁷⁾ But these arguments have been thoroughly debated already, leaving ART a generally accepted medical practice in Australia.⁽¹⁸⁾

What has been much less debated, however, and which is of particular interest here, is a further set of arguments specific to social infertility which, if they were successful, would have considerable potential to exclude socially infertile groups. These arguments approach the issue of social infertility from the point of view of what constitutes a medically acceptable practice, or a medically legitimate intervention. If providing medically assisted artificial insemination to physiologically fertile women falls outside the legitimate scope and purpose of medical practice or clinical activity, then this would be sufficient ground to exclude these women, no matter what the force of the other arguments for or against their access. Being properly within the domain of legitimate medical concern is a logically necessary condition for justified access to medical treatments and interventions. Such arguments, therefore, are of particular importance.

This paper seeks to explore and test the view that providing medically administered artificial insemination to socially infertile women is never a medically justified or legitimate intervention. This view is likely to be one that many would very readily agree with, or perhaps even take as self-evident. The arguments that might support it, however, are less readily produced and are not always clearly stated when they are. There may be a number of possible reasons for this, including perhaps the perception of self-evidence. Nonetheless, it can be contended that those arguments based on 'medical' reasons that either have or could plausibly be presented against providing ART to socially infertile women will all be variations on the following four key propositions. Each of these propositions purports to express a central property or characteristic of medically legitimate interventions, in terms of which ART for the socially infertile is argued to be medically illegitimate. The propositions do not all necessarily comport with each other under the one conception of medical legitimacy. They may arise from different views. But any of them, if valid, would be sufficient to exclude socially infertile women from ART.

1. An intervention is medically appropriate only when it is the only effective option. It will not be legitimate when there are other effective non-medical options available. There are other such options through which lesbians and single women can address their childlessness. There is adoption, and, because they are still medically fertile, there is also the option of intercourse with a man.
2. The domains of medical responsibility and of personal responsibility are distinct. An intervention is only medically legitimate when it is not called upon to correct the foreseeable consequences of people's choices, or to compensate for life circumstances that have merely been a little unfavourable in certain ways. Social infertility is either a foreseeable consequence of lifestyle choice, or else a predictable outcome of perhaps unfavourable, but not debilitating or out of the ordinary, life circumstances.
3. The purpose of medical interventions is to repair conditions. Medical interventions correct problems in people's normal functioning. They do not enhance people's well-being or capacities beyond their level of normal functioning.⁽¹⁹⁾ Lesbian couples do not, as a matter of their normal functioning, have a capacity to reproduce. Giving them that capacity would not be repairing a dysfunction as much as enhancing their lives. Strictly speaking, the inability of socially infertile single heterosexual women to conceive is not due to a dysfunction they have.
4. Medical treatments are for medical conditions.⁽²⁰⁾ To act otherwise by recommending a physiological intervention to overcome a non-physiological cause of childlessness is to act outside the confines of sound medical practice.

Needless to say, these propositions and their associated arguments, will intersect at various points. It is true that not all of the claims and arguments that have been advanced in the public debate exactly mirror these propositions as they have been expressed here. However, it will be held here that they do adequately capture the range of arguments and positions that could plausibly and seriously be put forward.

The purpose of this paper will be to examine these key propositions to determine how well they can be sustained under close scrutiny. If none of them can be sustained, that will be grounds to conclude that there is insufficient reason to think that providing assisted insemination to socially infertile women is medically illegitimate by definition. It is important to note with this, though, that even if that conclusion were to emerge, this would not automatically mean that socially infertile women should always (or even ever) be granted access in the end. Even if it were quite legitimate from a purely medical point of view to provide assisted insemination to socially infertile women, it may well be that other relevant considerations, when they are all factored in, militate against allowing some or even any of these women access. These considerations might include some of the arguments cited above, for instance, or ones connected with costs and limited resources,⁽²¹⁾ or rights of the child, or wider impacts on social expectations and family structures. Important as these other factors and considerations are, they will not be entered into here. The focus will be squarely on the concept of a 'medically legitimate intervention', and what implications this might have for socially infertile women.

The Notion of a 'Medically Legitimate Intervention'

Each of the four propositions above purport to specify its own criteria for something to count as a medically legitimate intervention. Each criteria would, if valid, disallow medically assisted insemination in the case of the socially infertile. To see if any of the four proposed criteria are indeed valid it will be necessary to look into what it might mean for an intervention to be 'medically legitimate'. It should be kept in mind that what is in question is medical legitimacy-what is medically acceptable, not what is medically necessary, which is a stronger notion concerning what should, or must, or needs to be done from the point of view of medicine.

It is important not to confuse two senses in which an intervention might not be medically legitimate. In one sense, the intervention might be medically illegitimate because it involves a treatment or procedure that has a very low degree of success compared to other available procedures. In the other sense, the treatment or procedure involved is effective, but it is medically inappropriate or unjustified for other reasons to provide it as an intervention in a particular set of circumstances. To illustrate the distinction between an effective treatment and an acceptable intervention, consider the case of amputation. Certain surgical techniques and procedures might be effective for the removal of a person's hand, and performing those procedures may well be considered a medically legitimate intervention if the hand is seriously damaged or diseased. But it will probably not be considered medically legitimate to apply those procedures to remove a person's (healthy) hand as a form of punishment for theft. (Many would count that an abuse of medicine).

Given that artificial insemination is a relatively effective means of conceiving, it is the latter sense of illegitimate intervention that is relevant here-one that is unjustified or inappropriate to provide for other reasons. If the other reasons alluded to here are not ones purely to do with the medical or instrumental effectiveness of the procedure, then what sort of reasons would they be? Arguably, they will relate to the point or purposes of applying medical interventions, and whether an application of the procedure in a particular situation conforms to those purposes. Without venturing too much into the question of what the ultimate purposes of medicine are, there are two observation that are worth making in this connection. The first is that medicine, as a practice, is a social practice: it both affects society in certain ways (e.g. heightens levels of personal and social well-being, influences norms and social expectations, uses social resources),⁽²²⁾ and its conduct is in turn affected by society (e.g. by changing social attitudes and regulations, by decisions about resource allocations, etc.). The perceived role of medicine has changed through modern history, and the accepted applications of artificial insemination, in particular, can be seen to have changed over the last century in relation to changing social circumstances and mores.⁽²³⁾

The second observation is that the purpose(s) of medicine are not neutral, or 'intrinsically' determined, or derivable from something like a purely scientific analysis of medical knowledge or procedures. Of course, the intrinsic possibilities and limitations of medicine as a discipline will shape and circumscribe what these purposes can sensibly be. But in their central respects, the purposes of medicine are value-based and normatively derived. When one states the goals or purposes of medicine, one is stating what medicine and medical interventions should rightly and properly be used for. To specify the medical goals of an intervention is to give the medically relevant reasons for intervening. To a certain extent, it will be a matter of rational analysis and ethical deliberation as to how these purposes, and the associated notion of 'medical legitimacy', ought or ought not be defined.⁽²⁴⁾

The upshot of these observations is that the question of what makes an intervention medically legitimate is implicitly a socio-ethical one. It is not something that is set in concrete. As the medical historian-sociologist, Simone Bateman Novaes observes, it is 'a practical normative construct in which medical and social justifications are woven together'.⁽²⁵⁾ The notion of medical legitimacy will be influenced by the social norms and practices in which it is embedded (and, of course, it is most deeply embedded in health care and medical practice). It will also be subject to disagreement, with different social groups laying claim to a role in its definition.⁽²⁶⁾ These facts, though, do not mean that all conceptions of medical legitimacy are of a kind, with none having greater validity than any other. Some views may well be more defensible than others, and all are subject to independent ethical or rational scrutiny and, perhaps, revision. This, too, goes for the four propositions above. Appendix 1 elaborates on what underlies the distinction between medically relevant factors and other relevant factors in deciding whether an intervention should go ahead.

