Assisted Reproductive Technologies and Women’s Choices: Autonomy, Harm and Gender Socialization

Abstract

In this paper, we examine and respond to two concerns associated with gender socialization and assisted reproductive technologies (ARTs). After providing an overview of feminist theorists’ work on gender socialization and women’s autonomy, we consider the concern according to which women’s choice to use ARTs to have children may be non-genuinely autonomous due to the influence of oppressive norms on this choice. We call this the autonomous-choice concern. In response to this concern, we argue that only subscribing to an overly demanding account of autonomy yields this conclusion and issue a caution about the implications of subscribing to such an account. Second, we consider the concern according to which the investment of public resources to make certain ARTs more widely accessible amounts to endorsing oppressive norms and may indirectly harm infertile and other women by making these norms more compelling. We call this the norm-legitimation concern. In response, we argue that when appraising the adverse downstream effects associated with legitimizing oppressive norms, it is necessary to consider both the costs of compliance and the costs of non-compliance with these norms, and that their consideration requires a differentiated approach. In closing, we consider objections to the arguments defended in the paper.

1. Introduction

Gender socialization is the process whereby individuals acquire gender-specific traits and internalize the beliefs, values, and norms that shape their lives as women and men within their social group. As feminist theorists have observed, while this process has an impact on everyone in society, it disproportionately disadvantages women by socializing them into subordinate roles (Mackenzie 2000). For instance, women are often encouraged to take up caring roles and be primarily responsible for child-caring and child-rearing. This comes at the expense of other life plans that they may want to pursue—as gender socialization constructs only some of them as appropriate for women. Moreover, it often results in holding disadvantageous positions within the family and, more generally, society. Consider, for instance, the choice to be a stay-at-home mother. This choice risks making a woman financially dependent on her partner, thus further entrenching her subordination. It additionally contributes to legitimizing norms and expectations according to which women are solely or at least primarily responsible for care and domestic labor. ¹ This partially explains feminist theorists’ skepticism towards some of the choices that women make under oppressive conditions. Often, women’s choices are taken to be non-genuinely autonomous, for gender socialization impinges on their ability to choose what may be best for them. Depending on the degree of disadvantage resulting from these choices and the account of autonomy that they subscribe to, feminist theorists advocate for various forms of intervention through policies or practices. ² These interventions are aimed at safeguarding women’s capacity to act autonomously; mitigating the disadvantage that may arise from their choices; and preventing downstream effects that their choices may generate for them as well as for other women.

In this paper, we discuss the phenomenon of gender socialization in relation to women’s ³ choice to use assisted reproductive technologies (ARTs) to have children. ⁴ First, we discuss the concern according to which women’s choice to use ARTs to have children may be non-genuinely autonomous due to the influence of oppressive norms on this choice. We call this the autonomous-choice concern. Second, we canvass the concern according to which the investment of public resources to make certain ARTs more widely available may indirectly harm infertile and other women by legitimizing oppressive norms. We call this the norm-legitimation concern. In response to the autonomous-choice concern, we argue that only subscribing to an overly demanding account of autonomy can yield the conclusion that women’s choice to use ARTs to have children is non-genuinely autonomous, and issue a caution about the implications of subscribing to such an account. In response to the norm-legitimation concern, we argue that, when appraising the adverse downstream effects associated with legitimating oppressive norms, it is necessary to consider the costs of both compliance and non-compliance with these norms, and that their consideration requires a differentiated approach.

There are many forms of ARTs. ⁵ In this paper, we use this term collectively in reference to medical interventions aimed at addressing people’s fertility issues. ⁶ Examples of these include in-vitro fertilization (IVF), intra-uterine insemination (IUI), gestational surrogacy, gamete and embryo cryopreservation, mitochondrial donation, and uterus transplantation (UTx). Since the early days of IVF, feminist theorists have criticized this and other reproductive technologies by drawing attention to the effects on women of liberalizing and supporting their provision. ⁷ More recently, the main target of feminist critiques of ARTs has been the investment of public resources to make specific ARTs, such as UTx, more widely available. However, while the critiques that we examine in this paper are aimed specifically at investing public resources into particularly risky and invasive ARTs, we believe that the reasons that authors supply to motivate them apply mutatis mutandis to investing public resources into most ARTs, and that IVF and other less fanciful technologies cannot be successfully insulated from these critiques. For instance, while IVF is not as risky or invasive as UTx, a woman going through an IVF cycle will be required to take medications with substantial side effects, undergo the procedure to retrieve her eggs while being fully sedated, and adjust her lifestyle and diet considerably. ⁸ While UTx is riskier and more invasive than IVF, it is unclear why only the former and not the latter technology raises the concerns that have been outlined in recent feminist literature on ARTs.

