Beyond Therapy: Biotechnology and the Pursuit of Perfection

Table of Contents

Better Children

I. Improving Native Powers:
Genetic Knowledge and Technology

A. An Overview

The possibility of using genetic knowledge and genetic engineering to improve the human race and its individual members has been discussed for many years, especially in the heady decades immediately following Watson and Crick's discovery, in 1953, of the structure of DNA. New life was breathed into old eugenic dreams, which had been temporarily discredited by the Nazi pursuits of a "superior race." As late as the early 1970s, serious scientists talked optimistically about humankind's new opportunity to take the reins of its own evolution, thanks to the predicted confluence of genetic engineering and reproductive technologies.¹ But as scientists have learned just how difficult it is to engineer precise genetic change-even to treat individuals with genetic diseases caused by a simple one-gene mutation-explicit talk about improving the species has largely faded. Instead recent years have seen, in its place, much talk about coming prospects for "designer babies," children born with improved genetic endowments, the result either of careful screening and selecting of embryos carrying desirable genes, or of directed genetic change ("genetic engineering") in gametes or embryos.

Interest in such possibilities has been fueled by recent developments in a number of related disciplines, beginning with the completion of the Human Genome Project. Knowledge of the complete chemical sequence of all human genes promises greatly increased powers for genetic screening of individuals and embryos. Numerous studies are already seeking to correlate phenotypic traits (and not only those connected with disease) with the presence or absence of certain genetic markers. Scientists have reported early success with directed genetic change in embryos of non-human animals (including primates²), though many more attempts have failed. And we are witnessing large increases in the use of assisted reproductive technologies, including for purposes that go beyond the mere treatment of infertility.³ Extrapolating from these developments, some scientists have predicted that parents, in the not-too-distant future, will be able to exert precise genetic control over many characteristics of their offspring.⁴ These predictions have been greeted both with enthusiasm-"At last, we can escape from the tyranny of fortune and bring our inheritance under rational control!"-and with alarm-"What hubris! Scientists are trying to play God!"

It is difficult to know what to make of these predictions, based as they are largely on speculation. In this enormously fertile and rapidly developing field, the future is unknowable. Thus, anyone can claim to be a prophet, and no one should confidently bet against any form of scientific and technological progress. Yet in our view, for reasons that we shall elaborate below, prophecies and predictions of a "new (positive) eugenics" seem greatly exaggerated. In consequence, much of the public disquiet created by loose talk of genetically engineered "designer babies" seems unwarranted. Nevertheless, the public's misgivings may contain a partial wisdom regarding practices in this area that are not far-fetched, indeed, that are already with us, including prenatal and preimplantation genetic screening. For, as we shall see, there is some reason to be concerned both about negative eugenics and about the practice of genetic selection of "better" children. Therefore, even as we try to calm down fears about genetic engineering of children, it behooves us to pay careful attention to the reasons behind them and to the human goods at stake. By this means, we may shed light on the meaning not only of things we might be doing in the future but also of things we are already doing in the present.

B. Technical Possibilities

One can distinguish several ways of trying to produce children with better genetic endowments. First is the use of directed mating, either choosing "superior" mating partners or using donor sperm or donor eggs (or both) obtained from "superior" individuals. Assuming that people with some superior natural ability or accomplishment are genetically better endowed, and, further, that such putative genetic excellence is heritable, directed mating of like with like, so the theory goes, would increase the odds of getting superior children. People seeking to initiate a pregnancy using artificial insemination by donor (AID) or in vitro fertilization (IVF) with donor eggs do check the pedigree (and will soon be able to check the genetic profile) of the prospective donor for general health and fitness, as well as for certain desired traits, from height and hair color to intelligence. In some notorious cases, people planning to undergo IVF have advertised in elite college newspapers, offering up to $100,000 for an egg donor with high SAT scores or "proven college-level athletic ability."⁵ Yet these approaches to genetic improvement are relatively crude and probably unreliable, since they all involve the uncertain lottery of chance inherent in all sexual reproduction, and they overestimate the degree to which heredity by itself determines traits such as intelligence or athletic ability. Moreover, most couples would rather have their own children than those they might get by using gametes from a "superior" donor.ⁱ We will not be discussing this approach further.

We concentrate instead on various powers that depend upon precise genetic knowledge and technique: (a) the ability to screen and select fetuses, embryos, and gametes (egg and sperm) for the presence or absence of specific genetic markers; and (b) the ability to obtain and introduce such genetic material in order to effect a desired genetic "improvement." The first, by itself, leads to two powers that merely select from among genetic endowments conferred by chance, the difference between them being the stage at which screening is done and whether selection is "negative" or "positive." Prenatal diagnosis during an established pregnancy (using amniocentesis or chorionic villus sampling) permits the weeding out, through abortion, of those fetuses carrying undesired genetic traits.ⁱⁱ Preimplantation genetic screening and selection of in vitro embryos, in contrast, permits pregnancy to begin using only those embryos that carry desired genetic traits.ⁱⁱⁱⁱ In contrast to both of these, a third power, directed genetic change (or genetic engineering), would attempt to go beyond what chance alone has provided, improving in vitro embryos directly by introducing "better" genes.^iv

In theory, these three prospects offer scientists and prospective parents a range of increasing genetic control, from (1) eliminating the bad ("screening out"), through (2) selecting the good ("choosing in"), to (3) redesigning for the better ("fixing up"). Each activity raises its own ethical questions, some of which we shall consider later. But in practice, they are not equally feasible as means of producing better children, and, for reasons discussed below, we believe that the scale of their use for this purpose will probably remain low.

We state the conclusion in advance: The first, prenatal diagnosis and selective abortion, widely practiced since the 1970s in order to prevent the birth of children with genetic or chromosomal abnormalities, is a weeding-out procedure; hence its potential to select "better than normal" babies is negligible, and it is unlikely ever to be effective or widely used for such purposes.^v The third and most ambitious, genetic engineering of improved children, is-contrary to much loose prediction-a most unlikely prospect, for reasons of both feasibility and safety. The second, selecting IVF embryos genetically predisposed to certain superior or desirable traits, might soon be possible for some relatively uncomplicated traits (for example, height or leanness). Yet even here, as we shall see, there will likely be large-perhaps insurmountable-logistical problems in obtaining a "genetically superior" embryo for any trait to which many different genes contribute. Moreover, absent certain innovations in technology (and greater insurance coverage for assisted reproduction procedures), this is unlikely to be a widespread practice in the near future, save for those who are willing and able to undergo IVF and to pay extra for the genetic screening. Finally, keeping in mind that most traits of interest to parents seeking better children are heavily influenced by environment, even successful genetic screening and embryo selection might not, in many cases, produce the desired result.

We look briefly at each of the alternatives.

1. Prenatal Diagnosis and Screening Out.

Genetic screening by amniocentesis or chorionic villus sampling is an established feature of prenatal care in the United States and other economically advanced countries. It is routinely offered to women of advanced maternal age or to parents known to be carriers of heritable disorders. Some prospective parents prefer not to screen and not to know, in many cases because they have decided that they will not abort, no matter what. But the use of the practice is growing, and it will in all likelihood continue to do so. The capacity for screening both parents-to-be and fetuses is certain to increase, thanks to the completed mapping of the human genome and to greatly improved efficiency of testing. In addition to detecting more genetic diseases, new screening powers may also be able to detect a growing number of genetic markers that correlate statistically with the presence (or absence) of certain heritable-and desirable-traits (for example, tallness, leanness, perfect pitch, longevity, and perhaps even temperament and eventually intelligence). For parents willing to abort and try again repeatedly, prenatal screening could in principle be used to try to land a "better"-and not just a disease-free-baby. But, in practice, such an approach-even leaving ethical issues aside-is unfeasible on scientific grounds. No genetic selection can "optimize" beyond what the parents have contributed to the fetus. Moreover, an enormous number of "trial pregnancies" would be needed to get an "optimum baby" for any polygenic trait. For all these reasons this entire approach strikes us as far-fetched, and we shall not consider it further as a realistic possibility.

Yet, before leaving this subject, we think it important to observe that the existence and normalization of prenatal diagnosis and abortion for genetic defect have already had significant effects on our thinking: about our genetic endowments, about reproductive choice and responsible parenthood, and about what constitutes a good or "good enough" child. Attitudes and opinions acquired in connection with this practice will certainly influence how we are likely to think about and deal with the coming new techniques for selecting or altering our prospective children. The ethical issues will be discussed in greater detail later, in the section devoted to them. To prepare that discussion, it is worth noting a few salient facts about the current practice of prenatal diagnosis and some of its social implications-regarding medicine, children, and parental prerogative and responsibility.