The validity of the criteria and supplementary arguments advanced in these propositions can be tested in a number of ways: by looking at their consistency, their presuppositions, and their underlying rationale. One useful way of testing consistency is to seek possible counter-examples to the criteria. That is, to seek instances of treatment interventions that we would, on careful and considered reflection, still be strongly inclined to call medically legitimate, even though the proposed criteria would definitely exclude them. It is also possible to scrutinise the assumptions the propositions make. It may turn out, on closer inspection, that some of those assumptions misconstrue the nature or circumstances of social infertility. Consequently the proposed criteria may simply miss their target. As well as this, it is particularly important to examine the underlying rationales and argument for each criteria, to test their strength and plausibility. The rest of this discussion applies these tests to the four propositions in turn.

The ineluctably socio-ethical nature of all these issues makes it inevitable that the discussion will delve into some of the abstractions, complexities and philosophical under-pinnings of medical justification. But without doing that, we could not be in any position, at the end, to comment on whether there are any solid medical reasons for excluding socially infertile women from assisted reproductive treatments.

Proposition 1: An intervention is medically appropriate only when it is the only effective option.

The Argument

Clearly, some personal conditions or problems can only be effectively overcome or addressed by medical means (for example, a ruptured spleen). But there are also some conditions or problems that can plausibly be addressed through non-medical as well as medical means. Some examples might include chronic obesity,⁽²⁷⁾ where diet change, exercise, and behaviour modification are options as well as pharmacotherapy or stomach surgery; and sleeplessness, where again, behaviour modification and relaxation techniques are options as well as medically prescribed hypnotics and sedatives.

There are two components to Proposition 1. The first is the view that it is only legitimate to apply a medical treatment when that treatment is the only option that has a reasonable chance of bringing about the desired outcome. The second is the supposition that fertile lesbians and single heterosexual women can either adopt children or, because they are fertile, can still engage in heterosexual intercourse to conceive in the traditional fashion, or else self-inseminate with sperm obtained from men known to them. They therefore have reasonably effective alternatives to medical intervention, and ex hypothesi, this means it would be illegitimate to provide medically assisted insemination in their case. On close examination, though, both of these components of Proposition 1 turn out to be questionable.

Medicine in the context of other effective approaches

Consider the view expressed in the first component. Initially, that view appears to have a ring of plausibility about it, particularly in the case of assisted reproduction. With physiologically infertile women/couples, treatment is usually only seriously entertained as a legitimate option when it becomes apparent that the other available option (regular heterosexual intercourse) is not effective.⁽²⁸⁾ Certainly, when there are no other reasonably effective non-medical alternatives, and assisted reproduction is the only effective means of conceiving, it does seem plausible that medically assisted reproduction would be a legitimate intervention in the circumstances. But, importantly, the converse does not necessarily follow. It is not at all obvious that when there is an alternative to medical treatment for a condition, that fact automatically renders the treatment illegitimate in the circumstances, or completely unacceptable as an option. Sure enough it would be odd to choose a medical intervention (which can often be invasive, uncomfortable and expensive), when there are other easier or more effective non-medical alternatives. But the fact that such alternatives exist arguably does not completely disqualify the medical intervention from being a legitimate option to be counted and considered along with the rest. It does not make the intervention medically improper, or an abuse of medicine. Moreover, it will very often turn out that the medical option is the most viable and effective in the circumstances. Here, the medical option would not only be an acceptable and legitimate intervention, it would be the most appropriate one. It would be odd not to choose it.

These few observations are corroborated by our current uncontroversial health care practices. Generally, we are quite happy to accept people having access to medical treatments to overcome conditions that can also be remedied in other ways. Take the simple examples of insomnia and obesity noted above, with the prescription of therapeutic drugs, rather than undertaking behavioural or life-style change.

So, it is not at all clear that the mere fact there are other possibilities in itself makes the option of medical treatment or assistance improper, or renders those who want to benefit from that assistance ineligible. This is not to suggest that the existence of other options plays no role whatsoever in therapeutic decision-making. The nature and effectiveness of the other possibilities may well be important in determining whether the medical option is the best one to choose in the circumstances. Also, the existence of other options might play a role in determining who should get priority access to medical treatment, a limited and sometimes costly resource. For instance, it is not implausible to argue that those who have no viable and effective alternative other than the medical to address a problem should be granted priority over those who do have safe and effective alternatives.

Consider now the second aspect of Proposition 1. Even if it turned out to be true that the use of medical treatments is improper when there are other options, it is not entirely clear that this would be relevant to the case of socially infertile women. The supposition, it will be recalled, was that fertile lesbians and single heterosexual women already have the (non-medical) option of adoption, or self-insemination, or heterosexual intercourse with a man just in order to achieve conception. Sure enough these are options, but it is not as sure a thing that they will be viable or realistic ones. The reasons for this are as follows.

Adoption and self-insemination as alternatives to ART

Single and lesbian women tend to face as many obstacles in adopting children as they do in accessing ART. Currently under Australian law there is no provision for joint adoption applications from same sex couples.⁽²⁹⁾ This makes adoption a less than realistic option. Moreover, the overwhelming desire these women have is often for a genetically related child. It is true that this could be achieved through self-insemination with sperm donated by male friends. But without the donated semen being medically screened for defects and transmissible diseases (including HIV), there are inherent dangers in self-insemination. Self insemination is also illegal in some jurisdictions, including Victoria, where it attracts a penalty of up to 4 years imprisonment.⁽³⁰⁾ In view of these reservations, self-insemination is not an acceptable option either. This leaves heterosexual intercourse as the remaining alternative to medically assisted insemination. How viable and realistic is this as an option for socially infertile women?

Sexual intercourse as an alternative to ART: Socially infertile heterosexual women

Take the case of single heterosexual socially infertile women-those who do not have the child they want because they have not had (and are unlikely to have) the opportunity to have sexual intercourse (with a view to conception) with a male with whom they consider this appropriate and acceptable. As an alternative to assisted insemination, it has been suggested these women still have the option of having intercourse with men, even though they will be men whom they would otherwise consider it unacceptable to have intercourse or conceive with.⁽³¹⁾ There are significant problems with this suggested alternative. Having penetrative sex is a deeply intimate act, and having penetrative sex (on perhaps a number of occasions) with a man with whom one considers this inappropriate and unacceptable, might not merely be distasteful or uncomfortable, but may well be deeply offensive or even traumatic. In view of this, it could not be uncontroversially considered a viable option. Added to this, the few studies available indicate that women in these circumstances sometimes have concerns about it being morally questionable, or even mercenary, to temporarily involve a man with whom they have no other personal involvement.⁽³²⁾

At this point it might be replied that these single heterosexual women may have ended up single and childless because they have set their sights too high in seeking an acceptable relationship. They ought to set about changing their preferences and standards to match their circumstances and the relationship/intercourse possibilities that are realistically open to them. But this suggestion does not seem compelling, either. Choosing sexual partners is again a deeply personal matter, and subject to a considerable degree of inter-personal complexity, as well as the vicissitudes of circumstance, not all of which are within a person's control. Sure enough, it is not impossible to change one's relationship preferences and standards, but the more deeply held they are and the less they are generated by rational intellectual considerations (as opposed to perhaps perceptual or even instinctual factors), the less likely they are to be straightforwardly amenable to deliberate modification. What is more, to the extent that they can be modified, the strong desire on the part of these women to have a child would itself have probably already acted as a force to moderate their standards in sexual partners, or men to conceive with. As an indication of the frustration that is sometimes felt by these single heterosexual women at the suggestion that they should simply try harder to find a partner, Leesa Meldrum, whose attempts as a single woman to gain access to fertility treatment in Victoria were the subject of the case of McBain v State of Victoria, has recently commented:

In view of all this, this supplementary suggestion is seriously questionable.