The concerns we canvass in this paper create a puzzle for those who defend the legitimacy of the publicly funded provision of ARTs. One of the key reasons expressed in its support is that these technologies are thought to promote women’s (and people’s) reproductive autonomy by expanding the range of options available to them and enabling them to have control over the realization of their reproductive aspirations and desires. However, if the choice to use ARTs to have children is non-genuinely autonomous, defenses of ARTs grounded in the value of promoting reproductive autonomy will lose their justificatory power. In addition to this, if fulfilling these desires and aspirations with ARTs harms infertile and other women by legitimizing oppressive norms, interventions aimed at restricting the investment of public resources in ARTs seem to be, at least prima facie, justified. It is for these reasons that it is important to engage with these concerns: they provide grounds for questioning the value of using ARTs to fulfill women’s reproductive aspirations and the legitimacy of public support for these technologies.

The paper is structured as follows: in section 2, we briefly sketch the debate on gender socialization and women’s autonomy. In section 3, we consider and respond to the concern that the choice to use ARTs to have children may be non-genuinely autonomous due to the influence of oppressive gender socialization. In section 4, we examine and respond to the second concern, namely that investing resources in the public provision of certain ARTs risks legitimating oppressive norms whose entrenchment indirectly harms women. ⁹ In closing, we consider and respond to three objections to our arguments.

2. Gender socialization and women’s autonomy

Gender socialization is particularly onerous for women as it often socializes them into subordinated roles while leaving broader prospects to men (Piper 2014). For instance, owing to their capacity to bear children, biologically essentialist perspectives on gender have contributed to bolstering the view that women’s “true nature” predisposes them to take up caring roles and responsibilities. As girls grow up, they are typically encouraged to be more attentive to other people’s needs and cater to them, and adult women are ordinarily expected to bear the lion’s share of care and domestic labor. This generates significant opportunity costs, such as foregoing alternative life projects they might want to pursue or having access to a more restricted range of career choices and opportunities. Due to the unequal distribution of child-caring and child-rearing responsibilities, women are more likely to take up positions that are less remunerated and that are compatible with the so-called “second shift.” However, privileging unpaid, reproductive labor rather than waged, productive labor generally results in financial dependency, which is both disadvantageous and difficult to reverse (Chambers 2007). ¹⁰

Women’s putative complicity in their subordination motivates what Natalie Stoljar refers to as the “feminist intuition,” according to which “preferences influenced by oppressive norms of femininity cannot be autonomous” (emphasis ours, Stoljar 2000, 95). Consequently, many have employed concepts such as “false consciousness” (Bartky, 1998), “internalized oppression” (Cudd 2006), or “adaptive preferences” (Nussbaum 2001) to capture the normative insufficiency of individuals’ choices vis-à-vis their embeddedness in a social context that systematically disadvantages them. In the context of this paper, given the opportunity costs that motherhood often generates for women, prevailing norms on gender and parenting, and that using ARTs to have children is physically and psychologically burdensome, feminist theorists have questioned the degree to which the choice to use ARTs to have children should be considered genuinely autonomous. ¹¹

Feminist relational accounts of autonomy have played a pivotal role in reconfiguring the debate on autonomy, emphasizing the detrimental effects of gender socialization. ¹² However, while proponents of these accounts broadly agree that gender socialization disadvantages women and contributes to bolstering their subordination, the necessary conditions for a choice to be considered autonomous remain disputed. These conditions vary significantly between accounts of autonomy that consider the influence of gender socialization, and result in heterogeneous attitudes toward paternalistic policies or practices. For instance, more substantive and demanding criteria for autonomy are more likely to legitimize various forms of interference with people’s choices. ¹³ Formal and more neutral criteria of autonomy lead to hands-off approaches to interference and respect for people’s choices as long as they result from appropriately conducted deliberative processes. ¹⁴ Therefore, from a shared skepticism towards the degree to which choices can be considered genuinely autonomous, feminist theorists of autonomy land on very different positions, which vary depending on the specifics of the account of autonomy that they subscribe to. ¹⁵

Another aspect that intersects with this discussion is the responsibility that women have towards other women if they comply with oppressive gender norms, ¹⁶ thus indirectly legitimizing them. For instance, Clare Chambers (2007) discusses women’s role in perpetuating gendered standards of appearance when they opt for cosmetic surgery. Chambers acknowledges that beautification brings some rewards. Getting breast implants may be in a woman’s best interests within a social context that primarily values her for her ability to conform to gendered standards of appearance. Yet Chambers additionally argues that this contributes to making cosmetic surgery more compelling for her as well as other women, and legitimizing the oppressive norms that dictate that women appear in gender-conforming ways. In pronatalist societies characterized by an unequal distribution of the responsibilities associated with care and domestic labor, while it may be rational for a woman to pursue motherhood (with or without ARTs), her choice can also disadvantage her along various axes. Moreover, it can reinforce existing norms concerning women’s appropriate role in society and negatively affect her as well as other women.

Having sketched the debate on gender socialization and women’s autonomy, we turn to the concern that the choice of using ARTs to have children may be non-genuinely autonomous. Before doing so, we briefly canvass social constructivist analyses of motherhood, infertility, and reproductive aspirations as they lend support to the view that the social environment in which women form reproductive aspirations influences them—a central aspect of feminist critiques of ARTs and their publicly funded provision.