First, prenatal diagnosis has enabled many couples to avoid the sorrows and burdens of rearing children with severe genetic and chromosomal disorders. Anyone who has been close to families having children with Tay-Sachs disease or anencephaly knows the anguish and misery that are now preventable by such means. Children born with these and comparable abnormalities endure serious and lifelong physical and mental disabilities. With certain of the conditions, postnatal care can restore some hope of a normal life; with others, such care is moderately palliative at best, and the children afflicted by these diseases are often destined to live relatively short lives marked by persistent physical pain and profound mental retardation. Without the option of prenatal screening, many couples at high risk for such genetic abnormalities would choose not to bear children at all; prenatal screening has also enabled women who have already given birth to an affected child or who are past the age of thirty-five (when the risk of chromosomal abnormalities begins to rise sharply) to become pregnant with some confidence of bearing healthy children.

Yet, second, to achieve these benefits prenatal diagnosis adopts a novel approach to preventive medicine: it works by eliminating the prospective patient before he can be born. This kind of preventive medicine is thus in fact a species of negative eugenics-elimination of the genetically unfit and a reduction in the incidence of their genes-albeit carried out voluntarily and on a case-by-case basis. It is true that the tests themselves are value-neutral and that many genetic counselors are committed to non-directive counseling, leaving prospective parents free to exercise their individual choices based on their own value judgments. Yet the very availability of these tests-accompanied in many cases by subtle pressures, applied by counselors (and others) to prospective parents, to abort any abnormal fetus-strongly implies that certain traits are or should be disqualifying qualities of life that justify prevention of birth.

Third, the practice of prenatal screening has established as a cultural norm (or at least as a culturally acceptable norm) a new notion about children: the notion that admission to life is no longer unconditional, that certain conditions or traits are disqualifying. To be sure, parents confronted with the painful decision whether or not to abort an affected fetus may feel deeply divided and moved by considerations on both sides of the issue, but there appears to be a growing consensus, both in the medical community and in society at large, that a child-to-be should meet a certain (for now, minimal) standard to be entitled to be born. Although, at least in the United States, the practice of screening and elimination is likely to remain voluntary, its growing use could have subtly coercive consequences for prospective parents and could increase discrimination against the "unfit." Children born with defects that could have been diagnosed in utero may no longer be looked upon as "Nature's mistakes" but as parental failings.

Finally, the practice of prenatal screening establishes the principle that parents may choose the qualities of their children, and choose them on the basis of genetic knowledge. This new principle, in conjunction with the cultural norm just mentioned, may already be shifting parental and societal attitudes toward prospective children: from simple acceptance to judgment and control, from seeing a child as an unconditionally welcome gift to seeing him as a conditionally acceptable product. If so, these changes in attitude might well carry over beyond choices confined to the presence or absence of genetic diseases, to the presence or absence of other desired qualities. Far from producing contentment and gratitude in the parents, such changes might feed the desire for better-and still better-children.

2. Genetic Engineering of Desired Traits ("Fixing Up").

With directed genetic change aimed at producing certain desired improvements, we enter the futuristic realm of "designer babies." Proponents have made this prospect look straightforward, and, on a theory of strict genetic determinism, it is. One would first need to identify all (or enough) of the specific variants of genes whose presence (or absence) correlates with certain desired traits: higher intelligence, better memory, perfect pitch, calmer temperament, sunnier disposition, greater ambitiousness, etc. Once identified, the requisite genes could be isolated, replicated or synthesized, and then inserted into the early embryo (or perhaps into the egg or sperm) in ways that would eventually contribute to the desired phenotypic traits. In the limit, there is talk of babies "made to order," embodying a slew of desirable qualities acquired with such genetic engineering. But in our considered judgment, these dreams of fully designed babies, based on directed genetic change, are for the foreseeable future pure fantasies. There are huge obstacles, both to accurate knowing and to effective doing. One of these obstacles-the reality that these traits are heavily influenced by environment-will not be overcome by better technology.

Most of the traits for which parents might wish to engineer improvements in their children-appearance, intelligence, memory-are most certainly polygenic, that is, traits (or phenotypes) that depend on specific genes or their variants at several, perhaps many, distinct loci. In such cases the relationships and interactions among these genes (and between one's genes and the environment) are certain to be enormously complex.^vi Isolating all the relevant genetic variants, and knowing how to work with them to produce the desired result, will therefore prove immensely difficult. To be sure, not every trait for which parents might wish to select need turn out to be highly polygenic: for example, height, skin color, eye color, or even the genetic contributions to sexual orientation or basic temperament might be heavily influenced by a very few genes. As we will see more fully in Chapter Four, one mutation in a single gene has been shown to result in enormous increases in the lifespan of flies, worms, and mice, and the same gene has been identified in humans. Yet even here there would be no guarantee that the predisposing genes, even if correctly and safely introduced into the zygote or early embryo, would necessarily express themselves as desired, to yield the sought-for improvement.

Even more of an obstacle to successful genetic engineering is the practical difficulty of inserting genes into embryos (or gametes) in ways that would produce the desired result and only the desired result. Getting the genes into the right place in the cell, able to function yet without disturbing regular cellular functions, is an enormously challenging task. Insertion of genes into the host genome can cause abnormalities, either by activating harmful genes or by inactivating useful ones. Recently, for example, children undergoing experimental gene therapy for immune system deficiencies have developed leukemia after retroviral gene transfer into bone marrow stem cells, very likely the result of activation of a cancer-producing gene by the virus used to transfer the therapeutic genes into the cell.⁶ And should introduced genes become inserted into inappropriate locations, normal host genes could be inactivated. Moreover, because many genes are pleiotropic-that is, they influence many traits, not just one-even a properly inserted gene introduced to enhance a particular trait would often have multiple effects, not all of them for the better.

Running such risks might be justified in gene therapy efforts for already existing individuals, where the genes hold out the only hope of cure for an otherwise deadly disease. But these safety risks will pose formidable obstacles to all interventions in gametes or embryos, especially nontherapeutic interventions aimed at producing children who would allegedly be, in one respect or another, "better than well." It is difficult to see how such an intervention could ever be considered ethical, especially since the negative effects might extend to future generations.

As a possible way around the hazards of gene insertion, some researchers have proposed the assembly and injection of artificial chromosomes: the new "better" genes could be packaged in small, manufactured chromosomal elements that, on introduction into cells, would not integrate into any of the normal forty-six human chromosomes. Such artificial chromosomes could, in theory, be introduced into ova or zygotes without fear of causing new mutations. But methods would have to be found to guarantee the synchronized replication and normal segregation of such artificial chromosomes. Otherwise, the package of improved genes, once introduced into the embryo, would not be conserved in all cells after normal mitotic division. Even more dauntingly, any gene introduced on such a chromosome would now be present in three copies (one from mother, one from father, and one on the extra chromosome) instead of the usual two, throwing off the normal balance of gene copies among all the genes. The consequences of such "triploidy" can be deleterious (for example, Down syndrome). All in all, safety and efficacy standards would seem to preclude doing such experiments with human subjects, at least in the United States, for the foreseeable future.^vii It is true that research along these lines might be undertaken in other countries (for example, China), by scientists unconstrained by these considerations, with eventual success in effecting directed genetic change in human embryos. But, at least for the time being, we believe that we may set this prospect safely to the side.

3. Selecting Embryos for Desired Traits ("Choosing In").

Unlike the prospect for precise genetic engineering through directed genetic change, the possibility of genetic enhancement of children through embryo selection cannot be easily dismissed. This approach, less radical or complete in its power to control, would not introduce new genes but would merely select positively among those that occur naturally. It depends absolutely on IVF, as augmented by the screening of the early embryos for the presence (or absence) of the desired genetic markers, followed by the selective transfer of those embryos that pass muster. This would amount to an "improvement-seeking" extension of the recently developed practice of preimplantation genetic diagnosis (PGD), now in growing use as a way to detect the presence or absence of genetic or chromosomal abnormalities before the start of a pregnancy.

