1Vakgroep Filosofie, Pleinlaan 2, Lok. 5 C 442, Vrije Universiteit Brussel, B-1050 Brussel, Belgium
Everybody agrees on the fundamental importance of the welfare of the child when judging the applicability of new reproductive technologies. However, this is as far as the consensus goes. The moral problem can be presented as the opposition between the right of the would-be parents to procreate and the well-being of the child. The positions on this issue are largely determined by one's view on parental responsibility (Pennings, 1995b). The most frequently mentioned difficulty is that there are no reliable predictive criteria for inadequate parenting and, thus, no criteria which can be used to guarantee the best interests of the child (Harris, 1990
). If we cannot reach a consensus on minimum parental capabilities, how can we introduce a licence or a test to exclude incompetent candidates for parenthood? (Lafollette, 1980
). The second, and less well-known, problem concerns the choice of the principle to interpret the level or measure of welfare. Even if we were able to calculate the `global welfare' of the child, we would still have to evaluate the result. The principles outlined in this article can be seen as positions on a continuum between the absolute autonomy of the parents on the one hand and exclusive concern for the welfare of the offspring on the other.
This contribution to the debate on the welfare of the child aims to highlight two points: firstly, that the standard for evaluating the welfare of the child differs according to opponents and proponents of certain applications and, secondly, that fixing on such a standard may reinforce the false idea that all moral positions are based on, and decided by, the outcome of that issue. Neglecting these distinctions seriously hampers the discussion on the acceptability of new reproductive technologies.
The measure of welfare does not allow us to reach a conclusion, unless we have a procedure to judge whether the amount is acceptable or not. Two different evaluation rules are used: (i) the maximum welfare principle; and (ii) the minimum threshold principle. Both rules help to determine whether the child that is born as a result of the infertility treatment is harmed by being brought into existence.
The maximum welfare principle
The maximum welfare principle implies that one should not knowingly and intentionally bring a child into the world in less than ideal circumstances. The following quotation from the editors of the Lancet illustrates this reasoning for post-menopausal pregnancies: `The long term well-being of the child should be of overriding importance. Research shows that children need a stable home with mature caring adults who themselves have a sound relationship. This need extends into the late teens, and even people in their 20s benefit from the love and support of their parents. Of course, many fertile couples have unplanned conceptions and some of their babies are born into circumstances that are far from ideal; we have little control over such `natural' events. However, ethical considerations inevitably enter into the decision to use high technology to give a woman a pregnancy. Thus the Human Fertilisation and Embryology Act (1990) insists that those providing in-vitro fertilization must take into account the welfare of the child, including the need of that child for a father' (Lancet, 1993). Since we can control (at least to a certain extent) the circumstances in which a child is made when the candidates are infertile, we ought to restrict our co-operation to those cases which maximize the welfare of the child. At the same time this fact explains why the standard for medically-assisted procreation must and can be higher than for natural reproduction.
The maximum welfare principle is used whenever the application of the new reproductive technologies is rejected because of some specific condition which deviates from the ideal of the family in Western societies (Acock and Demo, 1994). Every characteristic of those who request medical assistance that does not conform to the `heterosexual married parents with their genetically related children' pattern is assumed to result in negative consequences for the child (Golombok, 1998
). The comparison which underlies the moral assessment of the welfare is that the expected happiness of the child would have been greater if it had been born in ideal circumstances. Strictly speaking, this comparison is impossible since the same child could not have been born in other settings, to other parents and at a different time. This `non-identity' problem is based on the basic biological fact that every person could only have originated from the gametes from which he or she originated (Parfit, 1984
). This problem does not prevent us from comparing the quality of life of different children born in different settings. We can reject procreation in alternative settings because the welfare level of the existing child is lower than the expected welfare level of another possible child that would have been born in (what we consider as) ideal circumstances. The existing child is harmed to the extent that it could have had a better life. This is an extremely strong standard. When we take the time to scrutinize the consistent application of this rule, we will soon find out that this standard would exclude the overwhelming majority of the population from procreation. People who are poor, unemployed, handicapped, obese, workaholics and/or old should all be rejected as potential parents since the child they will have would have had a better life had it been born to other parents. The selective appeal on this rule shows the bias against some groups.