3. Social construction, oppressive norms, and reproductive aspirations: Are women’s choices to use ARTs genuinely autonomous?

There are many ways in which the claim that an object is socially constructed can be interpreted. ¹⁷ For instance, an object would be socially constructed if it was caused to exist or produced by social factors, such as conventions, institutions, and practices. Alternatively, it would be socially constructed if “particular features of social reality” were necessary to instantiate or constitute it (Griffith 2018, 395). Importantly, social constructivist analyses seek to emphasize the degree to which objects and phenomena considered natural, stable, or inevitable are, in fact, caused to exist or constituted by social norms, practices, and institutions. As different norms, practices, and institutions would have brought about different objects and phenomena, these are revealed to be contingent and can thus be “transformed by means of social action” (Diaz-Leon 2015, 1139). For instance, Ian Hacking (1999) argues that social constructivist analyses of motherhood have contributed to challenging essentialist views concerning women’s biological capacities and their “natural” role in reproduction by reminding us that “motherhood and its meanings are not fixed and inevitable, the consequence of child-bearing and rearing,” but, rather, “the product of historical events, social forces, and ideology” (Hacking 1999, 2). In this sense, there are reasons to be skeptical of seeing reproductive aspirations or, more specifically, the choice of having children (with or without ARTs) as necessarily following from having certain biological capacities, such as the capacity to bear and give birth to children. ¹⁸

In what can be considered a similarly spirited effort to reveal the contingency of certain phenomena, others have focused on the social construction of infertility, which they see as a “social condition that has been recast as a disease” (emphasis ours, Becker and Nachtigall 1992, 457) or a “socially constructed process whereby individuals come to regard their inability to have children as a problem” (Greil et al. 2011, 3). On this view, infertility is not a natural phenomenon, occurring due to a certain biological disfunction or other biological “facts about” a person, but something that is produced through and constituted by social norms concerning reproduction, heterosexuality, gender, and parenting. ¹⁹ Moreover, it is something that has been medicalized, i.e., it has been redescribed as a medical problem or pathology (Becker and Nachtigall 1992). As a result, the management of infertility has been left to medical professionals, and medical interventions, such as ARTs, have been privileged over “nonmedical solutions” (Kukla 2019, 4426) or “alternatives,” such as “self-definition as voluntarily childfree, adoption, fosterage, or changing partners” (Greil et al. 2011, 3).

Others draw attention to the social environment in which women form reproductive aspirations and to the influence of this environment on them. ²⁰ For instance, Angel Petropanagos (2017) argues that women’s choice to use ARTs to have children is influenced by two biases: “pronatalism” and “geneticism,” which indicate a bias in favor of gestation and genetic relatedness respectively. According to Petropanagos, these two biases can “interfere with autonomy by unduly influencing an individual’s reproductive values, preferences, and desires or by compromising her capacity for critical reflection or action” (emphasis ours, Petropanagos 2017, 133). Mianna Lotz (2021) contends that the publicly funded provision of UTx amounts to “legitimating and even endorsing certain problematic sociocultural attitudes and norms relating to reproduction” (Lotz 2021, 666). Following Lotz, these norms are ‘pronatalist, essentialist, and geneticist’ (hence Lotz’s acronym “PEG norms”). Pronatalist norms construct motherhood as a foundational aspect of the good life and essential to women’s flourishing, something that they ought to pursue to have a fulfilling life. (Lotz 2021) Essentialist norms identify certain, usually biological, properties as necessary to being classified as a woman, such as having a uterus, being able to procreate, and being nurturing. Lastly, geneticist norms ascribe superior value to genetic and biological ties over other kinds of ties between parents and children, thus using ARTs to have children becomes preferable to adoption, fostering, or pursuing other, non-traditional, family arrangements (Lotz 2021). Following these views, the social environment in which women choose to use ARTs to have children is biased towards certain kinds of life projects (such as motherhood) and certain kinds of family ties (such as biological and genetic ties). ²¹

Charlotte Witt (2014) refers to this social environment as one championing a “bionormative conception of the family,” i.e., what she takes to be the pervasive idea that “families that are formed via biological reproduction (in which there is a genetic relationship between parents and children) are, for that reason, superior to families formed in other ways” (Witt 2014, 50). Like others who lament the bias towards genetic/biological vis-à-vis other family ties, ²² Witt contends that, even if this conception of the family is prevalent in certain societies, there is nothing a priori special in parent–child relationships that entail biological or genetic links.