As currently practiced, PGD works as follows: Couples at risk for having a child with a chromosomal or genetic disease undertake IVF to permit embryo screening before transfer, obviating the need for later prenatal diagnosis and possible abortion. A dozen or more eggs are fertilized and the embryos are grown to the four-cell or the eight-to-ten-cell stage. One or two of the embryonic cells (blastomeres) are removed for chromosomal analysis and genetic testing. Using a technique called polymerase chain reaction to amplify the tiny amount of DNA in the blastomere, researchers are able to detect the presence of genes responsible for one or more genetic disorders.^viii Only the embryos free of the genetic or chromosomal determinants for the disorders under scrutiny are made eligible for transfer to the woman to initiate a pregnancy.

The use of IVF and PGD to move from disease avoidance to baby improvement is conceptually simple, at least in terms of the techniques of screening, and would require no change in the procedure. Indeed, PGD has already been used to serve two goals unrelated to the health of the child-to-be: to pre-select the sex of a child, and to produce a child who could serve as a compatible bone-marrow or umbilical-cord-blood donor for a desperately ill sibling. (In the former case, chromosomal analysis of the blastomere identifies the embryo's sex; in the latter case, genetic analysis identifies which embryos are immunocompatible with the needy recipient.) It is certainly likely that blastomere testing can be adapted to look for specific genetic variants at any locus of the human genome. And even without knowing the precise function of specific genes, statistical correlation of the presence of certain genetic variants with certain phenotypic traits (say, with an increase in IQ points or with perfect pitch) could lead to testing for these genetic variants, with selection following on this basis. As Dr. Francis Collins, director of the National Human Genome Research Institute, noted in his presentation to the Council, the time may soon arrive in which PGD is practiced for the purpose of selecting embryos with desired genotypes, even in the absence of elevated risk of particular genetic disorders.⁷ Dr. Yury Verlinsky, director of the Reproductive Genetics Institute in Chicago, has recently predicted that soon "there will be no IVF without PGD."⁸ Over the years, more and more traits will presumably become identifiable with the aid of PGD, including desirable genetic markers for intelligence, musicality, and so on, as well as undesirable markers for obesity, nearsightedness, color-blindness,^ix etc.

Yet, as Dr. Collins also pointed out to the Council, there are numerous practical difficulties with this scenario. For one thing, neither of the parents may carry the genetic variant they are most interested in selecting for. Also, selecting for highly polygenic traits would require screening a large number of embryos in order to find one that had the desirable complement. With only a dozen or so embryos to choose from, it will not be possible to optimize for the many necessary variants.^x

The practice of PGD and selective transfer is still quite new, and fewer than 10,000 children have been born with its aid. How likely or widespread such a practice might become is difficult to predict. As we have already indicated, a number of practical issues would need to be addressed before PGD could be extended to permit selection of desirable traits beyond the absence of genetic disorders. First are questions of possible harm caused by removing blastomeres for testing (up to a sixth or even a quarter of the embryo's cells are taken). Although current evidence (from limited practice) suggests that the procedure inflicts neither any immediately visible harm on the early embryos, nor any obvious harm on the child that results, more attention to long-term risks to the child born following PGD is needed before many people would consider using it for "improvement" purposes only. Because many of the desirable human phenotypic traits are very likely polygenic, the contribution of any single gene identifiable by blastomere testing is likely to be small, and the likelihood of finding all the "desired" genetic variants in a single embryo is exponentially smaller still. Testing for multiple genetic variants using the DNA from a single blastomere is likely to be limited-for a time-by the quantities of DNA available, the sensitivity of the genetic tests, and the ability to perform multiple tests on the same sample. But it seems only a matter of time before techniques are perfected that will permit simultaneous screening of IVF embryos for multiple genetic variants. And should some of the "desirable" genes come grouped in clusters, selection for at least some desired traits might well be possible.

Finally, even if PGD could be used successfully to select an embryo with a number of desirable genetic variants, there is simply no guarantee that the child born after this procedure would grow up with the desired traits. The interplay of nature and nurture (genes and environment) in human development is too complex and too little understood to make such results predictable. Given that IVF combined with PGD is an inconvenient and expensive alternative to normal procreation, and given that success is doubtful at best, the purely elective use of this procedure seems unlikely to become widespread in the foreseeable future. As Professor Steven Pinker put it, in his presentation to the Council:

The choice that parents would face in a hypothetical future in which even genetic enhancement were possible would not be the one that's popularly portrayed, namely, "Would you opt for a procedure that would give you a happier, more talented child?" When you put it like that, well, who would say no to that question? More realistically, the question that parents would face would be something like this: "Would you opt for a traumatic and expensive procedure that might give you a very slightly happier and more talented child, might give you a less happy, less talented child, might give you a deformed child, and probably would do nothing?"⁹

Nevertheless, we think it would be imprudent to ignore completely this approach to "better children." More and more people are turning to assisted reproduction technologies (ART): in parts of western Europe, roughly five percent of all births involve ART; in the United States, it is roughly one percent and climbing, as the average maternal age of childbirth keeps rising and family size keeps declining. More and more people are using IVF not merely to overcome infertility but to screen and select embryos free of certain genetic defects. Women who plan to delay childbearing are being encouraged to consider early removal and cryopreservation of their own youthful ovarian tissue, to be reintroduced into their bodies at sites easily accessible for egg harvesting when they decide to have children. Other novel methods of obtaining supplies of eggs for IVF-possibly including deriving them in bulk from stem cells¹⁰-would make the procedure less burdensome, and would, in theory, permit the creation of a large enough population of embryos to make screening for polygenic traits feasible.

The anticipated vast extension of genetic screening will make many more couples aware of the risks they run in natural reproduction, and they may choose to turn to IVF to reduce them-especially if obtaining eggs became easy. Once more and more couples start screening embryos for disease-related concerns, and once scientists have identified those genes that correlate with various admirable traits, the anticipated expansion of improved and more precise screening techniques might enable users of IVF to screen for "desirable genes" as well. People already using PGD to screen for disease markers might seek information also about other traits, as they have with sex or histocompatibility. And if, once screening becomes automated, its cost comes down, or if society decides to reimburse for PGD (regarding it as less expensive than the care of genetically diseased children), the use of this approach toward "better children" might well become the practice of at least a significant minority. Under these circumstances, should genuine and significant improvements be achieved for a few highly desired attributes (say, in maximum lifespan; see Chapter Four), one can easily imagine that there would be an increased demand for the practice, inconvenient or not. In the meantime, we would do well to consider the ethical implications not only of such future prospects but also of our current practices that make use of genetic knowledge.

C. Ethical Analysis

The technologies we have just considered range from the well-established (prenatal "screening out," using amniocentesis and abortion) to the speculative (embryonic "fixing up," using direct genetic modification of embryos or gametes), with special attention to the new and growing ("choosing in," using preimplantation genetic diagnosis followed by selective embryo transfer). It bears emphasis that genetic technologies have been and are being devised mainly with the intention of producing healthier children-not "enhanced children" or "super-babies," but children who are better only in the sense of being free of severe disease and deformity. As we have suggested, we have our doubts whether these powers will soon be widely employed for any other purpose. Yet there are ample reasons why we should not become complacent or take these matters lightly.

Powers to screen and select for one purpose are immediately available to screen and select for another purpose; the same is true for powers of directed genetic change. And, as already noted, it is sometimes hard to distinguish between desirable traits that one would call "healthy" and those that one would call "good in some other way": consider the case of leanness (non-obesity) or perfect pitch (non-tone-deafness) or attentiveness (non-distractibility). Moreover, there is ample reason to take stock of the ethical and social issues related to present and anticipated practices of screening and selection even if, as we have indicated, there is no reason for alarm regarding "designer babies." For the confluence of ever more sophisticated techniques of assisted reproduction with ever greater capacities for genetic screening and manipulation is already increasing the intrusion of science and technology into human procreation, yielding to scientists and parents ever growing powers over the beginnings of human life and the native capacities of the next generation. In addition to welcome consequences for the health of children, such practices may have more ambiguous or worrisome consequences for our ideas about the relation of sex and procreation, parents and children, the requirements of responsible parenthood, and beliefs in the equal worth of all human beings regardless of genetic (or other) disability.

Before one can decide whether these changes should be welcomed enthusiastically, tolerated within limits, or met with disquiet, one must try to think through what they mean-for individuals, for families, and for the larger society. In what follows, we shall examine, first, the reasons why many people welcome these technologies; second, concerns that might be raised about the safety of these procedures and about equality of access to their use; and, finally, more profound ethical questions regarding how these technologies might affect family life and society as a whole.

