The rights of intended children: The best interests of the child argument in assisted reproduction policy

Previous research on ART policies

Whereas Sweden’s early legal ban on donor anonymity placed it, according to Engeli and Rothmayr Allison (2017), in the category of countries with restrictive ART regulations, legislation that made lesbian couples and single women eligible for ART treatment in 2006 and 2016, respectively, moved the country into the permissive category. ART policy processes, like child rights governance in general, are influenced by the involvement of a range of actors. Alongside confessional political parties and the medical community, women’s movements have been identified as one of the key actors in policy processes concerning ART in several European countries (Engeli and Rothmayr Allison, 2017). Other recent groups of non-partisan political actors in Swedish ART policy processes are organisations that represent the interests of the involuntarily childless. The Swedish organisation Barnlängtan (Child Longing) cooperates with lesbian, gay, bisexual, and transgender (LGBT) organisations and an organisation for voluntarily single mothers and claims the equal rights of all – medically and socially – involuntarily childless to assistance (Gunnarsson Payne and Korolczuk, 2016). The role of child rights advocates, however, has hitherto attracted little interest in research on ART policies.

A recurring theme in research on ART policies and discourses is an interest in what family ideals and forms are supported through the regulation of access to ART (Baylis and McLeod, 2014; Lie and Lykke, 2017; Shanley, 2001; Tinnerholm Ljungberg, 2015). Primarily, the mobilisation of the BIP has been addressed in relation to the requirement for donor disclosure (e.g. Cohen, 2011) and questions concerning the legal status of the child–parent relationship in donor ART and surrogacy (e.g. Singer, 2006; Stoll, 2013). The consideration of the best interests of the child in assisted reproduction has, however, also been regarded as inherently discriminatory, both because such considerations are not applied to unassisted reproduction and because they are inevitably informed by dominant constructions of the family, motherhood and the child (Blyth and et al, 2008; see also Quigley, 2010). Despite the increasing acceptance of diverse family forms through the inclusion of lesbian couples and single women as eligible recipients of treatment with ART in many countries, it has been noted that ART policies continue to promote ideals of genetic relatedness and coupledom as the basis for the creation of families (Cutas, 2018; Nordqvist, 2006). Swedish ART legislation remains applicable only to single women or couples (married or common-law spouses or registered partners) and thereby excludes friends and other constellations from seeking assisted reproduction treatment together (SFS 2018:1283).

The BIP

In his philosophical exploration of the BIP, Archard (2002) suggests two analytical questions: what is being given weight (i.e. what meaning is the principle given in a specific context?) and how much weight is it given (in relation to the interests of other parties)? Discussing different weightings that can be given to the BIP, Archard (2002) states that a consideration that is primary is leading, whereas a paramount consideration trumps all other considerations. Lagerberg (2014) similarly distinguishes between an absolute and a relative application of the BIP. In contrast to an absolute application, a relative application can be expected to leave room for taking into consideration the interests of parties other than the child, for negotiations of acceptable risks or for the acceptance of less-than-optimal outcomes for the child. Furthermore, a distinction can be made between an interpretation of the BIP as a maximum welfare principle, on one hand, and as a minimum threshold principle, on the other. While the first mentioned implies that a child should not knowingly be brought into the world in less than ideal circumstances, the latter demands only that the intended child’s life should not be worse than not existing (Pennings, 2011). To some extent, the wording of the BIP determines the weight that may be attributed in an argument. However, wordings of the BIP may be inconsistent, as we will see, even within one and the same document.

In the drafting of the United Nations (1989) Convention of the Rights of the Child (UNCRC), the question whether unborn children are included in the definition of the child or not was left unanswered. Furthermore, the child in assisted reproduction is not only not yet born but also not yet conceived. Children, as philosopher David Heyd (1992) argues, cannot be harmed simply by creating them, and the interests of potential people cannot be taken into consideration. And yet, the BIP has recurrently been mobilised as a rhetorical resource in ART policy processes.