The maximum welfare principle can also be formulated in the following way: `if we can choose between having a happy child in circumstances x and having a less happy child in circumstances y, we must choose to create the child that will be the happiest'. This principle seems self-evident but it is not. It applies at the most in those situations where all other conditions besides the welfare are equal and in real life they rarely are. Suppose a 30 year old woman has to decide whether to postpone her pregnancy. She knows (in relative terms of certainty) that if she has a child now it will be healthy, while if she waits 10 years the child may be born with a congenital anomaly. In this situation, the principle prescribes that she has the child now. However, the circumstances are not the same. Almost all aspects of the life of a woman (partner relation, career, financial security...) may have altered in 10 years. Consequently, having a child 10 years earlier or later in your life may be incomparable. Moreover, the application of the rule entails that prospective parents have to consider all possible circumstances which might contribute to the welfare of their child during the whole period of their reproductive life span (and even beyond). They should try to figure out at what time and in which circumstances procreation would result in the child with the highest possible welfare. The factors include timing, reproductive partner (another partner might lead to happier children), economic situation, etc. In fact, in evaluating the decision of a person to procreate, we should make a comparison with all the possible children that the person could have had in the past and all the possible children he or she can have in the future. This is a very unrealistic and even irrelevant undertaking but due to the isolation of, and the specificity attributed to, the decision to conceive in the context of medically-assisted procreation, this is the kind of reasoning which is actually proposed by those who would like to restrict the applications in less than ideal circumstances.
The second point is how we fill in the `choice' condition. The application of the rule implies that the prospective parents have a choice between different possible children. In the debate on post-menopausal pregnancies one could hear the argument that `they should not have waited so long' and that `they had 30 years to have a child' (Pennings, 1995b). One repudiates treatment of this group because of the comparison between the child the parents could have had 20 years ago and the child they will have now. The same reasoning is applied to the other categories. Single women who want donor insemination should look for a man. Lesbians should become heterosexual. People who are poor should wait until they are financially secure, those who are living together should defer procreation until they are married and all those whose conditions are invariable or untreatable should forsake procreation. This kind of reasoning does not consider the decision to procreate as part of the life of the people who make it. It urges them to be different people than they are. Moreover, it shunts the difficulties by transferring the choice to the characteristics. A woman may not have a choice between two children but she has a choice on whether to remain single. This is a very simplistic approach and every physician and ethicist who has been confronted with real life stories knows that such a view does not do justice to the person's problem. We should judge the reproductive decisions of people as these decisions present themselves to the prospective parents as they are.
Finally, the rule only applies in exclusionary cases (when the creation of one child prevents the creation of another) and not when we are considering additional children. If we have to choose between creating a child in a normal family and creating a child in an alternative setting, we have to prefer the first option (always presupposing that the former in general leads to a higher welfare for the child). In conditions of scarcity (of gametes, physicians, infrastructure...), the maximum welfare rule can justify that priority is given to conventional families (as was done in Denmark). One could also argue that the available donor oocytes should be given to younger women, and this not only for reasons of efficiency and success rates. In practice we are only confronted with pure exclusionary situations when donor oocytes are needed. And even then, other solutions can change the circumstances in such a way that the exclusion is avoided. The post-menopausal women can be asked to bring their own donor to ensure that the supply of donor eggs is not diminished (National Advisory Board on Ethics in Reproduction, 1996). I want to emphasize again that before we apply the rule we need evidence to support the differential attribution of welfare to the different groups.