1. Benefits.

There is no question but that assisted reproductive technologies have, over the past few decades, enabled many infertile couples to conceive and bear children, and that the more recent addition of PGD holds the promise of helping couples conceive healthy children when there is a serious risk of heritable disease. The widespread practice of prenatal screening in high-risk pregnancies has enabled numerous couples to terminate pregnancies when severe genetic disorders have been detected. It is the natural aspiration of couples not only to have children, but to have healthy children, and these procedures have in many cases lent crucial assistance to that aspiration. People welcome these technologies for multiple reasons: compassion for the suffering of those afflicted with genetic diseases; the wish to spare families the tragedy and burden of caring for children with deadly and devastating illnesses; sympathy for those couples who might otherwise forego having children, for fear of passing on heritable disorders; an interest in reducing the economic and social costs of caring for the incurable; and hopes for progress in the overall health and fitness of human society.^xi No one would wish to be afflicted, or to have one's child afflicted, by a debilitating genetic disorder, and the new technologies hold out the prospect of eliminating or reducing the prevalence of some of the worst conditions.^xii

Should it become feasible, many people would have reason to welcome the use of these technologies to select or produce children with improved natural endowments, above and beyond being free of disease. Parents, after all, hope not only for healthy children, but for children best endowed to live fulfilling lives. At some point, if some of the technical challenges are overcome, PGD is likely to present itself as an attractive way to enhance our children's potential in a variety of ways. Assuming that it became possible to select embryos containing genes that conferred certain generic benefits-for example, greater resistance to fatigue, or lowered distractibility, or better memory, or increased longevity-many parents would be eager to secure these advantages for their children. And they would likely regard it as an extension of their reproductive freedom to be able to do so; they might even regard it as their parental obligation. In a word, parents would enjoy enlarged freedom of choice, greater mastery of fortune, and satisfaction of their desires to have "better children." And, if all went well, both parents and children would enjoy the benefits of the enhancements.

2. Questions of Safety.

Needless to say, the matter is hardly this simple. As with all biomedical interventions, a primary ethical concern is the matter of safety: the risks of bodily harm incurred by those subject to the procedures involved in genetic screening and manipulation. As with all biomedical interventions in reproductive processes, the safety issue takes on special gravity and difficulty, precisely because some of the hazards will be inflicted on the unconsenting child-to-be, and in the very activities connected with his coming-into-being. The Council has previously dealt at length with this issue in its report on human cloning, Human Cloning and Human Dignity, to which the reader is referred.¹¹

There are, first of all, hazards connected to the various technological means employed in genetic screening and manipulation: risks to the pregnant woman, the egg donor (if different from the mother-to-be), and, most important, to the offspring. In the case of prenatal screening, whether by amniocentesis or chorionic villus sampling, there are well known, albeit slight, risks of infection, trauma (to both pregnant woman and fetus), miscarriage, and premature labor. These risks are weighed against the hazards of not screening, when the mother is of advanced reproductive age or when there is other evidence suggesting heightened risk of genetic defects in the fetus. Of course, prenatal screening serves to prevent genetic defects only if it is followed up by abortion, which, besides destroying the fetus, involves some potential health risks to the woman.

Regarding direct genetic manipulation of the germ line, we have already examined some of the considerable associated risks and uncertainties in the course of arguing that this technology is unlikely to be applied to humans any time soon.

Regarding the topic of greatest interest here, preimplantation diagnosis and selection, there are questions as to the long-term safety of blastomere biopsy. Although the technique of removing one or two cells from the eight-cell embryo for chromosome or DNA analysis does not appear to harm the embryo (at least in those cases in which it goes on to become a child), there are as yet no studies looking at long-term consequences for children born after blastomere biopsy. Such currently imponderable risks might be thought to recede in importance when severe genetic diseases are in prospect. However, if PGD were to be undertaken, not to screen out genetic defects, but to improve native powers, there should be heightened scrutiny of any possible dangers involved in the procedure.

To date, ethical thinking about the hazards of the techniques of assisted reproduction has often been incomplete, partly as a result of the perceived desirability of the end. IVF and PGD are undertaken with the intention of producing healthy, fit children; put this way, the enterprise would seem to be much like other medical practices and, as such, amenable to the same ethical standards. But a medical procedure designed to produce a healthy person has a different character from procedures aimed at safeguarding or healing a patient who is already alive. Yet here our thinking is ill-served owing to a noticeable lacuna in our approach to the ethics of risky therapies and (especially) the ethics of research using human subjects.

Ordinarily, when new technologies are introduced into medical practice or when medical research is undertaken with human subjects, the safety of the patients or subjects is of paramount ethical concern. However, in the case of IVF, with or without PGD, the children who are produced as a result of these procedures are not considered subjects at risk, for the simple reason that the embryos being handled, tested, and manipulated are not regarded as human subjects. Thus, blastomere biopsy performed on a tiny eight-cell embryo is not treated as an experiment on a human subject or as diagnosis of a patient, even though the future health and well-being of the child are very much at stake. Instead, the ethics of IVF and PGD are generally dealt with as though the only patient involved were the mother.^xiii Whether or not one believes that the embryo here manipulated is a fully human being worthy of moral and legal protection, it is certainly the essential (and fragile) beginnings of the child who will be born and whose health and well-being should therefore be of overriding concern.

A deeper safety question connected with the goal of genetic screening is whether the normal ethical standard-"the best interests of the patient"-can be said to apply if and when PGD is used to select a "better" child. Even when PGD is used only to screen out genetic diseases-and all the more when it is employed to select positive traits-the parents are in effect choosing a particular genotype for their child. The question is, will this unprecedented power in the hands of the parents necessarily be used for the good of the child? Should parents be willing to gamble the safety of their children for the chance to make them "better than well"? What risks to their health and safety are worth taking in pursuit of improvement or perfection?

Ordinarily, in most matters regarding children, our society accepts the principle that each set of parents has authority and responsibility for the well-being of their own children. Yet there are circumstances that lead the state to step in to protect a vulnerable child against abusive or negligent parents. In such cases, the best of parental intentions do not exonerate. How should our society view parental (and biotechnical) discretion to seek to produce "better children" through procedures carrying unknown hazards to those children?

These questions take on greater poignancy once we recognize a novel but morally significant feature of embryo selection using PGD, absent in prenatal diagnosis. In intrauterine genetic screening, there is one fetus being tested, and the question at issue is a binary choice of "keep" or "destroy." In contrast, in preimplantation screening a whole array of embryos are scrutinized and tested, and the choice is not the either-or "yes or no" but rather the comparative choice of "best in the class." For if one is going to the trouble of doing IVF supplemented by preimplantation diagnosis, why not get "the best"-the healthiest and, perhaps soon, the "better-than-healthiest"? But in order to get the best, or even in order to get a non-diseased child, one must conceptually "bundle" all the separate embryos and regard them as if they were a single precursor. All will be subjected to testing so that the one who is chosen will be disease-free or better. Yet to make sure that the child who is to be born is the fittest, rather than his diseased or inferior brother or sister, the anointed one must bear potential risks (imposed during the testing) that he would not have borne in the absence of the parental desire for quality control. For the sake of which benefits to the child can we justify imposing on him what kinds and what degrees of risk?

Before leaving the subject of safety and the concern for the health of children, we observe an ironic feature of the search for better babies with the aid of genetic screening. What if, as a result of widespread genetic screening of adults and improvement in diagnostic screening of embryos, the practice of IVF with PGD came to be seen as superior to natural procreation in offering a greater probability of obtaining a healthy child? If the procedures became sufficiently routine and inexpensive (to the point, say, where they are covered by ordinary health insurance), prospective parents interested in healthier (or otherwise better) children might increasingly be tempted to consider IVF with PGD. Furthermore, couples who would then elect PGD in order to screen out genetic diseases might well be tempted to engage at the same time in some positive trait selection. In that case, what began modestly as a means to help the infertile bear children and continued as a way to screen out the worst genetic defects might ultimately stand as a competitor to natural reproduction altogether, with significant consequences for the family and for society at large.^xiv

As this discussion indicates, the issue of health and safety proves, on further reflection, to concern more than safety. When biomedical technology permits the substitution, for natural procreation and the rule of chance, of a procedure in which parents begin to control their child's genotype, reproduction becomes to some extent like obtaining or making a product to selected specifications. Even if the parents are guided by their own sense of what would be a good or perfect baby, their selection may serve to satisfy their own interests more than that of the child. The new technologies, even when used only to screen out and get rid of the sick or "imperfect," imply a changed attitude of parents toward their children, a mixture of control and tacit expectations of perfection, an attitude that might grow more pronounced as the relevant techniques grow more sophisticated. Apparently good intentions-to improve the next generation, to enhance the life of our descendants-will not guarantee that genetic screening will be an unqualified blessing for parents and children. (We return to this subject shortly.)