Sexual Intercourse as an alternative to ART: Socially infertile lesbians

What about lesbian women/couples? Is having penetrative sex with a man any more viable an alternative for them than for heterosexual socially infertile women? Arguably not. For many lesbians, their sexual and emotional orientation is a deep and inescapable fact about their life and person. It may even be considered by them an identity defining characteristic. And for these women, having penetrative sex with a man may be more than just deeply offensive. This is not to ignore the fact that some lesbians might, in the absence of any alternative, resort to sexual intercourse with a man they know in order to conceive.⁽³⁴⁾ However, it would not be clear why a such an option which is contrary to these women's deep preferences should be given credence over the much less offensive (and more strongly preferred) option of medically assisted insemination.

It should be observed also that it is usually lesbian couples who seek assisted reproduction. The prospect of distress or offence on the part of the lesbian partner needs to be taken into account as well. This tends to further weaken any perception that the proposed alternative of sex with men is a real and viable one for lesbians. This is reflected in some of the reported attitudes of lesbian women who choose assisted insemination because they do not want to violate their fidelity by sleeping with a man, nor introduce a third party into their family plans.⁽³⁵⁾ And, in case there is still some residual doubt about the suggested alternative, the question can be asked as to why the same option should not also be expected of heterosexual women whose male partners are infertile. If it were to be expected of lesbian couples but not heterosexual ones, that one partner should just sleep with someone else, in what would the medically relevant difference consist?

In all, Proposition 1 is less than convincing, as are the arguments it offers to exclude socially infertile women from ART.

Proposition 2: The domains of medical responsibility and of personal responsibility are distinct.

The Argument

This proposition concerns the issue of what properly belongs to the domain of personal responsibility as opposed to medical responsibility, and it is related in some ways to the previous one about alternatives. Underlying it, as its rationale, are two general and related (ethical) principles: that individuals should accept responsibility for their deliberate and conscious choices; and that individuals should also be expected to weather or endure the moderately unfavourable outcomes that life has to offer everyone from time to time. Accepting responsibility in these ways, according to this view, means that if people are willing to enjoy the benefits of their deliberate choices and the favourable ways life has turned out for them (i.e. their good luck), they should also be equally prepared to bear the 'costs' or burdens that might be consequent on making those choices, as well as moderately adverse life outcomes we all experience from time to time (i.e. their bad luck). Proponents of Proposition 2 would argue that, at bottom, socially infertile women are childless because of the choices they have made, or simply because of their mere bad luck in the social lottery. Their childlessness is a burden that they simply ought to endure, and not something that it would be legitimate to rectify through medically assisted insemination. The following paragraphs fill in more of the arguments behind these views.

Women who actively, deliberately and in full knowledge choose to live a lesbian lifestyle will be aware that, in the normal course of things, a foreseeable consequence of this is not being able to have one's own children. In having seriously made the decision to live a lesbian lifestyle, these women will probably have made their assessment of all the competing considerations, and will have judged that, for them, it is better to live that lesbian lifestyle even though it will probably mean foregoing children, than to have children through a heterosexual partnership and deny their real sexual preferences. Ending up with no children can be seen as a predictable trade-off that lesbians will have to make, or a risk they take, in deliberately choosing and pursuing the life they want and in which they find great benefit and satisfaction. The same applies to single heterosexual women who are childless at a late age because of the career choices they have made, for instance, choices which have left them no time or opportunity to develop relationships appropriate to conceiving or raising children. Just as was the case with lesbians, childlessness is an outcome these single women could have avoided if they had chosen to live differently. As autonomous adults, they are responsible for the decisions they make and the risks they take. Their childlessness is a condition they have brought about or contributed to through their own decisions, and is thus a burden they should properly bear and accept themselves. Given this, it would be improper for medicine to intervene to change that through assisted conception.⁽³⁶⁾ To do so would be to act outside the proper domain of medical responsibility.

This line of reasoning can also be extended to those single heterosexual women who have not borne children because the relationships prospects they have encountered have not been acceptable to them or have not met their personal standards. To the extent that personal standards are things that are chosen, or within conscious control, these women are again arguably responsible for the consequences of their choices. The personal responsibility argument also applies to single heterosexual women who have not borne children because they have not encountered much in the way of any real or enduring relationship prospects. Even if it is true that these women have not contributed to their condition through the choices they have made, and they have simply been unfortunate in not getting what they desire, it is not the purpose of medicine to correct for the misfortunes of those who miss out in the social lottery. This is especially so when not finding an acceptable partner is not an extraordinary or debilitating misfortune, and is something that everyone is generally at risk of.

Persuasive as these arguments about personal choice and responsibility might seem, they harbour significant weaknesses at a number of points. Firstly, they make questionable assumptions about the nature and circumstances of social infertility; secondly, they are not consistent with our other broader and considered views and practices concerning what is medically appropriate; and thirdly, the underlying rationale for the arguments (the general principles about personal responsibility) turns out not to apply to the issue of medical legitimacy and social infertility in quite the way it has been proposed. These weaknesses can be explored in turn.

Social infertility as the outcome of choice

As some of the observations made in earlier pages suggest, it is quite questionable that the childlessness of many socially infertile women is a result of the voluntary choices and decisions they have made in life. It was noted before that the deep psychological factors and 'standards' that play a role in the formation of people's personal relationships (i.e. the particular people they become attracted to, what they emotionally respond to in other people, and what desires and needs they have) are often not straightforwardly amenable to conscious scrutiny and are not usually thought of as things we choose. Given this, when people fail to form or develop personal relationships suitable to conception, those failures can not unequivocally be thought of as due to choices they have consciously made. Nor can the consequences of those failures, including childlessness. It would arguably only begin to be plausible to suppose this in the case of women who have freely and knowingly excluded the opportunity to develop such relationships because of the way they have consciously planned their lives (and where they were free to have chosen otherwise).⁽³⁷⁾

With respect to lesbianism, although it can certainly be a matter of choice or preference whether a woman has sex with another woman, it is not simply in terms of mere sexual activity that women define themselves or identify as lesbian. That identity is tied to their deeper affective dispositions. Even if people could, with some effort, regulate or control their emotional states, their affective dispositions-the underlying psychological (and perhaps biological) characteristics that dispose them to feel and perceive relationships in certain ways rather than others-are arguably less within conscious control. For those women who have committed themselves to a lesbian lifestyle, their lesbian identity is not likely to be something that is subject to choice, as much as being the perspective from which their other choices are made. It is broadly agreed among moral and legal philosophers that we can only begin to be held accountable or responsible for the outcomes of our actions and decisions if they are choices we have freely made between genuinely available alternatives.⁽³⁸⁾ It is not clear that denying one's identity to have children is a genuinely available alternative, and that there is any option other than to acknowledge and respond to one's deep and compelling self-perceptions, and face the prospect of childlessness. The bioethicists Tom Beauchamp and James Childress make a similar point in the broader context of health care provision:

A strong case needs to be made-much stronger than the one presented-that a lesbian's childless condition is an outcome that she could freely and with integrity have avoided.

So, it still remains to be shown that the childlessness of many socially infertile women is a matter of choice and personal responsibility. But even if that had been shown, there would still be problems with these arguments. For one, it is not clear that those arguments based on choice and responsibility would be confined only to socially infertile women, as the observations of Sheryl de Lacey bring to light:

Choice, personal responsibility and medicine: Some counter-examples

There is also a substantial question as to just how accurately the arguments about choice and personal responsibility underlying Proposition 2 reflect the notion of medical legitimacy. This can be seen when they are applied more broadly to other cases in the medical context. Those arguments, if valid, would serve to exclude from medical treatment all diseases, illnesses and conditions that have arisen through people's free and deliberate actions-cancer from smoking, respiratory failure from drug use, broken bones from bungee-jumping ... the list would be extensive. And, if taken to their conclusion, could even exclude the 30 to 50 per cent of medically fertile women de Lacey identifies above. All of these implausible implications are certainly out of step with our careful and considered perceptions of what legitimately warrants medical concern. Even when conditions or injuries are clearly foreseeable consequences of our actions (the sport of boxing), or even deliberately sought (self-harm and attempted suicide), they are not thereby justifiably excluded from medical consideration. The same can be said for ailments and conditions arising from accidents, misfortune, and sheer circumstance. It would be entirely implausible to suppose that the likes of minor infectious diseases and broken bones-misfortunes that normally befall many people-should go medically untended because it is not the purpose of medicine to correct the misfortunes of life's lottery which everyone is generally at risk of. And it is arguably just as implausible to suppose the same for women who are childless because, through sheer bad luck, they simply have not encountered any enduring relationship prospects.