The minimum threshold principle
The concrete circumscription of all dimensions of the welfare of the child proves almost impossible. This problem is closely linked to the fact that we have not yet been able to discover adequate criteria of a good parent. However, the experience in law taught us that a reasonable consensus can be reached on which conditions or characteristics of the parents are unacceptable. Previous criminal convictions for child abuse and serious mental illness of the parents are cited as examples of such absolute contra-indications (Mumford et al., 1998). A survey of fertility centres in the US revealed that four criteria were used for treatment rejection: substance abuse, physical abuse, severe marital strife, and coercion of one spouse by another (Leiblum and Williams, 1993
). Still, even this minimal threshold might be the subject of disagreement. The field of medical genetics presents several examples of fundamental discord on the right to procreate of parents who have an increased risk to have a child with serious malformations. A similar debate is currently going on concerning the right of parents infected with human immunodeficiency virus (HIV) to procreate (Arras, 1990
; Nolan, 1990
). The threshold principle is connected to a weak interpretation of the right to procreate: `a person has a right to rear children if he meets certain minimal standards of child rearing' (Lafollette, 1980
). This standard does not compare the welfare of the child-to-be-created with other possible children but it only verifies whether the quality of life of the future child is above the minimum threshold. This system is much more flexible since it does not require us to work with possible persons and does not demand that fine distinctions be made between different levels of welfare. There will, of course, be the unavoidable borderline cases but the general idea is that we should only condemn procreation in those circumstances where the expected welfare level of the child would be very low.
One of the most frequently used minimum thresholds can be called the `wrongful life' or the `worse than death' standard: `A child should not be brought into the world if and only if it would have been better never to have been born at all.' (Robertson, 1994; Strong and Schinfeld, 1984
). This is an extremely low threshold. If every act of procreation that fulfills this standard is considered acceptable, we have to accept some very counter-intuitive decisions. In fact, on this account we cannot even say that it is wrong to bring a child into existence whose prospects and opportunities are awful (Steinbock and McClamrock, 1994
). No one can seriously maintain that being born with a lesbian or single mother or being born to a post-menopausal mother is worse than not existing. However great the difficulties they will have to face, those children will not say that they would have preferred not to have been born at all. Nevertheless, this in itself is not sufficient to conclude that it is alright for these groups to have children. I think more can and should be expected from parents. The concept of parental responsibility would be a hollow notion if bringing, knowingly and willingly, children into existence who suffer devastating illnesses cannot be denounced.
The minimal threshold standard attributes a disproportionate importance to the autonomy principle as expressed in the right to procreate of the parents at the expense of the welfare of the child. According to this principle, the child is only harmed if it is brought into existence with a life not worth living. While this standard should be rejected for the moral evaluation of procreation, it might be an acceptable reference point for legal rules. Only those acts should be legally forbidden which lead to the creation of a child who would have been better off if it had never been born.
The reasonable welfare principle
We propose an intermediate principle that avoids the counterintuitive judgements of the two other principles and simultaneously conforms more closely with the way we look at procreation and parental responsibility in ordinary life. On the one hand, we do not have to reject or criticize people for bringing a normal child into the world because they could have had a happier one. On the other hand, we are not forced to accept decisions which result in the birth of seriously handicapped children because the net result is a life still worth living. Our standard is not the perfectly happy child but the reasonably happy child. Nobody's parents are perfect and no one is completely happy. We do not criticize parents for taking decisions which might negatively influence the well-being of their children (like moving, changing schools, working outside the home etc), if they have good reasons for doing so. We do not expect parents to live their lives with the sole purpose of maximizing the welfare of their children, even if this means that their own life plans are thwarted. In reality we only criticize parents when they pursue their own goals although this has disastrous consequences for their children. For the moral evaluation we use a standard of what is `acceptable' or `reasonable'. The rule to evaluate the applications of the new reproductive technologies could be stated as follows: `The provision of medical assistance in procreation is acceptable when the child born as a result of the treatment will have a reasonably happy life'.