While we cannot extensively review the evidence that critics of the “bionormative conception of the family” buttress in support of their claims, we find them compelling and are inclined to grant the empirical claim that biological and genetic ties are often ascribed a higher value than other forms of family formations, and that “pronatalism” and “geneticism” play a role in influencing women’s choice to use ARTs to have children. However, first, despite the merits of social constructivist analyses of infertility, reproductive choices, and motherhood, the evidence concerning reproductive aspirations, motivations, and intentions is complex. A heterogeneous set of variables has been employed to investigate psychological, social, and biological mechanisms influencing reproductive choices and decision-making, and to single out causative patterns, which casts doubts on purely socio-cultural explanations of reproductive aspirations and the desire to have children. ²³ A plausible explanation of the force that social norms exert on women’s reproductive aspirations is that these norms align with rather than counter evolutionary-endowed instincts and drives. The “bioproduct view” explains the desire to have children as a sort of side effect (or “epiphenomenon”) of basic drives, sex drive and nurturance instinct, ²⁴ for which there is robust empirical support across species. As a result, one’s social environment seems to find low resistance due to biological and evolutionary-endowed instincts that this environment aligns with rather than counters. Given these considerations, it seems that, while social constructivist analyses of infertility, reproductive choices, and motherhood rightly emphasize the influence that the social environment exerts on reproductive aspirations, their etiology is not straightforwardly attributable to socio-cultural explanations. Second, as Sally Haslanger (2012) argues, one of the “steps” of social critique “involves describing the social practice in question in a way that highlights those features that are relevant to normative evaluation” (Haslanger 2012, 16). Showing that something is partially or entirely socially constructed reveals certain aspects of a concept, a practice, or a norm to be contingent. This can inform their normative appraisals, and the permissibility of interventions aimed at transforming them. However, showing that a phenomenon could have been otherwise does not guarantee that the next interpretation is one we should favor or one that makes us better off. Moreover, even if the choice to use ARTs to have children is influenced by one’s social environment—a view that we accept with qualifications—this observation alone does not settle either whether it can be reasonably considered an autonomous choice or normative questions concerning the value of the publicly funded provision of ARTs.

3.1 The choice to use ARTs to have children and conditions for autonomy

We now turn to the autonomous-choice concern, i.e., the concern that women’s choice to use ARTs may be non-genuinely autonomous due to the influence of oppressive social norms on this choice. There are two ways in which this concern can be cashed out. According to the radical version, due to these norms, women’s choice to use ARTs cannot be genuinely autonomous. This version of the concern does not distinguish between degrees of influence on this choice. Thus, it deviates from what we take to be the common-sense view on autonomy-undermining influences, where the effects of such influences will be different for different people, depending on their social positions and other factors. For instance, Robyn Rowland (1992) argues that infertile women choose to use ARTs to have children in “a social context, constrained and shaped by the forces of economics, social ideology, personal psychology, and established power structures” and concludes that “this is not choice as feminists would construct it” (Rowland 1992, 279). The more moderate version of the concern acknowledges the influence of oppressive social environments on women’s choices to use ARTs but suggests that the choices of only some women might be non-genuinely autonomous. For instance, Petropanagos (2017) argues that:

Although not all women who choose gestational or genetic motherhood will have compromised autonomy, the social context is such that choosing against these norms or “failing” to achieve biological motherhood is potentially threatening or damaging to some women’s autonomy. As such, some women may be compelled to choose ART that they might not otherwise have chosen. (Petropanagos 2017, 133–34)

We start with the moderate version of the autonomous-choice concern, and argue that it is insufficient to justify withholding public funding of ARTs. Although it expresses relevant considerations about women’s reproductive aspirations in gender-oppressive contexts, it is pertinent to all medical interventions where some agents might be influenced by social norms. Typically, the recommended approach in these circumstances is a case-by-case evaluation of the agent’s autonomy rather than the withholding of public funding for these interventions. While we acknowledge that measures to ascertain the degree to which the agent’s choices are autonomous are not always entirely reliable, ²⁵ they avoid sweeping generalizations about her autonomy, which would negatively impact her as well as those whose choices may be less or not at all influenced by these norms. Concerns about an agent’s potential lack of autonomy in similar circumstances should be assessed carefully and individually, rather than broadly assumed. It is unclear why, when it comes to ARTs, we should opt for a different approach—one that would affect all women who wish to use ARTs to have children—even if only the choices of some of them are at risk of being influenced by an oppressive social context.

Our response to the radical version of the autonomous-choice concern is more articulated. First, we argue that only an overly demanding conceptualization of autonomy can lead to the conclusion that women’s choice to use ARTs is non-genuinely autonomous and justify withholding public funding for these technologies. Second, we contend that such a conception of autonomy risks undermining women’s agency and could lead to unwarranted interferences with women’s reproductive choices and aspirations.

Theorists who discuss women’s autonomy in relation to their choice to use ARTs to have children do not make explicit the conception of autonomy that informs their assessments. However, their critiques of ARTs or the publicly funded provision of these technologies often imply that, for an agent to be classified as autonomous, certain external social conditions must meet specific evaluative standards, such as the absence of oppressive gender norms. Therefore, their argument implicitly requires substantive and externalist criteria for autonomy. In the feminist literature on relational autonomy, constitutively relational accounts are usually associated with these two characteristics. They incorporate the external relational factor within the definition of autonomy, meaning that certain social conditions are part of the “defining conditions” of autonomy rather than merely contributory factors (Christman 2009). More specifically, these accounts make the presence of ideal social conditions necessary for autonomy—an approach that has the advantage of recognizing injustice even when those who are harmed by it appear to endorse norms that contribute to their subordination or, at least, do not protest against them. However, a competing concern arises when women’s choices are downgraded as non-genuinely autonomous due to external circumstances or the oppressive interpersonal relations that they sustain. Questioning these choices may come at the expense of belittling women’s agency, portraying them as “dupes of patriarchy” (Narayan 2002, 420) in need of salvation, and may lead to unwarranted paternalistic intervention (Khader 2011). We refer to this as the problem of victimization.