The redundancy of arguments based on personal responsibility

There is one final concern with the arguments associated with Proposition 2 above. The concern is a serious, but slightly elusive one. The views about personal responsibility appealed to in Proposition 2 are arguably subject to a crucial qualification which, when filled in, serves to undermine their efficacy in arguing against socially infertile access to ART. It does seem credible that mature, autonomous adults ought to accept the consequences of the choices they voluntarily and informedly make about their lives, but only, it can be argued, if those consequences are just and fair (or deserved) consequences. Exactly what the outcomes of our actions turn out to be depends very much on a myriad of factors, many of which are outside our control as individuals. These factors can simply be matters of random chance, but they will also include the influence of background social conventions, legal rules and institutional arrangements, not to mention the actions of others. When I overstay in a parking space, I am caught as a matter of luck, and fined as a matter of social procedure. There is arguably no moral compulsion on people to endure consequences and outcomes of their actions that are unjust, undeserved or unfairly imposed-either through chance, or through the influence of unjust and questionable rules or background social conditions.⁽⁴¹⁾ If I over-park for a short time and am fined, it is reasonable that I should expect to pay. But arguably not if the law is unfair and imposes a fine that is exorbitant and disproportionate, or if my car is clamped and confiscated without any warning at all.

So, even if the childlessness of socially infertile women were a consequence of their choices, whether it is something they simply ought to endure or not will depend on the fairness or justness of the background factors-social, cultural and personal-that have prevailed to actually make childlessness a consequence of the choices they have made. Importantly, this will include those social and professional norms and practices that act to exclude socially infertile women from opportunities such as medically assisted insemination, which would have enabled them to still have children. The important upshot of all this is that in order to determine whether socially infertile women are fully personally responsible for their childlessness (in the sense supposed by Proposition 2), we need to know antecedently and independently whether the rules, conventions and practices that exclude them from medically assisted insemination are just or fair. But now, if the justifiability (or otherwise) of this exclusion can be established independently, there would be no need to rely on any further arguments to do with personal responsibility. The case would have been shown already. The conclusion to emerge from all this is that, as it stands, the personal versus medical responsibility argument associated with Proposition 2 is incomplete. It presupposes a supplementary case. But once that case is provided and the arguments of Proposition 2 completed, those arguments simply become redundant.

Proposition 3: The purpose of medical interventions is to repair conditions.

The Arguments

Central to this proposition is the view that the sole object of medicine is the failure of people's normal functioning, and its purpose is to fix the causes and alleviate the consequences of that failure. As James Sabin and Norman Daniels characterise it:

According to this view, the aim of medicine is not to enhance people's capabilities to bring them equally into line with others' capacities, or to give them capacities they might not have had. It is simply to restore people to the range of capabilities they would ordinarily have had, in a world where it is normal for capabilities to be distributed unequally between people. And where the relevant incapacity itself cannot be repaired (i.e. restored as an ongoing capacity), the aim would be to correct the symptoms of that incapacity. Daniels argues that this distinction between treating a dysfunction (a loss of normal functioning) and enhancing existing 'natural' capacities or incapacities, captures the way we ordinarily think about acceptable and unacceptable medical interventions. For example, the distinction might explain why we may be more prepared to accept growth hormone therapy for children who are short because they have a growth hormone deficiency, than for children who have no deficiency but who, like many children in the normal run of things, are just naturally short. Similarly, it might explain why we are less inclined to condone giving prozac to someone who is just naturally shy, than to someone who is shy because of a diagnosable mental illness.⁽⁴³⁾ In the latter case, what would normally be an unshy disposition is inhibited by the illness, and therein counts as a dysfunctional shyness. In the case of natural shyness, however, the normal functioning view would acknowledge that 'many people are shy and withdrawn ... others are unusually outgoing and adept at making relationships' but it would argue that medicine 'is not designed to rectify the normal distribution of social skills, however much competitive disadvantage and suffering the lack of these skills might entail.'⁽⁴⁴⁾

This treatment/enhancement distinction might be applied to the case of social infertility as follows. It is part of the normal functioning of heterosexual couples to have a capacity to reproduce. A loss of this capacity (a dysfunction), through either physiological or psychological causes, merits medical concern. Lesbian couples, on the other hand, do not as a matter of their normal functioning, have the capacity to reproduce. This inability to reproduce is not, therefore, a deficit in their normal functioning, and does not warrant medical repair. As Robert Jansen, Professor of Clinical Medicine at Sydney University notes:

Providing medically assisted insemination to lesbian couples would be an enhancement of their normal capabilities, and not a legitimate medical intervention to address a loss of normal function. For heterosexual couples on the normal functioning model, having a child through ART would be the fulfilment of their normal possibilities. But for lesbian couples, it would be adding a possibility that enhances their lives.⁽⁴⁶⁾ Parallel points can be made in relation to socially infertile heterosexual women. In the normal run of things it simply turns out that some women will miss out in the social lottery and have no children. This is a situation that is normal to expect (like turning out short), and not a dysfunction.

How well does the normal functioning proposal hold up to closer scrutiny? There does seem to be something in the general idea that medical interventions are primarily reparative, and that 'normality' (or something like it) is important in some way when gauging what counts as a reparation. With this said, however, the devil lies in the detail, and there are some concerns as to whether the normal functioning view tells the entire story about what is medically legitimate. Firstly, it is not clear that it is always medically improper or illegitimate to enhance a person's capacities or well-being beyond what might be normal for them. Secondly, there is question as to what 'normality' is meant to include or exclude, anyway. And thirdly, it is not clear that the normal functioning model actually excludes assisted insemination for fertile lesbians in the way that it suggests.

Medicine and 'normality'

With respect to enhancement, the current medical establishment and the general community recognise a range of medical practices and interventions that are specifically designed to enhance people's well-being and capacities. Cosmetic procedures to improve (acceptable) normal appearance, or lipo-suction for those who are naturally very large, would be examples. No one is suggesting that these are anything more than elective and discretionary procedures, and few would defend them as medically necessary in any sense. However, they arguably still would not count as an abuse of medicine nor would they be otherwise medically improper. Ostensibly, they are medically acceptable procedures.

There is also difficulty in pinning down what counts as normal, for the purposes of determining what is legitimately treatable and what is not. The point was made in the example above about natural shyness that it was normal for many capacities to be distributed unequally in the community, and that many of the incapacities people have, like shortness or shyness, are normal incapacities-incapacities they have as part of their normal condition. These 'natural' incapacities were to be distinguished from incapacities that are a divergence from their normal condition (dysfunctions), brought about by some cause or identifiable factor. However, there seems to be two senses of normality operating here, and they tend to run into each other when distinguishing natural incapacities from dysfunctions.