It is rather difficult to give an elaborate description of which construction constitutes the normal state of happiness or welfare. We have to rely to a considerable extent on our common sense. Several authors have tried to define the concept of a decent welfare level. Daniels refers to the abilities that are required for an individual to enjoy a normal range of opportunity in his society (Daniels, 1985). Kavka uses the notion of a `restricted' life, that is a life `significantly deficient in one or more major respects that generally make human lives valuable and worth living' (Kavka, 1982
). Kleinig defines the welfare of a person by `the absence of defects and irregularities with respect to some conception of its normal functioning' (Kleinig, 1978
). Although these constructions all remain fairly vague, they contain a common core which can be used to demarcate a reasonably happy life. I would suggest that an individual has a decent welfare level when he has the abilities and opportunities to realize those dimensions and goals that in general make human lives valuable. All those conditions and defects which obstruct the pursuit of the normal human interests should be considered as harm to the person (Morreim, 1988
). The deviations which cause the welfare level of the child to fall significantly below this baseline comparison state are sufficient reason to refuse treatment. If, for instance, psychologists found that children conceived in lesbian households have serious problems with their gender identity, this constitutes a reason to deny access to treatment for this group (Golombok and Fivush, 1994
). Having trouble identifying as male or female will very probably gravely interfere with their ability to form lasting intimate relationships in adult life and the formation of stable and non-superficial relationships is one of the basic conditions of a happy life (Kleinig, 1978
). Another possible example of an infringement of the right to a reasonably happy life could be the demonstration that donor offspring who know about the method of conception have serious identity problems if they are unable to acquire the name of their genetic father. In this case, the main point is not a total prohibition of all treatments with donor material but a complete stop on all anonymous donations. The standard may not only specify the criteria for access to treatment but may also require amending the existing procedures. However, this argument simplifies things and needs specification. It is, for instance, very unlikely that all children raised in some type of household or all children who do not have the name of their donor will suffer to the same extent. The percentage of children with the disadvantage should be combined with the seriousness of the disadvantage to decide what should be done. Moreover, social circumstances should be taken into consideration. To take the hypothetical example of the children of lesbians again, the gender identity problems may be reduced by introducing a male parental model into the family. Or, the self-concept problems of the donor offspring caused by the lack of information on the donor may disappear or largely diminish if non-identifying information is provided. The context in which the child is conceived as well as the environment in which it is raised may provide several elements to prevent the breach of the reasonable welfare standard. These elements may be introduced as necessary conditions for treatment. However, it is an illusion to believe that the procedure could be specified to such an extent that no prospective parent will be excluded who in fact would have provided adequate upbringing and that all accepted candidates will be competent parents.
By accepting lesbians, single women etc. for infertility treatment, we do not have to deny that heterosexual parents are better or that younger people are more competent parents. Our position is not threatened if this ever could be shown to be true. We only have to show, and the evidence at the present time is already very convincing on this point, that they make acceptable and suitable parents. The amount of welfare of the child may be lower than the level that could be expected in ideal circumstances but it may still be optimal given the concrete circumstances and characteristics of the parents. The use of the reasonable welfare principle implies that conception at a certain point in time and in certain circumstances might be worse than conception at a different time and in different circumstances but that does not entail that the decision is unacceptable. In other words, the welfare of the child might not be as high as it could have been (i.e. had the child had different parents) but it is sufficiently high to be considered a positive gift to the child.
The reduction of the moral domain
The present point can be added to the remarks expressed by Blyth and Cameron, who made some reservations about the welfare of the child test and warned against putting too great a weight on this aspect (Blyth and Cameron, 1998). While their warning concerns the possibility that contentious issues like the individual rights and informed consent might be neglected due to the status accorded to the welfare of the child, I caution about the one-sided attention paid to consequentialist arguments. Mumford et al. (Mumford et al.,1998
) stated that the welfare of the child-to-be `is also the basis of ethical and conscientious opposition of some medical staff to the provision of treatment to certain groups of people.' However, the deontological arguments should be clearly separated from the consequentialist arguments. Mixing them can lead to disappointment for those who think they have a decisive and irrefutable argument in hand to decide the issue once and for all. The welfare of the child, i.e. the extent to which its needs are fulfilled, is to a large extent objectively measurable in the sense that it can be determined whether there is a deficiency in the satisfaction of one of the child's needs. This deficiency will be expressed in some emotional, relational, behavioural or other problem. The empirical psychological (and medical) research is meant to verify whether the means of conception or the use of donor gametes has a negative influence on the offspring born by means of any kind of medical assistance. The body of evidence on the lesbian families for instance shows that children in these families do not differ significantly from children in heterosexual families on gender development, emotional and behavioural adjustment and social relationships (Brewaeys and van Hall, 1997
). We undoubtedly need these long-term follow-up studies of the children to decide the consequentialist debate (Brewaeys, 1998
). But, even if this debate could be closed, a large part of the people who currently refuse access to certain groups will not be converted by such findings. This is not because they are irrational or stubborn, but because this argument is directed at the wrong cause. The welfare of the child is associated with the conception of the ideal family (Golombok, 1998
). This ideal is a moral background theory that is linked with religious and secular conceptions of the family, procreation, sexuality, marriage etc. Although there is an indirect link with the happiness of the people involved (the ideal would not be an ideal if the people who lived in that kind of family would be very distressed) the well-being is not the ultimate justification of the ideal. A person should not deviate from the norm even if he is terribly unhappy with his relationship (most religions oppose divorce). It is precisely the independence of the ideal of the family from the happiness of the people involved which explains why scientific information about the welfare of the children cannot decide the issue. The opponents of the controversial applications will, of course, use all the arguments that reinforce their position (including consequentialist ones) but they will ignore the welfare argument if it contradicts their ideal of the family.