In line with existing critiques of these accounts, ²⁶ we contend that due to their substantiveness and externalism constitutively relational accounts of autonomy are particularly vulnerable to the problem of victimization. The primary concern is that defining ideal external social conditions as constitutive of autonomy renders the oppressed ipso facto less autonomous than the dominant (Holroyd 2009). A corollary of this view is that the oppressed have a lesser entitlement than the dominant to make decisions about their own lives. As a result, this process risks not only undermining oppressed people’s choices but also justifying additional forms of oppression on them. An example of this is “epistemic oppression,” which occurs when oppressed individuals are considered “bad knowers or morally defective” (Khader 2020, 504), licensing greater scrutiny of their choices. Moreover, it paradoxically presents them as less eligible for participation in democratic decision-making processes, as their perspectives and value orientations are questioned (Christman 2004). Implicitly relying on a similarly substantive and externalist conception of autonomy to argue against investing public resources to make ARTs more widely available has the same undesirable implications in matters of reproduction. It risks undermining women’s agency by not recognizing them as reliable judges of their own good and might inadvertently justify unwarranted paternalistic interventions in their reproductive choices, based solely on the existence and influence of oppressive norms.

Proponents of constitutively relational accounts contest the victimization charge. They argue that denying women’s autonomy in adverse external conditions does not equate to denying their agency, as the two concepts are not co-extensive. This rejoinder relies on the distinction that these theorists draw between “global” and “local” autonomy. Global autonomy corresponds to a temporally extended condition in which agents have “de facto power and authority over choices and actions significant to the direction of their lives” (Oshana 2006, 2), while local autonomy refers to the ability to make particular choices at particular times. If this distinction between global and local autonomy is taken into account in the substantive evaluation of women’s choices, the choice to use ARTs to have children may be deemed only locally non-genuinely autonomous, but the agent would nonetheless retain her global autonomy and other components of her agency.

However, this way of reframing the issue is self-defeating in relation to the choice to use ARTs to have children, since these choices plausibly fall within the choices that are significant to the direction of one’s life, thus qualifying as part of global autonomy. If constitutively relational theorists wish to consider women who choose to have children with ARTs as only locally non-autonomous, they must downgrade reproductive choices below the class of those significant for one’s life direction. This would leave us with a substantive theory of autonomy that excludes reproductive choices from the domains of life needing safeguarding. But a similar move might inadvertently justify interferences and limitations not only on the use of ARTs, such as restricting public funding for research and provision, but also in the broader domain of reproductive choices. On this revised account, women would be deemed to retain their global autonomy, but at the cost of being classified as locally non-autonomous with respect to reproduction in general, which seems to be an unpalatable conclusion, especially from a feminist point of view. ²⁷ Alternatively, the constitutively relational theorist might consider women who choose to use these technologies as impaired in their global autonomy, yet this would reenact the previous worries about victimization. More specifically, it would lead to a more refined version of the problem of victimization, resulting from the hierarchical relation that constitutively relational theory usually draws between global and local autonomy. For instance, Oshana (2006) argues that if an agent is impaired in her global autonomy, it is sometimes permissible to override her locally autonomous choices to contribute to its development. Therefore, the distinction between local and global autonomy restores the concerns of detractors of constitutively relational accounts, rather than dispelling them. Moreover, even if we set aside the problem of victimization and the associated risk of unwarranted interference with women’s choices, there are reasons to be cautious about policies that might end up frustrating individual aspirations, particularly those related to the domain of reproduction. Such interventions can be profoundly disruptive to an individual’s life and negatively impact her well-being (Brock 2005). We explore this concern further in the next section, where we discuss the problem of double jeopardy.

4. ARTs, norms legitimation, and indirect harms

While some theorists acknowledge that a woman’s choice to use ARTs to have children can be considered genuinely autonomous within a specific social context, they contend that the state should not invest resources in particularly risky or invasive ARTs—as state support for these technologies amounts to publicly endorsing oppressive norms. For instance, Lotz (2021) argues that—to the extent that pronatalist, essentialist, and geneticist (PEG) norms are considered an instance of oppressive gender socialization—their promotion via the investment of public resources in expensive, invasive, and risky ARTs, such as UTx, should be considered morally suspect. At the root of this claim is the view that there seems to be a contradiction between the commitment to resist oppressive norms, on the one hand, and public support for the very thing that is valorized by these norms, on the other. We refer to this as the norm-legitimation concern.