When looked at from the point of view of how things are normally distributed in the community, it is normal to expect many people to experience dysfunctions. That is, many people will have or experience dysfunctions as part of their normal condition. Are these dysfunctions then natural incapacities? It is not always clear what sort of incapacity something is. Take, for example, the occasional difficulties people have in sleeping. Does this incapacity to sleep, when it happens, count as medically treatable (e.g. with prescribed pills) on the normal function model because it is a divergence from a person's normal pattern of sleep (due to some intervening cause)? Or is it not legitimately treatable because normally everyone finds it hard to sleep occasionally, and consequently it looks more like a natural incapacity? The crucial distinction that the normal functioning model relies on, and in terms of which fertile lesbians are deemed not to be legitimately treatable, does not seem to be clearly sustainable. If the distinction between natural incapacities and dysfunctional incapacities is to be maintained, it will have to be in terms of some factor other than what is 'normal'.⁽⁴⁷⁾

The importance of the impacts of incapacities

Another important observation is worth making. Even if this confusion about normality could be clarified in a way satisfactory to the normal functioning model, that model would still arguably leave out something that seems crucial in deciding whether a medical intervention is legitimate-the impacts or consequences of the incapacity. Even when some incapacity is not a dysfunctional one, like natural shyness, if its constant presence leads to ongoing emotional discomfort and stress, or acts to cut off valuable opportunities that a person would otherwise be able to take advantage of, then arguably that incapacity legitimately warrants medical attention. In ignoring the impacts of incapacities, the normal functioning view fails to adequately reflect what seems to be a crucial consideration in deciding whether an intervention is medically legitimate. This oversight is of particular significance in the context of social infertility, where the inability to conceive is often a source of considerable stress and deep dissatisfaction.

Repairing the social 'incapacity' to conceive: A lifestyle enhancement or a return to normality?

Finally, there is the question of whether it is true, as suggested earlier, that the 'incapacity' to conceive that socially infertile women experience is something that should be considered normal for their circumstances (a natural incapacity on the normal function model). Despite what was said earlier about the fertility incapacity of lesbian couples, it can nonetheless be argued that having the capacity to reproduce is part of the normal functioning of socially infertile women as women, regardless of their relationship status. Taking that as the norm, it can be argued further, that the relationship status or sexual preferences of these women (whether chosen or not) actually detracts from this normal capacity. To that extent, social infertility could be thought of as an impediment to normal functioning (i.e. a dysfunction), and something that would be entitled to medical repair (through medically assisted insemination). As Professor John Pearn of the Brisbane Royal Children's Hospital observes in a similar connection:

As a characterisation of the legitimate aims of medicine, the normal functioning model underlying Proposition 3 is limited, as is its force in excluding medically assisted insemination for socially infertile women.

Proposition 4: Medical treatments are for medical conditions

The Arguments

Though it has been left till last, this view is probably the one that most readily comes to mind when considering medical interventions for non-medical conditions like social infertility. It is most readily thought of because it seems simple and self-evident, and it readily generates the following argument: Medical techniques involving physiological interventions are designed to overcome physiological impediments to well-being. In characterising this view, de Lacey writes 'medicine is historically grounded in positivism wherein a prescribed intervention follows the identification of a cause for illness, and for which a positive outcome is predicted.'⁽⁴⁹⁾ So, in the case of reproductive techniques, their point will be to overcome physiologically caused obstacles to pregnancy. Medical interventions are illegitimate in cases where there is no medical condition to treat, as in social infertility. Straightforward as this argument is, it turns out to be inaccurate in some key, but instructive, respects.

Medical interventions and infertility: Causes or condition?

Reasonable and acceptable medical interventions are not always confined to conditions with an identifiable physiological basis. In fact, with ART it is not always known what the exact nature of the impediment to pregnancy might be,⁽⁵⁰⁾ and it is generally recognised that some occasions of persistent infertility in heterosexual couples can be due to male psychological factors. But what is more, there are many ostensibly legitimate medical interventions that are not designed to treat, repair or otherwise address 'causes' of illness or poor wellbeing at all, whether physiological or not. For example, pharmacological palliatives for headache, insomnia and asthma; certain surgical procedures to relieve pain; and many other medical and health care interventions, do not treat causes, but address symptoms. They intervene not to remove or change the physiological or psychological causes of conditions, but to block or remove or change the effects of those causes. Indeed, it can be argued that this is exactly what ART does. 'The cause of the infertility is not the issue; like deafness or paraplegia, it is the disability itself that is important.'⁽⁵¹⁾ As was said, in many cases, the causes of the infertility remain unassailable. With donor insemination for physiologically infertile couples in particular, the procedure is performed on the woman, even when the causes of the infertility reside with the male partner. Assisted reproduction assists the completion of the reproductive process, not by removing or repairing the physiological or psychological causes of infertility, but through avoiding or bypassing their impeding action. ART, and assisted insemination, is less a treatment for infertility and more an alternative mode of conception to heterosexual intercourse.⁽⁵²⁾ It does not repair the cause, it addresses the symptom-childlessness-or to be more precise, the distress and felt social stigma that some people acutely experience in being childless.

There is a considerable body of survey and interview-based evidence to consistently indicate that not being able to have a child (when one strongly wants one) can be a distressing and devastating experience. These studies reveal that being childless can be associated with feelings of loss of status and self-esteem, and the questioning of identity, particularly in the case of women for whom the social and gender-based expectation to procreate is strong.⁽⁵³⁾ It has been theorised that the distress that may be associated with childlessness, particularly in women, is social in origin and based in gender-oriented social norms, expectations and constructions surrounding femininity, procreation and motherhood.⁽⁵⁴⁾ As Sheryl de Lacey notes, 'motherhood has historically been constructed as a biologically predetermined, natural and therefore inevitable function of women, through a discourse of 'biological destiny'.⁽⁵⁵⁾ If this thesis is accurate, being childless is a problem for some women because of the social factors they are subject to. It is ironic that even when the condition of childlessness has a physiological cause, what makes that condition distressing, and so makes it something we consider worthy of medical repair-is social in nature. The thing that legitimates medical assistance for infertility, whether social or physiological, has a social basis. So, there is a clear sense in which all (problematic) infertility is a social condition.

There is no obvious reason to think that the distressing condition of childlessness will be any less acutely felt by lesbians and single heterosexual women, than by heterosexual couples who are physiologically infertile. And if, as just suggested, it is this condition that is the real attractor of medical concern, and for which the application of ART procedures is thought appropriate for physiological infertility, then parity of concern would suggest that it is just as legitimate to apply those procedures in the case of lesbians and single women. Perhaps this line of reasoning would have been more obvious from the start if the focus had been more on the consequences rather than the causes of childlessness. These women are better described as 'socially childless' rather than socially infertile.

From medical legitimacy to medical necessity: the issue of public subsidy

The limited nature of our conclusion is reinforced by the fact that we have been discussing medical legitimacy-what it is medically permissible to do, and not medical necessity-what should (or perhaps must) be done to maintain a satisfactory level of health and well-being. The latter is a stronger condition, and will need to take account of a broader range of factors, including ones that we have put aside in this discussion. The question of what medical necessity consists of, and how it differs from mere legitimacy or permissibility, will not be entered into here, either. Nonetheless, the critical clarifications that have been made in the previous discussion do serve to bring to the fore a further important question relating to the public subsidising of ART. This paper will end by noting in a very preliminary way some of the implications the previous discussion has for that question.

Medicare support: Some preliminary reflections

Should the provision of ART for socially infertile women be subsidised through Medicare? Two factors are relevant in answering this: (i) medical necessity; and (ii) parity. Consider the first factor. Arguably only procedures that are necessary for medical treatment are eligible for rebate according to the Health Insurance Act 1973. The likes of cosmetic surgery to improve already acceptable appearance would not be covered, for instance. Clearly, an extended discussion would be required to decide once and for all whether, and when, assisted insemination is medically necessary for socially infertile women. Nonetheless, it is still possible to advance some respectable preliminary considerations based on 'harm-minimisation' to the effect that sometimes it could well be medically necessary.