In her study on Finnish ART policy discourses, Kanckos (2012) refers to the best interests of the child as a floating signifier that has been used in arguments both for and against making ART accessible to same-sex couples and single women. Whereas the interests of the child in having two parents were pitted against single women’s equal right to ART treatment by conservative MPs, liberal MPs argued that it was in the best interests of the child that the family forms they were born into were accepted. The use of the best interests of the child on both sides of the argument indicates not only its emptiness as a signifier but also its hegemonic status as a rhetorical resource. In Swedish ART policy debates, too, the best interests of the child argument has played a prominent role. According to Jonsson Malm (2011), the Insemination Inquiry of the early 1980s and its commissioner, Tor Sverne, came to portray themselves as advocates of the interests of the child against unscrupulous involuntarily childless adults, doctors and scientists. To not make the interests of the child a paramount concern was, according to Jonsson Malm (2011), represented as unethical by the inquiry. Liljestrand (1995) too ascribes considerable power to the best interests of the child argument through her representation of the BIP as ‘a smoke screen’, a rhetorical cover-up for other interests (cf. Finlayson, 2007), such as the social control of individual citizens and their procreation through the ban on donor anonymity (p. 275).

Hence, it is reasonable to expect that the BIP has played a significant role as a rhetorical resource in Swedish ART policy processes. However, as governments are obliged to take into consideration not only the interests of the child but also of the childless adults, I am particularly interested in the meaning and weight that is attributed to the BIP in relation to policy decisions where these interests are claimed to be in conflict.

Political argument, governance and governmentality

The goal of arguments during policy processes is generally to convince others of a certain policy content, and persuasion is therefore a central aspect of policy-making (Fischer and Gottweis, 2012; Gottweis, 2012). Discourses determine which arguments will be regarded as legitimate and which not, but they generally leave room for different lines of argumentation and rhetorical choices (Fischer and Gottweis, 2012).

Contemporary research on governance and governmentality alike has an interest in the diffusion of power from the state to civil society (Bevir, 2011). Holzscheiter (2005) emphasises the possibilities for NGOs to not only implement agreed norms but to influence discourses, norms and the setting of agendas in policy-making. Hence, in a political context, in which the BIP is agreed upon as a primary consideration, acknowledged child rights advocates can be expected to have considerable capital in policy matters that concern children. This makes their input to the policy process interesting to investigate in an analysis of the negotiation of the BIP in ART policy processes. Criticising the recent emphasis on the role of NGOs in global governance literature, Bevir (2011) maintains that the state remains a powerful actor within the policy process. The policy agenda set by governments, as well as governmental response to the input of child rights advocates is therefore of equal importance.

Data and analytical procedures

The primary data for the analysis consist of the reports of government inquiries, government bills, ministry memoranda and reports by the Swedish National Council on Medical Ethics. ² In addition, the referral statements submitted in response to the inquiries and memoranda by BRIS, Rädda Barnen and Barnombudsmannen have been analysed. As child rights advocates, their rhetorical use of the BIP, and the weight they attribute to it as an argument in their referral statements, serve to contrast the manner in which the BIP is used as a rhetorical resource in government inquiry reports, memoranda and bills.

In the Swedish policy process, it is the government that both defines the policy question under consideration and appoints the inquiry that is commissioned to investigate the matter. Referral bodies, such as NGOs and authorities, are invited to respond to the inquiry’s proposal by submitting referral statements, before the government proposes a bill to be voted on by parliament. Through their referral statements, child rights advocates may offer alternative interpretations of the BIP, and they may also offer a different weighting of the BIP. The subsequent government bill presented to parliament is an indication of the extent to which the opinions of various referral bodies have been taken into consideration by the government (Lundqvist, 2007).

The analytical procedures draw upon the reconstruction of argumentative discourse (Van Eemeren, 2010) and a discourse-historical approach (Wodak, 2015). They entail the identification of the point of departure of the argumentation, the explicit and implicit arguments and the outcome of the argumentation. All documents were searched for sections where the phrase ‘the best interests of the child’ (barnets bästa), or similar phrases addressing the rights or needs of the child, was used. I then examined how the BIP was referred to linguistically in order to assess whether its wording suggested an absolute or relative interpretation of the BIP. Furthermore, the meaning that the best interests of the child was attributed in each case was identified, alongside the policy arguments that the reference to the BIP was used to support. Explicit and implicit shifts and discrepancies regarding the wording of the BIP were identified within the documents, as well as in relation to specific policy questions that were processed over a longer time-span. Below, I will first give an overview of the mobilisation of the BIP and the UNCRC in Swedish ART policy processes, before I will move on to the analysis centred on the limits of the scope of relevance of the BIP, the negotiation of the weight of the BIP and the negotiation of the meaning of the BIP.