3. Questions of Equality.

Many observers have noted with concern that, owing to the sheer expense of IVF and PGD-a successful assisted pregnancy costing, on average, roughly $20,000-$30,000^{12 xv}-not all couples who could benefit from these procedures have unfettered access to them. If PGD were to become an established option, but only for the affluent, one envisages the troubling prospect of a society divided between the economically and genetically rich, on the one hand, and the economically and genetically poor on the other. Severe inherited diseases might disappear except among the poor, while genetic enhancement through screening and selection might be a privilege enjoyed exclusively by the rich. These concerns would, of course, diminish (though they would not disappear) if, as seems likely, the costs of the procedures in question come down and access to these services grows wider.^xvi

Yet these legitimate concerns about equality of access rest, ironically, on certain inegalitarian assumptions that need to be brought to light. First, the goal of eliminating embryos and fetuses with genetic defects carries the unspoken implication that certain "inferior" kinds of human beings-for example, those with Down syndrome-do not deserve to live. The assumption that the genetically unfit ought to be prevented from being born embodies and invites a profoundly denigrating and worrisome attitude toward those who do get to be born. How will we come to regard the many people alive today who carry genetic defects that in the future will be screened out, or the many people, even in a future age of more widespread screening, who will still be born with the abhorred disabilities and diseases? The worry over unequal access to PGD is, in effect, a worry about the inability of the economically poor to practice the ultimate discrimination against the genetically poor.

Second, when new techniques permit parents to be the partial authors of their child's genetic makeup, the inequality between parents and children is substantially increased. Parents thereby acquire the power, not just of giving life to their children, but of shaping (or trying to shape) the character of that life. Of course, through education and upbringing parents have always had an enormous influence on the lives of their children; but inasmuch as the consequences of genetic screening are imposed before birth and are carried as the child's permanent biological destiny, the inegalitarian effect of the new technology is unprecedented and irreversible.

In response to these concerns, it will be pointed out, rightly, that genes are not exactly destiny, and that it will prove very difficult to intervene genetically at the embryonic stage in ways that will guarantee the appearance of the desired "improvements" in one's children.  But much mischief can be done to a child simply from the enhanced parental expectations, all the more so if the child fails to attain the superior native gifts for which he was selected. And as we shall soon see, we are already witnessing certain subtle forms of genetic discrimination even though the technology of screening is still very undeveloped.

4. Consequences for Families and Society.

Beyond questions of safety and equal access, there is reason to believe that the advent of expanded genetic screening and its uses in reproduction could have a profound impact on human procreation, family life, and society as a whole. At present, fewer than 10,000 children have been born following PGD, and the screening procedure itself is being used to diagnose only a limited number of chromosomal and genetic ailments. For these reasons, it is both difficult to predict and also easy to underestimate the societal import of marrying genomic knowledge with established techniques of assisted reproduction, should the practice become widespread.

To make vivid the possible implications, it may therefore be helpful to imagine a future time at which all external barriers to the use of these procedures have been largely removed.^xvii Suppose that, a decade from now, IVF and PGD have been perfected to the point where preimplantation screening is safe and effective, not prohibitively expensive, and capable of identifying a wide range of markers for heritable disorders. Suppose, in other words, that prospective parents (perfectly fertile) routinely have the option of using these technologies in order to select an essentially disease-free embryo for transfer to the mother's womb.^xviii

Under such circumstances-admittedly quite hypothetical-might not the practice become moderately widespread? Could many people come to regard using IVF plus PGD as safer (for the child) than the randomness of sex, and therefore preferable to natural procreation even when there is no particular history of genetic disease? In societies in which people are limited-or limit themselves-to only one child, might they not increasingly turn to these techniques to ensure that their child might be as "perfect" as possible? And, should this procedure begin to compete with or even to supplant sex as the more common route to conceiving children, in what ways would the meaning of childbearing be altered?

The hypothetical case just sketched may seem like science fiction, but the important questions it raises are, in fact, implicated in the current practice of genetic screening. Even though the practice of PGD is still in its infancy, its availability has begun to influence our thinking about childbearing. Already the goals of assisted reproductive technologies are changing, from the original modest aim of providing children for the infertile to the novel and more ambitious aim of producing healthy children for whoever needs extra assistance in obtaining them.^xix Anticipating the coming of augmented powers of genetic screening and selection, people are expanding the idea of "a healthy child" and therewith almost certainly the aspirations of prospective parents. In his presentation to the Council, Dr. Gerald Schatten, a leading researcher in the field of reproductive biology, stated that the overall goal of assisted reproductive technology is "to help prospective parents realize their own dreams of having a disease-free legacy" (emphasis added).¹³ The dream of a disease-free legacy-as stated, a goal that looks beyond merely the next generation-seems rather different from the merely hopeful wish for a healthy child. And even without such a broad ambition, the intervention of rigorous genetic screening into the order of childbearing will likely involve raising the standard for what counts as an acceptable birth. The likely significance of this fact is subtle but profound. The attitude of parents toward their child may be quietly shifted from unconditional acceptance to critical scrutiny: the very first act of parenting now becomes not the unreserved welcoming of an arriving child, but the judging of his or her fitness, while still an embryo, to become their child, all by the standards of contemporary genetic screening. Moreover, as the screening technology itself grows more refined, more able to pick out serious but not life-threatening genetic conditions (from dwarfism and deafness to dyslexia and asthma) and then genetic markers for desirable traits, the standards for what constitutes an acceptable birth may grow more exacting.

With genetic screening, procreation begins to take on certain aspects of the idea-if not the practice-of manufacture, the making of a product to a specified standard. The parent-in partnership with the IVF doctor or genetic counselor-becomes in some measure the master of the child's fate, in ways that are without precedent. This leads to the question of what it might mean for a child to live with a chosen genotype: he may feel grateful to his parents for having gone to such trouble to spare him the burden of various genetic defects; but he might also have to deal with the sense that he is not just a gift born of his parents' love but also, in some degree, a product of their will.

These questions of family dynamics could become even more complicated when preimplantation genetic screening is used to select embryos for some desirable traits. While current negative screening is guided by the standard of a healthy or disease-free baby, the goals of prospective positive use are in theory unlimited, governed only by the parents' ideas of what they want in their child. Today, parents using PGD take responsibility for selecting for birth children who will not be chronically sick or severely disabled; in the future, they might also bear responsibility for picking and choosing which "advantages" their children shall enjoy. Such an enlarged degree of parental control over the genetic endowments of their children cannot fail to alter the parent-child relationship. Selecting against disease merely relieves the parents of the fear of specific ailments afflicting their child; selecting for desired traits inevitably plants specific hopes and expectations as to how their child might excel. More than any child does now, the "better" child may bear the burden of living up to the standards he was "designed" to meet. The oppressive weight of his parents' expectations-resting in this case on what they believe to be undeniable biological facts-may impinge upon the child's freedom to make his own way in the world. Here we see one of the ethically paradoxical consequences of the new screening technologies: designed to free us from the tyranny of our genes, they may end up narrowing our freedoms as individuals even further.

In addition to changes in the parent-child relationship, there are reasons to be concerned about the wider social effects of an increased use of genetic screening and selection. There is, first of all, the prospect of diminished tolerance for the "imperfect," especially those born with genetic disorders that could have been screened out. It is offensive to think that children, suffering from "preventable" genetic diseases, should be directly asked, "Why were you born?" (or their parents asked, "Why did you let him live?"). Yet it is almost as troubling to contemplate that "defective" children and their parents may be treated contemptuously and unfairly in light of such prejudices, even if they go unspoken. Already, parents who have a child with Down syndrome are sometimes asked, "Well, didn't you have an amnio? How did this happen?" Many of these parents are people who, for their own ethical reasons, have chosen to proceed with the pregnancy even after learning the results of genetic screening, electing to love and care for the children that it has been given to them to love. Yet as the range of detectable disorders increases, as adult screening becomes ubiquitous and every pregnancy is tested, and as the economic cost of caring for the afflicted remains high, it may become difficult for parents to resist the pressure, both social and economic, of the "consensus" that children with sufficiently severe and detectable disabilities must not be born.