It was noted earlier that self-insemination by lesbians and single women is thought to be common. A recent study has indicated that of the lesbian women surveyed who intend to become mothers within the next five years, 70 per cent of them intend to conceive through self-insemination⁽⁵⁶⁾ by arranging an informal sperm donor. If there is no medical screening of this at-risk sperm for genetic defects and transmissible diseases, there will be a substantial proportion of socially infertile women who intend to engage in a potentially harmful practice, harmful to themselves and to the wider community. Given this, it is certainly not improbable that there will be cases where the health-related risks and harms are potentially serious enough to require the provision of appropriate medical interventions to avoid them. In other words, to make necessary the provision of medically assisted insemination, where the sperm used is routinely screened and tested.

It may be replied here that socially infertile women seeking to self-inseminate face no greater risk than most other women seeking to become pregnant with their male partner. However, there are relevant differences in the nature of the risks experienced. Women in an ongoing heterosexual partnership are likely to have a greater knowledge of the male's family history and personal background, and consequently the possible risks that are involved in conception with the male's sperm. Socially infertile women will not necessarily have the same level of information to assess risks. Similarly, in heterosexual partnerships, whatever risks there are, they will be shared risks for both partners (assuming the child will be raised by both, and that a harm to one will be a cost to the other). With socially infertile women seeking insemination, however, the risks are not necessarily shared by the inseminator. And importantly, if there is the suspicion of a problem before conception, heterosexual couples still have the option of testing, which socially infertile women do not.

Even if the chances of acquiring a sexually transmitted condition from unscreened sperm were relatively low, a cost-benefit analysis would still argue strongly in favour of subsidised screening. Such a subsidy would be low in cost to the Commonwealth compared to the very substantial amounts involved in treating, say, a HIV/AIDS infection. So, there are at least the beginnings of an argument on grounds of harm-minimisation, that there may be cases and circumstances where it is medically necessary to provide assisted insemination to socially infertile women.⁽⁵⁷⁾

The practice and circumstances of self insemination also involves a range of other risks and harms, although not of a strictly medical nature. For instance, lack of clear legal regulation or protection concerning obligations of paternity, or level of paternal involvement. A further, and significant harm which is often overlooked is that of the criminalisation in some Australian states of inevitable behaviour, where socially infertile women see themselves as having no viable option but to engage in an illegal activity.

Turning now to the question of parity. One of the more important points to emerge from the previous discussion is that it is not the cause, but the condition of childlessness that seems most important in determining whether a medical intervention is justifiably warranted. According to current medical practice that condition has a sufficiently serious symptomatology to warrant medical repair on many occasions. But then, if it is this condition that is the medical warrant for providing physiologically infertile couples with ART-a provision which is currently publicly subsidised-then it ought on grounds of parity be similarly the case for all women who experience the same distressing symptoms of that condition, including socially infertile women.

It should be noted with this argument that there is a question about the appropriateness of subsidising ART for physiologically infertile couples, particularly when there are limited medical resources and urgent medical needs that are sometimes hard to meet. Notwithstanding this, physiologically infertile couples are currently subsidised, and without a great degree of controversy. In this existing policy situation, the onus is to show why social childlessness should not be treated like-wise. This onus becomes even more pressing when the extra public costs of subsidising assisted insemination for the socially infertile would be low compared to the existing level of Commonwealth expenditure on assisted reproduction.⁽⁵⁸⁾

These brief remarks by no means constitute anything like a full defence of medicare support. But they are arguably sufficient to shift the burden of proof onto those who want to maintain that ART should never be subsidised for socially infertile lesbians and single women.

Conclusion

It was contended at the beginning that all of the credible arguments that could be brought against the idea that it is medically illegitimate to provide ART to socially infertile women are variants of the four propositions just discussed. That discussion has shown these propositions not to hold up under scrutiny. We can therefore, and with some degree of confidence, conclude that these leading reasons for denying socially infertile women access to ART on medical grounds have not sufficiently made their case.

It was also pointed out at the start that this conclusion will be a limited one. The fact that there are no sound medical reasons for excluding these women, does not mean there are no sound reasons of some other sort for excluding them (either on some occasions or in every case). So it has not been shown that socially infertile women ought to simply be granted access to ART. Considerations to do with, for example, women's reproductive rights, equal access and non-discrimination, parenting, and the interests of the child, would all need to be taken into account, and these issues have been expressly set aside in the discussion in this paper. But even if the conclusion here is limited in this way, it is nonetheless, a quite important one. If it did turn out that ART was medically improper in the case of medically fertile women, these other factors and considerations would be purely academic. The virtue of the previous discussion is that it has illuminated a central and necessary question concerning ART for socially infertile women. The leading views that it is medically illegitimate to provide it can be successfully challenged.

Appendix 1: The distinction between medically relevant factors and other relevant factors

It was said that even when a procedure is justified on medical grounds, there might be other reasons or factors that justifiably prevail against the procedure being applied in a particular set of circumstances (or perhaps even in general). Purely medical reasons do not exhaust all of the considerations that are relevant to deciding whether an intervention can acceptably go ahead in the end. Personal factors will play a role (the patient's consent, and the doctor's own willingness, for instance), as will broader social factors such as the availability of resources, the priorities that are right to assign in the light of those resources, the interests of immediately affected parties, and other emergent social impacts.⁽⁵⁹⁾

If all these various factors are alike in playing a potential justificatory role in whether an intervention should go ahead, then how are the factors that are specifically medical among them to be identified and distinguished from other factors that are relevant but non-medical? Why can it not be argued, for instance, that the broader social impact of a treatment or procedure is a medical factor or consideration, and that the notion of medical legitimacy should be taken to incorporate such factors? And in the particular case of social infertility, why should not the potential impacts on family structures, or the impacts on the future child of not having a male parent, be considered medical reasons for providing or not providing assisted insemination? How is the boundary between purely medical criteria and other relevant factors to be drawn in a way that does not beg the question, especially when the notion of what is properly medical (i.e. what is medically legitimate) is subject to argument? The value-laden nature of medical criteria seems to throw the whole question of what is legitimate on medical grounds, or what is in accordance with medical purposes, open to any set of values or ethical interpretations.

The fact that medical criteria are value-based, however, does not mean that simply anything goes, or that any normative view will do when it comes to presenting a plausible definition or characterisation of medical factors. Just how plausible some definition of medically relevant criteria is will depend partly on how well it reflects the core and fundamental way the concept 'medical' is generally applied and understood in our community. Part of the point of a definition or analysis of a concept is to reveal and clarify the underlying properties and suppositions in our existing understanding of it. Even in the case of ethically contested concepts like 'being medically legitimate', or 'medical purposes', an argued defence of a set of defining criteria still purports at some level to be a description of a concept we use and are familiar with. To be plausible, the description needs to be recognisable to us, and therefore anchored at some point in our existing understanding (though not necessarily entirely, if it is a critical revision or an argued analysis of the concept).

Returning to our original question, when judged against the background of our current understanding, it can be argued that it is less rather than more plausible to count the broader personal and social impacts of an intervention as being medical factors. The argument for this is as follows.

We normally understand medical interventions to be physiological or psychological interventions. But not all physiological or psychological interventions would count as proper or legitimately medical ones (stabbing for instance, or water-torture). Medical interventions are distinguished by the nature of their goals and, as was said, these are subject to argument. Medical goals provide the medically relevant reasons for an intervention, without which reasons there would be no point at all in intervening.

It is a fundamental and widely accepted feature of our existing understanding that the goals of a medical intervention will primarily be to benefit the well-being of the individual who undergoes the intervention. At this intuitive level, if there is no reason to benefit the well-being of the individual concerned, there would seem little reason to intervene. In this sense, the well-being of the individual seems to have a privileged place as a goal of medical intervention in our underlying understanding of medicine.

These few reflection provide two key questions for judging whether some factor or consideration would plausibly count as a specifically medical one: (i) to what extent does it intuitively provide a reason for a physiological/psychological intervention (i.e. does it behave like a goal or purpose)?; and (ii) to what extent is it compatible with benefiting individual well-being? These questions can be asked of the sorts of factors mentioned a few paragraphs ago, and also of those considerations relevant to assisted insemination and social infertility which have been set aside in this paper. Groups of such factors will be addressed in turn.