The deontologists frequently refer to the rights of the child to condemn certain interventions. However, moral rights are meant to protect a person's interests. We have to know what interests a person has before we can postulate the existence of a right. In the context of the new reproductive technologies, all kinds of queer rights are invented: the right to be raised by one's genetic parents, the right to be brought up in a stable family, the right to have a father, the right to have young parents etc. Without preceding proof of the link between the interests of the offspring and the rights, these so-called rights are nothing more than expressions of the ideal of the family. They are mere reformulations which add nothing new to the moral position to which they already confessed. The opponents claim a right to have a father which is violated by the treatment of a lesbian and subsequently they conclude that the child is harmed. But this is not the way it should work: they have put the cart before the horse. Once it has been shown that for instance being brought up by lesbian parents causes considerable harm to the child, one can introduce a right to protect the interests of the children. But since it has been demonstrated neither for single nor for lesbian parents that the children have serious psychological or social problems because of the setting in which they are raised, there is no reason to posit a right to have a father. I want to add three short arguments to demonstrate the independence of the deontological and consequentialist arguments and the limited usefulness of the welfare of the child argument. Firstly, most people would not conclude that procreation by lesbians should be promoted if it were proven that their parenting is superior to that of heterosexual couples. This point indicates the most important counter-argument against the supposed link between some controversial aspect in the parents' life and the welfare of their children. The true determining factors for the child's well-being (strong desire for parenthood, warm and supportive relationships etc) do not coincide with and are not (mainly) determined by the sexual orientation, the number of parents or the genetic relatedness (Golombok, 1998). If we have the welfare of the child in mind, we ought to select on those characteristics and conditions which have a proven influence on the well-being and happiness and not on ideologically or religiously based features. Secondly, the `conscience clause' for medical personnel implies a pluralism of moral opinions which cannot be reduced by objective data (Pearn, 1997
). If the question of the welfare of the child were the only conclusive element to decide whether an application is right or wrong, it would be strange if some people could state against the evidence that for them the application is still not right. If they do, this is not because ethics is a subjective matter. It is because the consequentialist argument of the welfare of the child is not the sole, let alone the most important relevant moral factor in the evaluation. Deontological reasons and normative conceptions do intervene in the moral decision making process.
The `conscience clause' should only be abolished when a general consensus is reached that the denial of access of some controversial groups to infertility treatment is a form of discrimination. Until then, everyone, and more specifically gamete donors whose contribution is also essential to the process, should have the right to appeal to such a clause (Pennings, 1995a). Finally, if the welfare of the child were the matter which really concerns us most, we would try to improve the welfare of the children regardless of the way they were conceived. The decrease of the happiness due to the fact that the child lives in a one-parent household may be negligible compared to the decrease in welfare caused by the poor financial situation of the mother. By not supporting needy parents more, we state that conforming to the ideal of the family is more important than the welfare of the children. This also shows that isolating the evaluation of the medically assisted procreation techniques from the general social, cultural and economic context leads to biased decisions.