In response, we argue that failing to publicly fund ARTs ends up restricting access to these technologies and that this cannot be justified by an appeal to general duties to counter oppressive gender socialization. More specifically, our contention is that this entails the withholding of a highly valuable flourishing opportunity from a group whose opportunities to flourish are already limited. In what follows, we articulate and defend this view. We briefly examine the reasons that feminist theorists advance to criticize PEG norms and their influence on women’s lives and well-being. Then, we consider the costs of non-compliance to these norms and argue that we should not impose these costs on the oppressed as part of a socially reformist strategy.

According to those who criticize the public provision of particularly invasive or risky ARTs, PEG norms are problematic for two interrelated reasons. First, as with gender socialization more generally, they socialize women into a limited set of flourishing opportunities at the expense of other life projects that they may want to pursue. Second, they present these opportunities as the most valuable for women, despite the costs of both compliance and non-compliance attached to them. To illustrate this, imagine a woman, Sally, who has a range of aspirations and who, given her natural endowments, upbringing, social status, and preferences could choose to pursue only some of them. Given internal and external constraints, Sally could be a professional tennis player, a teacher, or a digital nomad. However, gender socialization would construct some of these aspirations as more appealing than others, and, importantly, the aspiration to become a mother would be constructed as the aspiration she should, all things considered, pursue. ²⁸

We begin with the costs of compliance. These costs result from taking up a certain (socially prescribed) role R in a (non-ideal) context C and they are exemplified by the disadvantageous positions that women who take up the R of mothers in C will suffer. As we have shown, women may shoulder the bulk of the responsibilities associated with domestic and care labor and, as a result, have access to career options that are unfulfilling or less remunerated. The costs of non-compliance consist of a diminished capacity to flourish within a given socio-political environment. Recall Sally. She may, for example, opt to forego motherhood or forego having a second child to pursue a career as a professional tennis player or another profession that is likely to be difficult to square with motherhood or with the kind of mothering that is often expected and, sometimes, required from women. ²⁹ Therefore, for her, the costs of non-compliance may result from being stigmatized for prioritizing a certain less (socially constructed as) meaningful life project (such as playing tennis) over others (such as child-bearing and child-rearing). ³⁰ But they may also result from, for example, being excluded from the social relations that taking up the role of the parent (and, specifically, the mother) can give rise to. ³¹ This is to say that Sally need not be perceived as self-interested for non-compliance to negatively influence her capacity to flourish in a given social environment.

Let us turn back to infertile women who wish to have children and choose to use ARTs to achieve this goal. Infertility can be considered a form of non-voluntary non-compliance. Yet the costs attached to it may exceed those of voluntary non-compliance. Both women who choose to forego motherhood and women for whom childlessness is not a choice risk being stigmatized as a result of their lack of compliance with PEG norms. But, unlike voluntarily childless people, involuntarily childless people cannot comply with these norms. In addition to being stigmatized as a result of their childlessness, they often end up blaming themselves for it—even though fertility is something in large part beyond one’s control. Carolyn McLeod and Julie Ponesse (2008) explain that this is a feature of women’s experience of living in pronatalist societies precisely because of gender socialization: they internalize the belief that their prominent social role qua women is to reproduce.

Contra critics of the publicly funded provision of certain ARTs, we argue that the costs of both compliance and non-compliance demand a differentiated approach to resisting oppressive gender socialization. With respect to the costs of compliance, it is necessary to alleviate the costs resulting from taking up certain, socially prescribed, Rs in non-ideal Cs. This can be pursued in different ways, which would chiefly aim to contest and resist the norms that uphold unequal social arrangements. An example of this would be to advocate for and implement strategies aimed at a fairer distribution of caring responsibilities between parents or at making certain career opportunities more accessible to primary carers. However, with respect to the costs of non-compliance, a different approach is required, especially when non-compliance is non-voluntary, as in the case of infertility. Therefore, the alleged tension between the feminist commitment to undermine oppressive norms and the intention to provide publicly funded ARTs to realize the very preferences that these norms influence derives from failing to acknowledge the necessity of this differentiated approach.

The norm-legitimation concern exclusively focuses on the costs attached to compliance, without taking into account the effects on women who do not comply or, as in the case of infertility, cannot comply. When these costs are considered in the appraisal of the different strategies aimed at bringing about desirable social reforms, there is no contradiction between the two commitments: both are ultimately aimed at catering to women’s opportunities to flourish, only in two different sets of circumstances. Undermining oppressive norms is a more forward-looking practice of resistance, aimed at protecting the interests of future generations of women who will be harmed by these norms in the two interrelated ways that we have described above. Publicly funding ARTs instead caters to the interests of present women who are being harmed by oppressive norms. ³² As a result, restricting access to ARTs and, thus, withholding from them the possibility of having children with ARTs, disadvantages them even further. It removes what they consider a valuable flourishing opportunity from the already limited set of opportunities that their society makes available to them. For this reason, we contend that a similar resolution amounts to a form of double jeopardy. It places on infertile women’s shoulders the unwarranted burden of resisting social norms for future generations, on top of experiencing the harms deriving from oppressive gender socialization. ³³ Moreover, considering that those troubled by norm-legitimation and indirect harms do not object to ARTs per se, but to its provision via state funding, there is an additional class-based consideration associated with our double jeopardy argument: the lack of publicly funded provision primarily disadvantages less well-off women, while leaving the option of using ARTs open to those who can afford to do so. As a result, the burden of defying oppressive norms would fall on those who are already facing other forms of disadvantage, such as disadvantage resulting from their economic circumstances.