1. Personal factors such as the consent of the patient, and the willingness of the doctor: As noted, these are certainly relevant to whether a medical intervention should go ahead or not. But they do not of themselves provide reasons for intervening (either necessary or sufficient). An independent reason(s) for intervention needs to exist in the first place, for them to even qualify as relevant considerations.
2. The availability of resources and the allocative priorities that might fairly apply: Very relevant as these factors are, they do not logically behave like (medical) goals. That there are enough resources for an intervention, or that the intervention falls within certain allocative priorities, do not themselves provide reasons for intervening. Again, independent reasons for intervention are presupposed.

Rather than being medical goals or purposes, these factors behave more like qualifications or side-constraints on the pursuit of medical goals. That is, even if there is independent medical reason for intervening (i.e. an intervention would achieve certain medical goals), those goals should not be pursued (and the intervention should not go ahead) if pursuing them would be against the consent of the patient, or the will of the doctor, or appropriate allocative priorities, or would be an unfair use of resources, etc.

Similar things can be said of factors such as:

3. The potential negative impacts of an intervention on the interests of third-party individuals
4. Potentially disruptive broader social impacts of an intervention on, say, family structure or social expectations.

With respect to social infertility and assisted insemination, the salient third party will be the future child. It is often argued that if the psycho-social or psycho-sexual development of a child is sufficiently damaged by being born into a family with no male parent, this is ground to withhold assisted insemination from single women and lesbian couples. These potential adverse effects are best thought of as constraints against undertaking the intervention (and thus against pursuing whatever medical goals apply). It is not plausible to see the avoidance of these effects as a goal in itself, or a purpose for providing assisted insemination.

Similarly with adverse emergent social impacts of interventions; it is sometimes argued that assisted insemination for socially infertile women will unacceptably disrupt normal patterns of family formation, or social expectations about the family unit. If there are such impacts, and if they can be argued to be socially detrimental, then it would be less than plausible to regard their avoidance as a purpose for providing medical insemination. Any such social detriment is better understood as a consideration that might block the intervention (and the achievement of its particular goals).

Impacts on third-party individuals and on society can, of course, be thought of in positive terms, as things to deliberately and directly seek. Thought of in this way, such positive impacts can have the character of goals. Positively promoting the interests of third-party individuals is already recognised by current medical practice as a medical goal in some cases. Organ donation and transplants are instances where the express medical goal is for a physiological intervention on one person to benefit another person. Perhaps also at the social level, certain physiological interventions on individuals (e.g. immunisations) can be provided with the goal of producing social states such as a healthy or disease free society.

As was said, it is possible to mount arguments that the attainment or maintenance of certain social states or third-party interests should be considered appropriate goals for medical intervention. The question then becomes what these social states might be, and how the physiological/psychological interventions that seek them are to be reconciled with benefiting individuals' well-being-something that our current understanding of medicine takes to be a primary goal of medical intervention.

In its discussion of the medical legitimacy of assisted insemination for socially infertile women, the current paper takes the medically relevant factors to be those associated with the goal of benefiting the individual's well-being. The paper in turn understands the other relevant factors (such as the interests of the future child, the quality of same-sex and single-woman parenting, impacts on social expectations and family structures, resource priorities) to be non-medical side-constraints on the medical goal of benefiting well-being. Because the concern of the paper is whether a particular sort of intervention is illegitimate on medical grounds, it does not address these other factors.

It should be said that in no sense does characterising these other factors as side-constraints diminish their force or relevance. Nor does the focus on benefiting individuals' well-being disallow argument and normative debate about medical purposes. There are many questions about the sense of well-being that is most important from the point of view of medicine, the role that the origins or consequences of an individual's state of well-being have in legitimising an intervention, and questions about what form of benefit to well-being is appropriate. In critically examining the four key propositions, the paper takes up these questions in so far as they relate to providing assisted insemination to socially infertile women.