To conclude, when we have the results of the long-term empirical research projects on the welfare of the children brought into existence by means of the new reproductive technologies in `special' circumstances, we will know whether they are harmed by those circumstances. Although the research at present does not exclude conclusively that this is not the case, the findings in general point at a welfare level which seems to fall well within the range of acceptable happiness. Once we discard the maximum welfare principle and adopt the reasonable well-being standard, as we do for the evaluation of parental responsibility in other instances, there are no indications that the technology is getting out of hand. Still, some extreme (even if rare) instances (as very old mothers, some examples of posthumous reproduction) do request continued vigilance. The renewed attention for the welfare of the child can serve as a counterweight against the overextended autonomy of the parent(s) in those situations.
Notes
This debate was previously published on Webtrack 54, February 23, 1999
References
Acock, A.C. and Demo, D.H. (1994) Family Diversity and Well-being. Sage, Thousand Oaks, London, New Delhi.
Arras, J.D. (1990) AIDS and reproductive decisions: having children in fear and trembling. Milbank Mem. Fund Q., 68, 353382.
Blyth, E. and Cameron, C. (1998) The welfare of the child: an emerging issue in the regulation of assisted conception. Hum. Reprod., 13, 23392342.
Brewaeys, A. (1998) How to care for the children? The need for large scale follow-up studies. Hum. Reprod., 13, 23472349.
Brewaeys, A. and van Hall, E.V. (1997) Lesbian motherhood: the impact on child development and family functioning. J. Psychosom. Obstet. Gynecol., 18, 116.[ISI][Medline]
Daniels, N. (1985) Just health care. Cambridge University Press, Cambridge.
Editorial (1993) Too old to have a baby? Lancet, 341, 344345.[ISI][Medline]
Golombok, S. (1998) New families, old values: considerations regarding the welfare of the child. Hum. Reprod., 13, 23422347.
Golombok, S. and Fivush, R. (1994) Gender development. Cambridge University Press, Cambridge.
Harris, J. (1990) Wrongful birth. In Bromham, D.R., Dalton, M.E. and Jackson, J.C. (eds), Philosophical Ethics in Reproductive Medicine. Manchester University Press, Manchester, New York, pp. 156170.
Jecker, N.S. (1987) Reproductive risk taking and the nonidentity problem. Soc. Theory Practice, 13, 219235.
Kavka, G.S. (1982) The paradox of future individuals. Phil. Public Affairs, 11, 93112.[ISI]
Kleinig, J. (1978) Crime and the concept of harm. Am. Phil. Quart., 15, 2736.
Lafollette, H. (1980) Licensing parents. Phil. Public Affairs, 9, 182197.[ISI]
Leiblum, S.R. and Williams, E. (1993) Screening in or out of the new reproductive options: who decides and why. J. Psychosom. Obstet. Gynecol., 14, 3744.[ISI]
Morreim, E.H. (1988) The concept of harm reconceived: a different look at wrongful life. Law Phil., 7, 333.
Mumford, S.E., Corrigan, E. and Hull, M.G.R. (1998) Access to assisted conception: a framework of regulation. Hum. Reprod., 13, 23492355.
National Advisory Board on Ethics in Reproduction (1996) Report and recommendations. In Cohen, C.B. (ed.), New Ways of Making Babies: the Case of Egg Donation. Indiana University Press, Bloomington, IN, USA, pp. 233320.
Nolan, K. (1990) Human immunodeficiency virus infection, women, and pregnancy. Obstet. Gynecol. Clinics North Am., 17, 651668.[ISI][Medline]
Parfit, D. (1984) Reasons and Persons. Clarendon Press, Oxford, UK.
Pearn, J.H. (1997) Gatekeeping and assisted reproductive technology: the ethical rights and responsibilities of doctors. Med. J. Austr., 167, 318320.[ISI]
Pennings, G. (1995a) Should donors have the right to decide who receives their gametes? Hum. Reprod., 10, 27362740.[Abstract]
Pennings, G. (1995b) Age and assisted reproduction. Med. Law, 14, 531541.[Medline]
Robertson, J.A. (1994) Children of Choice: Freedom and the New Reproductive Technologies. Princeton University Press, Princeton, NJ, USA.
Steinbock, B. and McClamrock, R. (1994) When is birth unfair to the child? Hastings Center Report, 24, 1521.
Strong, C. and Schinfeld, J.S. (1984) The single woman and artificial insemination by donor. J. Reprod. Med., 29, 293299.[ISI][Medline]