5. Objections

In this last section, we consider and respond to three objections to the arguments that we have defended in this paper. To recap: first, we argued that only subscribing to an overly demanding and idealized account of autonomy yields the conclusion that women’s choice to use ARTs to have children may be non-genuinely autonomous. However, this undermines women’s agency and licences paternalistic interventions in a domain that should be insulated as much as possible from these interventions: the domain of reproduction. For this reason, we believe that the publicly funded provision of certain ARTs cannot be objected to on this ground. Second, we have argued that—in considering the adverse downstream effects that the investment of public resources in ARTs has in terms of legitimizing oppressive norms—it is necessary to consider both the costs of compliance and the costs of non-compliance, and that their consideration requires a differentiated approach.

A first objection to our arguments could weaken the case for realizing expressed preferences that disadvantage women. Considering the costs associated with ARTs and motherhood more broadly, even if infertile women autonomously choose to use ARTs to have children, one could question whether doing so is in their best interests. This objection is sometimes raised against practices of beautification, such as cosmetic surgery, in feminist literature on oppressive gender and beauty norms. On this view, beautifying is not considered to be in a woman’s “genuine” interest but rather something that benefits men in the context of male sexual desire (Bartky 1998). On our reading of Lotz (2021), she might be raising a similar objection, which considers that there might be substantive differences between interests and preferences. Lotz hints that if the desire to have children with burdensome and risky ARTs is a desire that infertile women have rather than something in their best interests, we should not prioritize the investment of public resources in UTx, the technology whose publicly funded provision is Lotz’s main critical target. While Lotz does not pursue this line of argument further, we respond to it to forestall objections of this kind.

First, we believe that, in this and other domains, we should start from what we take to be the common-sense assumption that autonomous agents are ordinarily the best judges of their own interests (Dworkin 1988). We have argued for the compatibility between the influence of gender socialization on women’s choices and the genuinely autonomous nature of these choices in section 3.1. A similarly lengthy discussion about rationality and interests is beyond the scope of this paper. However, it is worth noting that questioning an agent’s preferences based on what we, as external judges, might think is “really” in her best interests can constitute, following Seana V. Shiffrin, “a failure to respect either the capacity of the agent to judge, the capacity of the agent to act, or the propriety of the agent’s exerting control over a sphere that is legitimately her domain” (Shiffrin 2000, 220). Second, given infertile women’s social environment, ARTs are in their best interests. This argument aligns with Heather Widdows’s objection to the view that cosmetic surgery is not genuinely in a woman’s best interest but rather favors the male gaze. Widdows (2018) contends that in contexts where lookism is rampant, engaging in processes of beautification, including cosmetic surgery, is in one’s best interests as it leads to concrete advantages. ³⁴ Similarly, in contexts where motherhood is valued and bionormativity is the standard, using ARTs to have children may indeed be in a woman’s best interests.

A second objection could be directed at our double jeopardy argument and note that, in realizing infertile women’s reproductive aspirations through the publicly funded provision of certain ARTs we are prioritizing these women’s interests over those of future fertile and infertile women. Arguably, the latter groups would benefit from the weakening of oppressive norms that non-compliance with these norms can eventually bring about. But, first, this response prioritizes future women’s interests over those of current women, something that, as we have argued above, should not be accepted without further justification. Second, to succeed, this objection also needs to proceed under the assumption that, in the absence of the publicly funded provision of ARTs, future infertile women would not experience, or experience to a much lesser extent, the costs of non-compliance resulting from their inability to have children. This, however, seems implausible as it relies on an overly simplistic understanding of gender socialization and the possibility of social change. In a counterfactual world without any publicly funded provision of ARTs, it seems unlikely that oppressive gender and parenting norms would disappear, for the content of these norms tracks flourishing opportunities that many consider valuable. ³⁵

Lastly, one could argue that the state should not engage in the promotion of oppressive norms, regardless of the costs of non-voluntary non-compliance. To illustrate this point, Lotz (2021) draws a parallel with skin-whitening treatments. She argues that, even if skin-whitening treatments could reduce the degree to which people of color are harmed by racist attitudes, the state should not invest resources to make them more accessible via the publicly funded provision of these treatments—as doing so would undermine efforts aimed at eradicating racist attitudes by legitimating the very same norms that sustain them. While Lotz does not establish that all the norms associated with skin-whitening are racist, we proceed under the assumption that many or most of them are. If this was the case, there might be good reasons to refrain from publicly funding skin-whitening treatments. Our response to this objection is twofold.