Endnotes

1. For example, 'Limit IVF to stable heterosexual relationships: a child has the right to expect the love of a mum and a dad', Greg Sheridan, The Australian, Friday August 4, 2000; 'Father of all debates is about child's right', Bettina Arndt, The Sydney Morning Herald, 5 August 2000.
2. McBain v State of Victoria 2000, FCA 1009.
3. For some of the background and possible legal implications of the challenge to the Victorian Infertility Treatment Act 1995, see Katrine Del Villar, 'McBain v State of Victoria: Access to IVF for all Women', Research Note No. 3, 2000-01, Department of the Parliamentary Library; and 'McBain v State of Victoria: Implications Beyond IVF', Research Note no. 4, 2000-01, Department of the Parliamentary Library.
4. Medical infertility is usually understood in the medical profession to mean 'the inability of a couple to attain or retain a pregnancy following 12 months of regular sexual intercourse without contraception' (Access to Reproductive Technology: Final Report to the Australian Health Ministers' Conference, National Bioethics Consultative Committee, 1991). However, it should be noted that there is still question as to how medical infertility should be characterised.
5. 'Exploding the myth of the nuclear family', Australian Medicine, 4 September 2000, p. 14.
6. Which is approximately 1 per cent of the annual Australian IVF treatments. David Molloy (IVF Directors Group of Australia) and Benjamin Haslem, 'IVF Battle over just 150 Women', The Weekend Australian, 5 August 2000. Confirmed also in personal communication.
7. Ian Johnston, Chairman, Reproductive Technology Accreditation Committee, Fertility Society of Australia. Correspondence, 13 October 2000.
8. Survey conducted by Significant Others Market Research for Lesbians on the Loose magazine, with a national readership of 45 000. Cited in Canberra Times, 3 August 2000.
9. Tom McNamee, 'Lesbian, gay parents increasing, but hard to count', Chicago Sun-Times, 28 November 1994.
10. Office of Technology Assessment, US Congress, Artificial Insemination: Practice in the US: Summary of a 1987 Survey, US Government Printing Office, Washington DC, 1988. (Cited in Judith Lasker). There is generally a lesser degree of legislative regulation of ART in the US than in Australia, and access tends to be less restrictive. In some cases there is even mail-order type access. See http://206.117.149.143/index2.cfm.
11. Helen Szoke, 'Regulation of assisted reproductive technology: the state of play in Australia', in Ian Freckleton and Kerry Petersen, eds, Controversies in Health Law, Federation Press, 1999. The Northern Territory is guided by the South Australian Reproductive Technology Act 1988.
12. Ethical Guidelines on Assisted Reproductive Technology, National Health and Medical Research Council, 1986. A previous version of those guidelines, however, did advise that donor insemination should only be provided to those in 'accepted family relationships'. However there was apparently disagreement and unclarity as to what this included and excluded, and the provision was deleted in favour of an emphasis on the rights and welfare of the child as paramount. Despite seeming less restrictive than the former guidelines, the current ones nonetheless advise that where state statutes or providers' codes of practice may be in breach of the federal Sex Discrimination Act 1994 (in excluding unmarried women), exemptions from that act could be sought. Jenni Millbank, 'Every sperm is sacred?', Alternative Law Journal, vol. 22, no. 3, 1997.
13. European women who are denied access to ART in their own country (a significant proportion of whom are older and post-menopausal) often arrange privately paid treatment in Italy, for example, where there are fewer restrictions on access to ART. D. Evans and M. Evans, 'Fertility, infertility and the human embryo: ethics, law and practice of human artificial procreation', Human Reproduction Update, vol. 2, no. 3, 1996, pp. 208-224.
14. To the extent that opinion surveys can be taken as an indication, one recent poll indicates that 47 per cent of Australians support single women, and 44 per cent support lesbians having access to IVF (Herald-AC Nielsen, reported in the Sydney Morning Herald, 15 August 2000). Another poll (The Herald Sun Voteline, 3 August 2000) indicates a much smaller minority (10 per cent) who agree that these women have a right to access IVF, and a larger majority (90 per cent) who disagree.
15. Small scale surveys on this issue have produced conflicting results, even within the one region. For example, David Molloy (a Brisbane IVF specialist) reports that 80 per cent of sperm donors in one Brisbane clinic say they do not want to donate for single women, and 90 per cent do not for lesbians. Judith Whelan, Sydney Morning Herald, 19 August 2000. Another survey of men who donated sperm in Brisbane indicated that 64 per cent had no objection to donating to lesbians, and 79 per cent no objection to donating for single women. Heather Pollock, Queensland Fertility Group, Fertility Society of Australia Conference, 1997, http://www.nor.com.au/community/aisg/article01.htm.
16. Bernard Lane, 'Bishop's IVF action', The Australian, 27 October 2000, quoting Archbishop Francis Carroll.
17. These objections are predominantly religiously based. Some Roman Catholics will argue that assisted reproduction is an unacceptable way of becoming a parent because it separates the 'unitive' and the 'procreative' aspects of reproduction. See The Congregation for the Doctrine of the Faith, Instructions on Respect for Human Life in its Origin and on the Dignity of Procreation, Vatican City, Vatican Polyglot Press, 1987.
18. For an indication of this, in the 1983 Australian Values Study, Roy Morgan Research Centre, 77 per cent of the 1228 Australians surveyed, approved of IVF as a technique for helping infertile married couples; and in a 1997 Western Australian study, 89 per cent of respondents approved of married couples using reproductive technologies, and 74 per cent supported its use by de facto heterosexual couples. 'Attitudes toward access to reproductive technology', unpublished paper, Dr. Pia Broderick, Murdoch University Department of Psychology.
19. See, for example, Norman Daniels and James E. Sabin, 'Determining "medical necessity" in mental health practice', Hastings Center Report, November-December 1994, pp. 5-13. Also, Norman Ford, 'Access to Infertility Clinics for Single Women and Lesbians?', Chisholm Health Ethics Bulletin, vol. 6, no. 1, Spring 2000.
20. See, for example, Robert P. S. Jansen, 'Reproductive medicine and the social state of childlessness', in The Medical Journal of Australia, vol. 167, September 1997, pp. 321-23. Also E. J. Cassel, The Nature of Suffering and the Goals of Medicine, Oxford University Press, New York, 1991 (cited in Jansen op. cit.).
21. In the case of artificial insemination, though, the resource-cost arguments can be overstated, since the procedures are fairly straight-forward, and the costs relatively low. Currently, donor insemination procedures roughly cost around $600 per attempt at pregnancy.
22. It can also impact on particular social groups. Much of the feminist critique of reproductive technology, for instance, focuses on its perceived role in further entrenching oppressive gender relations and its perceived male-dominated contribution to the social construction of gender. See, for example, R. Klein, The Exploitation of a Desire: Women's Experience with IVF, Deakin University, Geelong, 1988. See also, D. Steinberg, 'The depersonalisation of women through the administration of In Vitro Fertilisation', in M. McNeil, I. Varcoe and S. Yearley, eds, The New Reproductive Technologies, Macmillan, UK, 1990.
23. See Simone Bateman Novaes, 'The medical management of donor insemination', in K. Daniels and E. Haimes, eds, Donor Insemination: International Social Science Perspectives, CUP, 1998. Lynn Payer also documents some significant cultural variations between the practices of physicians in the USA, England, West Germany and France, with respect to procedures such as hysterectomies and caesarean sections. See Medicine and Culture, Penguin Books, New York, 1988.
24. See, for example, R. M. Veatch, The Patient-Physician Relation, Indiana University Press, 1991, for an account of the ways in which medical decision-making and practice are value-laden.
25. Bateman Novaes, op. cit., p. 117.
26. The medical profession has traditionally claimed a privileged and sometimes exclusive role in deciding what is medically appropriate. See, for example, J. D. Keeping, 'Should lesbians receive donor sperm?: Clinical, not personal, guidelines', Australian Medicine, 5 May 1997, pp. 8-9. Also, the definitions of 'clinically relevant' and 'inappropriate medical practice' in section 3 and section 82 of the Health Insurance Act 1973, rely heavily on peer-based judgements in the medical profession. Alternatively, cultural theorists such as Sheryl de Lacey ('Assisted reproduction: who qualifies', Collegian, vol. 5, no. 4, 1998, pp. 28-36) argue that the medical profession has been invested with too much control over what does and does not count as medically appropriate, and consequently has had too much power to determine what social groups do and do not have access to a particular social resource (ART).
27. George A. Bray, 'Drug therapy of obesity', Medline UpToDate, October 2000 http://www.medscape.com/28038.rhtml?scrmp=ms-100600.
28. Many clinics will only provide artificial insemination to people who have attempted unsuccessfully for twelve months to achieve pregnancy through intercourse.
29. Tanya Canny, 'Same sex couple adoption: the situation in Canada and Australia', Research Note No. 29, Department of the Parliamentary Library, April 2000; also J. Lasker, 'The users of donor insemination', in K. Daniels and E. Haimes, eds, Donor Insemination: International Social Science Perspectives, CUP, 1998.
30. Infertility Treatment Act 1995 (Vic), s. 7. And despite this prohibition, it is thought that self-insemination is widely practised and hard to detect. See Gabrielle Wolf, 'Frustrating sperm: regulation of AID in Victoria under the Infertility Treatment Act 1995 (Vic)', Australian Family Law Journal, vol. 10, no. 2, 1996. Self-insemination is also legislatively prohibited in Western Australia and South Australia.
31. Why unacceptable in this way? Because if a single heterosexual woman already had the opportunity to conceive with a man with whom she considered this appropriate and acceptable, she would be unlikely to be seeking assisted insemination in the first place. She would not, in other words, be in the group of single socially infertile women that the current paper is concerned with.
32. See Y. Englert, 'Artificial insemination of single women and lesbian women with donor semen', Human Reproduction, vol. 9, no. 1, 1994, pp. 969-971; and R. S. Leiblum, M. G. Palmer, and I. P. Spector, 'Non-traditional mothers: single heterosexual/lesbian women and lesbian couples electing motherhood via donor insemination', Journal of Psychosomatic Obstetrics and Gynaecology, vol. 16, 1995, pp. 11-20.
33. Australian Story, ABC Television, 12 October 2000.
34. Surveys indicate that some lesbians do have intercourse with men on occasions, most often gay or bisexual men. For instance Sydney Women and Sexual Health Survey 1996, AIDS Council of NSW; L. Remez, 'As many lesbians have had sex with men', Family Planning Perspectives, vol. 32, no. 2, p. 97. With respect to the mode of conception (in the current situation where medically assisted insemination is rarely available for lesbians), one recent survey indicates that the overwhelming preference among lesbians is for self-insemination. See Significant Others Market Research, Parenthood Intentions of Lesbian Women, 2000.
35. Englert, op. cit.
36. This general form of argument based on personal responsibility and voluntary risk-taking is outlined and explored in Chapter 6 of Tom L. Beauchamp and James F. Childress, Principles of Biomedical Ethics, OUP, 1989.
37. See, for example, Arnold S. Kaufman, 'Responsibility, moral and legal', in Paul Edwards, ed., The Encyclopedia of Philosophy, vol. 7, Collier-Macmillan, 1970, p. 183.
38. Kaufman, op. cit.
39. Beauchamp and Childress, op. cit., p. 282.
40. de Lacey, op. cit., p. 32.
41. Although there are limited occasions where this observation will be made in the context of health-care, it is often appealed to in relation to broader social and political issues-in the context of justifications for civil disobedience and conscientious objection, for instance. See John Rawls, 'The justification of civil disobedience', in Hugo A. Bedau, ed., Civil Disobedience: Theory and Practice, Pegasus Books, New York, 1969, pp. 240-255. Also in the question of what level of assimilation is appropriate for ethnic minorities voluntarily immigrating to another country, see M. Rickar