In all likelihood parents will increasingly feel pressure to conform to shifting social standards of what is genetically fit. Along with the freedoms bequeathed by the new technologies comes a certain danger of social coercion and tyranny of public opinion. Furthermore, as our table of detectable genetic markers grows more complete, there is the prospect of using genetic screening to weed out not only the most devastating genetic disorders but also heritable conditions that are bad but manageable, or even merely inconvenient. In practice, it is likely to prove very hard to draw a bright line between identifiable defects that might justify discarding an embryo or preventing a birth and those defects that parents might (or should) be able to find acceptable. It is not clear what resources our society will be able to draw upon to assist parents in making such important decisions.

Should PGD and IVF, contrary to current expectations, ever become widely used for positive screening of desirable traits, the impact on society could be even greater. Our knowledge of the human genome and our powers of genetic selection might grow so great as to unleash competition among parents eager to bear children who are biologically destined to be taller, thinner, brighter, or better-looking than their peers.

It should be noted that the social consequences of the widespread use of genetic screening alone are likely to outstrip the actual biological enhancements: those "unfortunate" enough to be born with genetic "defects" that might have been detected by screening might well be subject to discrimination, even without waiting to see how they turn out. The thoughtful (if not quite scientifically accurate) film Gattaca explores some of the chilling social implications of a human future in which genetic screening of children has become the norm. To the careful observer of current practices, the risks of such discriminatory implications are already evident.

II. CHOOSING SEX OF CHILDREN

There is one area in which parents are today already able to choose an important inborn characteristic of their children: sex selection and control. This practice is widespread in many countries around the world, and there is some evidence that it is being used with growing frequency in the United States.¹⁴ Strictly speaking, choosing the sex of children is not exactly a choice for a "better" child, save in those cultures in which one sex (usually male) is held to be superior or privileged (or more rewarding to the family economically). But, if "good" means "that which is desired," it is a choice for a child thought by the parents to be "better" in the limited, but significant, sense of "more wanted." In choosing a child of the preferred sex, the parents are acting to satisfy their own desire for what, to them, is better (at least here and now).

While it is true that what is being chosen here is nothing new or different-selection is confined to one or the other of the eternal alternatives, male or female-the choice is not for that reason trivial or free from moral implications. Parents choose a supremely important aspect of their child's lifelong identity, yet in most cases they do so not for the child's sake. They choose not because they think that the child will be better off being male rather than female, or the reverse, but because they now want a boy or a girl, or because they want to balance a family now lacking in one sex or the other.^xx

The seemingly innocent practice of sex selection in fact raises many of the larger ethical concerns introduced above: about changing the relations between parents and children, moving procreation toward manufacture, and expanding parental choice and mastery over the next generation. Moreover, what happens in the area of sex-selection may have implications for other, more far-reaching efforts to choose or control the genetic makeup of our offspring, if and when that becomes possible. Both for itself and as a precedent, it is worth considering on its own this more modest form of seeking "better children."

In considering the ethical implications of sex selection, we must attend especially to the social consequences not just of the fact of choice but of the choices made. For the private choices made by individuals, once aggregated, could produce major changes in a society's sex ratio, with profound implications for the entire community-and also its neighbors. Over the past several decades, disturbing evidence has accumulated of the widespread use of various medical technologies to choose the sex of one's child, with a strong preference for the male sex. The natural sex ratio at birth is 105 baby boys born for every 100 baby girls. But in several countries today the ratio approaches or even exceeds 120 baby boys born for every 100 girls. There is also evidence that the ratio at birth of boys to girls is rising among certain ethnic groups in the United States. This phenomenon especially calls out for our attention and demands a broad-ranging ethical and social evaluation.^xxi

A. Ends and Means

Sex selection offers a stark example of the marriage that can occur between modern technique, on the one hand, and ancient custom or primordial desire, on the other. For the human desire to choose the sex of one's offspring-usually to have a son rather than a daughter, but also on occasion a daughter rather than a son-is hardly new. The folk wisdom of times gone by attests to the enduring power of this human want, found in mothers and fathers alike. In ancient Greece, it was believed that if men had sex while on their right side, a boy would result; and in eighteenth-century France, it was recommended to men who wanted sons to tie off their left testicle during intercourse. In our own time, books that claim to reveal the secrets of having a boy or a girl abound, with one bestseller recounting myriad methods but recommending the timing of sexual intercourse as the key. Indeed, the importance to all of us of a baby's sex is revealed in the first question we nearly always ask upon news of a newborn (assuming that we have not already found out by sonogram): "Is it a boy or a girl?"

If the central importance of a baby's sex and our desires to choose it are old, the medical techniques for realizing our desires are new. The principal means for doing so are, first, prenatal diagnosis (either using a sonogram to disclose the genitalia or using amniocentesis or chorionic villus sampling to disclose whether the karyotype is XX, female, or XY, male), followed by abortion of fetuses having the unwanted sex. Second, preimplantation genetic diagnosis (PGD) followed by selective transfer of embryos having the desired sex. And third, a less certain technique, pre-fertilization separation of sperm into X- and Y-bearing spermatozoa,^xxii followed by artificial insemination or in vitro fertilization. The first two techniques select post-conception; the last seeks to produce the desired sex at the time of conception.

These methods were developed (or at least the first two were) to prevent disease. However, as with many other medical technologies, nontherapeutic uses were quickly discovered and put into practice. The techniques of amniocentesis and sonograms have been available respectively since the 1970s and 1980s and have become increasingly widespread. Amniocentesis can make a determination of sex at 16 to 18 weeks of gestation; sonograms at 15 to 16 weeks. PGD, the procedure (described earlier) to screen IVF embryos for chromosomal abnormalities and genetic diseases, has been available for about ten years. The newer and less tested sperm-sorting technology was originally a creation of the U.S. government, invented by a Department of Agriculture scientist in the 1980s for the purposes of selecting sex in livestock. The Genetics and IVF Institute in Fairfax, Virginia, developed the technology for humans and currently has an exclusive license on it-the technology is known as "MicroSort." The Institute charges about $2,300 per try, and currently claims a 90 percent success rate for girls and 73 percent success rate for boys. It offers this service only for the purpose of "family balancing"-that is, for achieving a mix of boys and girls in a family.

Even in just the short time that these various methods of sex selection have been available, they have had dramatic effects on sex ratios in many parts of the world. Generally, any variation in the sex ratio exceeding 106 boys born per 100 girls born can be assumed to be evidence of the practice of sex selection. Here, from the most recent figures available, are just a few examples of skewed sex ratios around the world today. The sex ratio at birth of boys to 100 girls in Venezuela is 107.5; in Yugoslavia 108.6; in Egypt 108.7; in Hong Kong 109.7; in South Korea 110; in Pakistan 110.9; in Delhi, India, 117; in China 117; in Cuba 118; and in the Caucasus nations of Azerbaijan, Armenia, and Georgia, the sex ratio has reached as high as 120.^xxiii While the sex ratio in the United States has remained stable at 104.8, certain American ethnic groups have seen a statistically significant rise in their sex ratios. In 1984 the sex ratio for Chinese-Americans was 104.6 and for Japanese Americans 102.6; in 2000, these ratios had risen respectively to 107.7 and 106.4.¹⁵

Imbalances in the sex ratio are certainly not evenly spread across every region of the globe. However, one cannot but be impressed by the fact that distortions in the sex ratio afflict developed as well as underdeveloped nations, Hindu and Moslem populations as well as Christian populations, Western as well as non-Western nations, wealthy and educated regions as well as those that are less so. Although the practice is, for now, greater outside than within the United States, the other nations are mainly using technologies that we have developed (albeit for other purposes). One can only expect in the future that technologies of sex selection will be further refined and that new and cheaper technologies will emerge on the market. In the absence of some system of regulation, nothing stands in the way of a continuation and expansion of substantial distortions in the sex ratio, at least in some parts of the world and among some communities in the United States.^xxiv

B. Preliminary Ethical Analysis

Previous public discussions of the ethics of sex selection, conducted largely in terms of "sex bias" and "reproductive freedom," have been oddly ambivalent. On the one hand, despite the widespread and growing practice of sex selection, it has attracted few overt defenders or partisans, at least in the United States. Almost no one argues openly in its favor, and those who do rarely offer up the single most important reason for its spread-the desire for sons over daughters (though, as we shall see, this taboo may be changing). To date, several special panels and advisory bodies in the United States have considered the ethics of sex selection.¹⁶ None of these has condoned the practice; all have raised serious ethical concerns. Yet, on the other hand, all have insisted that sex selection should not be made illegal and may at least in some instances be defensible. Even those who condemn the practice urge that there is nothing we can do about it without violating our most cherished principles of reproductive freedom and individual autonomy.