First, we contend that there is a difference between racist norms and norms that influence the choice to use ARTs, even risky and invasive ARTs, to have children. Racist norms are a paradigm example of norms that are oppressive and disadvantage certain groups, and that do not track any positive or, at the very least, neutral attitude towards these groups. They are intrinsically morally contentious and endorsing them negatively impacts groups that are already disadvantaged along various axes. For these reasons, publicly funding skin-whitening treatments and lending support to these norms is unlikely to ever be morally permissible. We thus concur with Lotz that the state should not engage in the promotion of racist norms. Norms that are “pronatalist and geneticist” are different in this respect. We take pronatalism to be objectionable insofar as it reduces available opportunities for flourishing, causally contributes to stigmatizing those who do not conform, and disadvantages both those who conform and those who do not conform. Geneticism, as described by critics, is problematic insofar as it stigmatizes families that do not conform to dominant views of family relations, also referred to as “the bionormative conception of the family.” However, not all the norms that influence the choice of using ARTs to have children are pronatalist and geneticist, and not all attitudes that these norms track are problematic. The choice to use ARTs to have children rather than, say, remain childless or adopt, when it is autonomous, and when it tracks attitudes that are non-discriminatory or non-essentialist, is not intrinsically morally contentious. ³⁶ Moreover, at least some uses of ARTs are inherently resistant to dominant gender and parenting norms, more broadly conceived. For instance, same-sex couples, transgender parents, or those who fall outside of the gender binary resist heteronomativity and cissexism. While these groups are “compliant” insofar as they opt into genetic and biological parenthood, they do not seem to raise similar concerns related to reinforcing oppressive norms. On the contrary, they may contribute to undermining at the very least the essentialist components of these norms.

Second, we argue that, even when there is nothing redeemable about certain social norms, we should adopt a pluralist approach to resisting their entrenchment. The risk of not doing so is to place additional burdens on the groups that are targeted by these norms and vulnerable to the harms that they bring about. The case of hymenoplasty represents an interesting example in this respect. This controversial practice consists of a few temporary stitches on the remnants of the hymen, after the rupture of the membrane that is widely, yet falsely, believed to typically occur the first time that women have intercourse (Saharso 2022). Women living in communities that consider virginity a requirement for marriage may ask for hymenoplasty to hide that they have had premarital sex. The practice is thus deeply contested for perpetuating false myths about virginity and social norms that restrict women’s sexual autonomy (Chambers 2004). Nonetheless, the costs of non-compliance faced by women who do not have access to the practice may actually trump, in specific circumstances, these principled considerations. ³⁷ For instance, a ban on hymenoplasty not only prevents women who have had premarital sex from marrying within their community but, in more treacherous circumstances, also makes them more vulnerable to possible violent repercussions from their community (Cesarano 2023). Therefore, in resisting the entrenchment of oppressive norms, we need to consider carefully what approach should be adopted when moving from principles to practices and interventions. Sometimes, more direct approaches, such as bans, may backfire and disadvantage the groups that these policies intend to safeguard. More indirect measures, such as sex education classes and providing support to women who do not intend to adhere to the virginity norm, might be more effective in achieving the desired outcomes and expressing respect towards these women, and their autonomy and well-being. As we argued with respect to women’s choice to use ARTs, the costs of compliance and non-compliance demand a differentiated approach, one that is sensitive to these “local” considerations.

6. Conclusion

Our aim in this paper has been to engage with two concerns related to gender socialization and ARTs. In response to the autonomous-choice concern, we have argued that only an externalist and substantive conception of autonomy can accommodate the view that women’s choice to use ARTs to have children may be non-genuinely autonomous—and warned against the implications of subscribing to it, especially in the context of reproductive choices and aspirations. In response to the norms-legitimation concern, we have argued that the costs of compliance and non-compliance require a differentiated approach to resisting their entrenchment and mitigate indirect harms to infertile and other women.

By way of offering some synoptic considerations, we would like to stress the non-ideal focus of our responses to these concerns. Ideally, fertile and infertile women would be free to decide whether or not to pursue motherhood, and when and how to pursue it. However, various biological and social constraints affect their ability to do so. Public funding for ARTs does not necessarily make women freer, as many feminist theorists have noted since the early days of IVF. However, it seems to us that such an intervention expands rather than reduces the set of flourishing opportunities available to them and expresses endorsement for their ability to choose what is best for them. While feminist theorists have become less critical of ARTs over the years, some look warily toward novel ARTs or novel ways to use these ARTs. Recent critiques of the publicly and privately funded provision of UTx or so-called “social” egg-freezing exemplify this skepticism. While we agree with critics that laissez-faire or more procedural approaches to conceptualizing and appraising autonomy risk overlooking the influence of oppressive gender socialization on women’s reproductive choices and aspirations, our paper has been motivated by the worry of imposing overly demanding criteria for autonomy and the related implications for women’s freedom in matters of reproduction.