The best interests of the child, the UNCRC and Swedish ART policies

In Swedish government inquiry reports and bills on assisted reproduction, early references to the best interests of the child were made in the first inquiry on insemination published in 1953 (Statens offentliga utredningar [SOU] 1953:9). Subsequent inquiry reports contained numerous references to children’s rights (e.g. SOU 1983:42: 17; SOU 1985:5: 47) or to the BIP as ‘the lodestar’ (SOU 1983:42: 14). In the second insemination inquiry’s report, the claim that donor-conceived children should have a right to information about the donor’s identity was primarily supported by references to experiences from adoption (SOU 1983:42). Its commissioner, Tor Sverne, had previously served as commissioner of the government inquiry into the rights of the child, which in its first report (SOU 1978:10) argued in favour of a legal ban on the corporal punishment of children. The connection between the two inquiries through Sverne has been ascribed importance in explaining the prominent role that children’s rights came to play in the inquiry’s proposal to prohibit the use of anonymous donors (Assisted Reproduction in the Nordic Countries, 2006).

Sweden ratified the UNCRC in 1990. In the bill preceding the decision, Sweden’s legal ban on donor anonymity was mentioned in relation to Article 7 on the child’s right to know her or his parents (Bill 1989/90:107). In 1994, the Committee on the Rights of the Child noted the possible contradiction between Article 7 and ART policies that allowed donors to be anonymous (paragraph 10) and explicitly interpreted Article 7 as including the right of children born through donor ART to have access to information about their genetic origins in 2016 (paragraphs 33–34). In the same year, the Ministry of Health and Social Affairs commissioned the Swedish National Council on Medical Ethics (SMER) to investigate the question whether the use of donor gametes in IVF and oocyte donation should be legalised. Proposing the legalisation of both, SMER addressed the rights of the child, however, without referring to the UNCRC (SMER, 1995). Instead, the UNCRC was mentioned in the referral statement by the 2 years previously appointed first Children’s Ombudsperson Louise Sylwander, who anchored her disapproval of SMER’s proposal in UNCRC Article 3 (Barnombudsmannen dnr 9:235/95), and in the referral statement by the Swedish Paediatric Society (Svenska Barnläkarföreningen) (Ds 2000:51).

After that references to the UNCRC began to appear more frequently. In particular, the inquiry reports that were published after the Swedish government’s decision to work towards the incorporation of the UNCRC in 2013 contained lengthier accounts of the UNCRC and its content. I will return to the UNCRC and the role that it played in the argumentation concerning the legalisation of ART treatment for single women towards the end of the analysis. The articles that were cited most frequently were Article 3 (the best interests of the child) and Articles 7 and 8 about the child’s right to know her or his parents and to preserve her or his identity (e.g. Ds 2004:19; SMER, 2013; SOU 2019:24). However, the BIP was referred to as alternately ‘the most central ethical premise’ (e.g. SOU 2014:29: 105) and as one (of several) central principles (e.g. Bill 2014/15:127: 32), which indicates a negotiation of its weighting. In the next section of the analysis, I will attend to one area of ART policies – the regulation of ART treatment with the candidates’ own gametes – which illustrates the limits of the BIP’s scope of relevance as an argument.

The limits of the BIP

In non-donor insemination and non-donor IVF, the gametes of the man and woman undergoing treatment are used, and no donor is involved. Current Swedish legislation on ART distinguishes between donor and non-donor treatment by demanding that an assessment of the candidates’ suitability as parents be made only in donor treatment. ³ The distinction illustrates that the scope of relevance of the BIP as an argument is limited to certain areas of assisted reproduction.

In its first report, the Insemination Inquiry concluded ‘that donor insemination shall be accepted only if it can be assumed that the future child will grow up under good circumstances’ (SOU 1983:42: 87, italics added by author) – treatment that would result in a complete genetic bond between parents and child. The inquiry stated was comparable to ‘natural’ conception (SOU 1983:42: 86). However, the exclusion of non-donor insemination from the ‘good circumstances’ requirement is not further commented upon or explained. One possible interpretation of the distinction is that it was discursively impossible to suggest the regulation of natural conception, not least because of Sweden’s inter- to post-war history of forced sterilisation (Lind, 2008).

Explicitly targeting the distinction between donor and non-donor treatment in the legislation as a problem, in 2000 the Ministry of Health and Social Affairs (Ds 2000:51) proposed the removal of the sections containing the ‘good circumstances’ requirement in the insemination act and the act regulating conception outside the body (SFS 1984:1140, §3:2; SFS 1988:711, §5:1). The ministry did not offer any further comments on its proposal. It also did not cite the content of the section, thereby drawing minimal attention to the fact that the proposal would entail the removal of a requirement that was motivated with reference to the BIP.