Typifying this approach, the one previous presidential commission to consider the topic gave several reasons to support its judgment that the use of amniocentesis and abortion for sex selection was "morally suspect." First, such a practice was "an expression of sex prejudice." Second, it was incompatible with the findings of developmental psychology that the parent-child relationship depends upon "the attitude of virtually unconditional acceptance." Third, sex selection treated the child "as an artifact and the reproductive process as a chance to design and produce human beings according to parental standards of excellence"-an attitude that the commission condemned.¹⁷ Yet despite these powerful objections, the commission did not see the matter in black-and-white terms either, and its policy recommendations were mild:

This is not to say that every decision to undergo amniocentesis solely for purposes of sex selection is subject to moral criticism. Nonetheless, widespread use of amniocentesis for sex selection would be a matter of serious moral concern. Therefore, the Commission concludes that although individual physicians are free to follow the dictates of conscience, public policy should discourage the use of amniocentesis for sex selection. The Commission recognizes, however, that a legal prohibition would probably be ineffective and, worse, offensive to important social values (because vigorous enforcement of any such statute might depend on coercive state inquiries into private motivations).¹⁸

One factor distorting the ethical discussions of sex selection in America is that it has become entangled-as has the debate over stem cells and human cloning-in the controversy over abortion. Certain widely accepted political and ethical principles, such as individual autonomy, equality, the right to choose, and "non-directiveness," are thought to be threatened by any thoroughgoing critique of sex selection. In the early years, when post-conception determination of sex followed by abortion was the only means of sex selection, it was widely argued by many feminist-oriented scholars, as well as other liberal thinkers, that any legal or policy actions taken against abortion for sex selection would put the abortion right itself at risk.

The practice of sex selection also throws other cherished principles into disarray. Since the end of World War II, genetic counselors have adhered to the ethical norm of "nondirectiveness." It was hoped that by this principle they would avoid the coercive eugenic policies of the past, from forced sterilization to genocide. Yet by mandating the moral neutrality of genetic counselors, nondirectiveness in fact makes it easier for individual couples to practice sex selection as a matter of personal choice. And here too the culture wars over abortion play a part. In one study it was found that genetic counselors were reluctant to recommend against sex selection since they considered it a "logical extension of parents' rights to control the number, timing, spacing, and quality of their offspring."¹⁹

But three new developments conspire to invite a serious reexamination of this matter. First, there is the growing cultural heterogeneity of American society, with a rise in subgroups with distinct preferences for males. Second, there are growing commercial prospects for these services. Although the sex-selection technologies were originally developed within the moral framework of medicine and were directed towards disease prevention, the commercial possibilities of these technologies are becoming increasingly evident. Sex-selection services are openly advertised on the Internet, and sex selection could in the future become a big business.^xxv Third, perhaps related to the second, resistance to this practice is weakening, including among those who are keepers and purveyors of the technologies.

In 1999, the American Society for Reproductive Medicine (ASRM) criticized the use of PGD and sperm sorting for sex selection, fearing that such practices might contribute to gender stereotyping and discrimination.²⁰ In 2001, however, the ASRM relaxed its opposition to sperm sorting if used for the purpose of "family balancing,"²¹ and, later that year, the chairman of ASRM's ethics committee appeared to endorse the use of PGD for the same purpose. When this produced considerable public controversy, in part based on concern over the destruction of embryos involved in PGD, the ASRM reaffirmed its position that PGD for sex selection should be discouraged, in deference to concerns about gender bias as well as about the moral status of the embryo. But the Society's recommendations are not enforced, and several of its members are openly offering sex selection to their clients.

In sum, although the practice of sex selection continues to grow, the American public debate over sex selection has never been aired in full. The new impetus to the growth of this practice, from multiculturalism to commercial interests, will make it difficult to slow its future spread. All the more reason to try now to evaluate its significance, beginning with the most common arguments for and against the practice.

*     *     *

There are a number of reasons given to support the practice of sex selection. The most common rationale today for sex selection is that it permits family balancing, enabling a couple to achieve its as-yet-unfulfilled wish to raise both sons and daughters. Many parents have had three or four girls (or boys) in a row, and really want a boy (or girl); effective sex selection would satisfy this wish without any risk of continued "failure." More generally, sex selection is defended on grounds that it could increase the happiness of the parents by enabling them to fulfill their desire for one or more sons or daughters. Sex selection is also supported because it may help to slow population growth (since many families continue to have children only to achieve a particular balance of boys and girls); because it may enable parents to fulfill religious or cultural expectations (since some cultures attach great importance to or impose special obligations on male heirs); and because it may make children feel more wanted and comfortable with their sex (since they will know that they were in fact chosen to be whichever sex they are).

In certain cultures, the desire of parents for sons is extremely powerful; in traditional Islam, for example, parents are expected to continue bearing children until they have at least one son. A strong preference for sons also appears prevalent in most (though not all) of the countries of Asia. Sex selection can therefore be defended on "multicultural grounds," as helping parents to achieve not merely individual preferences but also traditional and religious aims.

A common objection voiced against sex selection is that, in its most prevalent practice today, it almost always involves the abortion of (otherwise healthy) fetuses of the unwanted sex.^xxvi However, sex selection by IVF with PGD involves instead the selective transfer of embryos of the desired sex and the discarding of any embryos of the other sex; some people, for this reason, regard this approach as less morally objectionable than the one that requires abortion, while others see no moral difference. No such stigma attaches to the practice, still nascent, of sex selection by sperm sorting; whether used with artificial insemination or in conjunction with IVF, sperm sorting reduces the need to discard embryos of the unwanted sex. Should ongoing research eventually produce selective spermicides that would permit sex selection via natural intercourse, all such objections to the means would be much diminished or even disappear. We would be left to evaluate only the end itself.

The objection most often raised to sex selection, especially as it is practiced throughout the world today, is that it reflects and contributes to bias or discrimination against women. Sex selection has involved the abortion of female fetuses on a massive scale, or, in a few cases only, the selection of male embryos over female ones for implantation. As we have seen, sex ratios in some communities have been altered sharply in a very short period of time. Yet, criticism of this phenomenon has tended to be muted because of the connection between sex selection and abortion; those who support the right to an abortion have generally been reluctant to argue that abortion for the sake of sex selection should be restricted. The "pro-choice" idea of "every child a wanted child" establishes the rule in reproductive matters of the supremacy of parental "wants." Ironically, the "right to choose," which was and is defended in the name of equality for women, has in this way made permissible the disproportionate choice of aborting female fetuses. It is open to question whether the cause of equality has been well served by this development.

Paradoxically, the anti-female bias thought by critics to be implicit in sex selection might in fact redound to the advantage of women, at least regarding marriage: their relative scarcity could give them greater selectivity, choice, and control of partners. In certain Asian countries for example, where the ratio of boys to girls at birth has been severely skewed by sex selection, young men of marriageable age are already facing a severe shortage of young women to marry. Thus one might oppose sex selection as much for the actual harm it does to men as for the prejudice it expresses against women.

But sex selection is ethically troubling for reasons that go beyond both its potentially discriminatory use and the necessity, under current procedures, of destroying fetuses or embryos of the unwanted sex. One of the fundamental issues has to do with the limits of liberty.

C. The Limits of Liberty

As we noted earlier, few policy makers or opinion leaders argue openly in favor of sex selection. Rather, the assumption is made that our most cherished ideals of individual autonomy and the right to choose preclude an unambiguous condemnation of sex selection or public polices that might curtail it. Yet this assumption is questionable.

Our society, to be sure, deeply cherishes liberty and rightfully gives a wide berth to its exercise. But liberty is never without its limits. In the case of actions that are purely self-regarding-that is, actions that affect only ourselves-society tends to give the greatest protections to personal freedom. But as we move outward, away from purely self-regarding actions to those actions that affect others, our liberty is necessarily more liable to societal and governmental oversight and restraint. Sex selection clearly does not belong in the category of purely self-regarding action. The parents' actions (their choice of a boy or a girl) are directed not only toward themselves but also toward the child-to-be.

One might argue that, since each child must be either a girl or a boy, the parents' actions in selecting the sex do not constitute much of an intrusion on the prospective child's freedom and well-being. But the binary choice among highly natural and familiar types hardly makes the choice a trivial one. And having one's sex foreordained by another is different from having it determined by the lottery of sexual union. There is thus at least a prima facie case for suggesting that the power to foreordain or control the nature of one's child's sexual identity is not encompassed in the protected sphere of inviolable reproductive liberty. It is far from clear that either the moral or the legal right to procreate includes the right to choose the sex-or other traits-of one's children.

But it is not only that sex selection affects the individual child-to-be that puts it in a class of actions fit for oversight, regulation, and (perhaps) curtailment. Sex selection, if practiced widely, can also have powerful societal effects that reach far beyond individuals and their families to the nation as a whole. The dramatic alteration in sex ratios in such countries as South Korea and Cuba bear this out. Whether or not one views the preference of individuals for sons over daughters as rational, taken together these individual preferences could and do have serious society-wide effects. The males may have diminished chances of finding an acceptable mate, while the broader society may suffer from higher crime, greater social unrest, increased incidence of prostitution, etc.-social troubles closely associated with an abnormally high incidence of men, especially unmarried men.^xxvii One could argue that the choice of a male child is individually rational for parents, given the strong preference in certain cultures for males. But such individual choices may be socially costly-a case where individual parental eugenic choices do not yield a social optimum. Indeed, unrestricted sex selection offers a classic example of the Tragedy of the Commons, in which advantages sought by individuals are nullified, or worse, owing to the social costs of allowing them to everyone.²² In such cases, it is acceptable (and arguably necessary) for a liberal polity to place limits on individual liberty.

D. The Meaning of Sexuality and Procreation

The two aspects of sex control-it is control of sex, and it is a form of control of offspring-locate the deeper significance of this practice in two important human contexts: the meaning of sexuality, and the nature of procreation and family relations. A discussion of these matters shows why there is more at stake here than personal liberty.

The arguments previously advanced against sex selection, based on concerns regarding sexual bias, have been less than satisfactory. Some have argued, for example, that sex selection would reinforce gender stereotypes and threaten gender equality-presumably because it would manifest preferences for boys. Yet these critics do not specify what they mean by "gender stereotypes" and "gender equality." Sometimes it seems that they are worried that expressed preference for males would lead to a return to the world of 1950s-style stereotypes, with men and women playing distinct social roles. But it sometimes seems that they are also worried that sex selection would threaten a positive goal, a movement toward a more genuinely gender-neutral or socially androgynous society, one in which our socially constructed human identities would triumph over the mere biology of sexual difference. But in such a gender-indifferent society, it would presumably make no difference whether you are a girl or a boy, a woman or a man. And thus the choice of parents of a boy rather than a girl, or vice versa, would have no negative implications of gender stereotyping and would not threaten the equality of the sexes. The choice between a girl and a boy would be purely an aesthetic choice-as between pink and blue. And who could then object to letting parents choose? The very logic and language of gender equality, taken in its androgynous direction, would seem to soften opposition to sex selection. Further, there seems to be a contradiction between arguing that "sex should not count" in opposing the right of parents to choose boys rather than girls, while at the same time implying that "sex counts plenty" in approving sex selection for "family balancing." If, as the critics say, sex does not or should not count, why could they think a sexually balanced family humanly better than an unbalanced one? By selecting sex for any reason, does one not in fact acknowledge that it is very important?

As one of its arguments against the use of PGD for sex selection, the ASRM has suggested that it might "trivialize human reproduction by making it depend on the selection of nonessential features of offspring."²³ But if sexual identity is non-essential for many purposes (for example, at least in theory, in employment or other areas where the law forbids discrimination), for other purposes it is central to who and what we are. Humanity exists as a sexually differentiated species; it is constituted in part by the sexual difference. The reason is that our bodies are integral to our humanity. There is no generic or androgynous human "self" to which, as a kind of accidental addition, either a male or female body is then appended. Were that the case, sexual identity really would be "nonessential" or "inessential" to our self. It would not in any sense help to constitute a person's identity.

If, however, we do not accept that kind of dualism in which the real self simply is attached to and makes use of a (male or female) body, then we will have to take sexual identity seriously as given with our body. Every cell of the body and the entire body plan and form mark us as either male or female, and it is hard to imagine any more fundamental or essential characteristic of a person. It is surely odd, to say the least, to deny the importance of sexual identity in the very activity of initiating a life.

Seeing this, we can understand why it often seems so important to people that they have either a boy or a girl. Indeed, it would be surprising if people did not care about a difference so fundamental. But acknowledging this, we can also understand why we should be reluctant to see ourselves as people who may appropriately dictate such a crucial part of the identity of our child. Many prospective parents will say quite honestly that they don't care whether their baby is a boy or a girl; they'll be happy to have either. That attitude is desirable not because the sex of the child is a matter of indifference but because it counts for so much. Far too much to be seen as their responsibility to determine.

In a previous Council report, on human cloning,²⁴ we emphasized how cloning-to-produce-children alters the very nature and meaning of human procreation, implicitly turning it (at least in concept) into a form of manufacture and opening the door to a new eugenics. Sex selection raises related concerns.

The salient fact about human procreation in its natural context is that children are not made but begotten. By this we mean that children are the issue of our love, not the product of our wills. A man and a woman do not produce or choose a particular child, as they might buy a particular brand of soap; rather, they stand in relation to their child as recipients of a gift. Gifts and blessings we learn to accept as gratefully as we can; products of our wills we try to shape in accordance with our wants and desires. Procreation as traditionally understood invites acceptance, not reshaping or engineering. It encourages us to see that we do not own our children and that our children exist not simply for our fulfillment. Of course, parents seek to shape and nurture their children in a variety of ways; but being a parent also means being open to the unbidden and unelected in life.

Sex selection challenges this fundamental understanding of procreation and parenthood. When we select for sex we are, consciously or not, seeking to design our children according to our wants and desires. The choice is never merely innocent or indifferent, since a host of powerful expectations goes into the selection of a boy or a girl. In choosing one sex over the other, we are necessarily making a statement about what we expect of that child-even if it is nothing more than that the child should provide sexual balance in the family. As fathers, we may want a son to go fishing with; or as mothers, we may want a daughter to dress for the prom. The problem goes deeper than sexual stereotyping, however. For it could also be the case that we may want a daughter who will become president to show that women are the equal of men. But in making this kind of selection we have hardly escaped the problem, for the child's sexual identity would be determined by us in order to fulfill some particular desire of our own. If this were not the case then there would be no felt need to choose the sex of our child in the first place. And thus does it happen that in practicing sex selection our acceptance of our children becomes conditional-a stance that is fundamentally incompatible with the deeper meanings of procreation and parenthood.

The truth of this matter is paradoxically displayed by a small fact connected with current American practices of sex-selection. The assisted reproduction clinics that offer elective sex selection (through sperm sorting or PGD) require their clients to agree in advance that they will accept whatever child results, even if the child is not of the sought-for sex. The clinics are no doubt mainly protecting themselves against legal liability for a wrong result. Yet their need to insist on accepting an undesired "product" shows how the practice itself must make into a matter of compulsory agreement what the idea of parenthood should take for granted: that each child is ours to love and care for, from the start, unconditionally, and regardless of any special merit of theirs or special wishes of ours.

III. Improving Children's Behavior: Psychotropic Drugs

In addition to trying to enhance or control the inborn capacities of their children, parents can try to improve what their children do with the capacities they have. They can help them improve specific native gifts (musical, artistic, athletic, etc.) through practice or training. They can stimulate interest, develop tastes, and enlarge horizons through reading, travel, and exposure to culture. They can try to improve their moods, attitudes, and, of course, their behavior: how they act at home and school, how they respond to authority, how they comport themselves with family and friends. They can try to improve their ability and willingness to be considerate, show respect, pay attention, carry out assignments, accept responsibility, deal with stress and disappointment, and practice self-control. In these efforts, parents continue to use, as they always have, our time-honored methods for child rearing and education. But they may be acquiring extra help from biotechnology and the novel approaches to behavior modification that make use of drugs and devices that work directly on the brain.

Opportunities to modify behavior in children using psychotropic drugs are growing rapidly, and the young but expanding field of neuroscience promises vast increases in understanding the genetic and neurochemical contributions to behavior a