In its referral statement, the child rights organisation BRIS proposed instead that the requirement that the child can be expected to grow up under good circumstances should concern both donor and non-donor insemination and IVF and argued strongly against the removal. Furthermore, BRIS argued that

[a] general goal in other areas of legislation is to signal through rules of law that the best interests of the child should be the primary consideration in decisions that concern children. There is no reason to move in another direction when it comes to assisted reproduction. (Dnr 2000/5967/HS)

The absence of explanations and the proposal’s low profile in the memorandum suggests that, even if the consequences of taking the best interests of the child into consideration – in this case the requirement that the future child can be expected to grow up under good circumstances – were unwanted, the ministry was not prepared to argue explicitly in favour of a rejection of the relevance of the BIP. The ‘good circumstances’ requirement, as well as the distinction between donor and non-donor treatment, remained in place in the subsequent government bill (Bill 2001/02:89) as in subsequent legislation despite the Swedish National Council on Medical Ethics proposal in 2013 ‘that an assessment that considers the best interests of the child is made in all treatment forms’ (SMER, 2013: 27). While the above demonstrates that the BIP’s scope of relevance was limited to assisted reproduction treatment with donor gametes, the next section will discuss a policy process that demonstrates the manner in which the weight of the BIP was negotiated when consideration of the BIP as an absolute principle seemed to threaten a desired outcome of the policy process.

Negotiating the weight of the BIP

The proposed ban on donor anonymity was controversial as it was claimed to discourage possible donors to the extent that it would be impossible to continue to offer donor insemination (Jonsson Malm, 2011; Tinnerholm Ljungberg, 2015). As response to the Insemination Inquiry’s proposal to prohibit the use of anonymous donors, both supporters and critics of donor disclosure mobilised the BIP in support of their standpoints. While those in support of donor disclosure claimed the right of the child to knowledge about her or his genetic origin, those against stated that disclosure of the identity of the donor would have a negative effect on the relationship between parents and child in the insemination family (Bill 1984/85:2). One of the main concerns in relation to the legalisation of IVF, by contrast, was related to the possible medical risks that this technology entailed for the child. While an alternative interpretation of the BIP made it possible to claim, albeit unsuccessfully, that donor anonymity was indeed in the best interests of the child, it was rhetorically impossible to frame medical risks as being in the best interests of the child. Rather than competing meanings of the best interests of the child, the negotiation concerns the weight attributed to the BIP. This makes an analysis of the mobilisation of the BIP in this policy context particularly interesting.

When IVF was first legalised in 1988 neither BRIS nor Rädda Barnen submitted any referral statements on the inquiry report (Bill 1987/88:160). A few years later, however, the Swedish Paediatric Society claimed in its referral statement on a report by the Swedish National Council on Medical Ethics (SMER, 1995) that the child was ‘subjected to significant risks in order to satisfy the needs of parents to carry a pregnancy’ through all forms of IVF (Ds 2000:51: 32). The Society demanded further investigations into the medical risks for children, and in 1998 and 2000 the National Board of Health and Welfare (NBHW), together with the Swedish Society for Obstetrics and Gynaecology and the Swedish Paediatric Society, published two reports examining the medical risks to children born after non-donor IVF between 1982 and 1995 (SoS-rapport 1998:7; SoS-rapport 2000:4).

Both reports, as well as the subsequent recommendations by the NBHW, primarily targeted the medical risks resulting from the until-then common practice in IVF of multiple embryo transfers (SoS-rapport 2000:4). The solution recommended by the NBHW was a change in routines from multiple to single-embryo transfers to avoid multiple pregnancies, a solution that faced objections because it was thought to decrease the chance of achieving pregnancy in each treatment cycle (SoS-rapport 1998:7; SoS-rapport 2000:4). However, the reports also indicated that the prevalence of preterm births and low birth-weight was slightly higher in single IVF pregnancies too, as were the risks of being born with cerebral palsy, of having psychomotor developmental delays and being in need of habilitation services, even if the number of affected children was small (SoS-rapport 1998:7; SoS-rapport 2000:4). According to the second report, this indicated ‘that other factors than multiple pregnancies also contribute to causing these risks’ (SoS-rapport 2000:4: 23).

In a subsequent memorandum on the use of donor gametes in IVF, the Ministry of Health and Social Affairs (Ds 2000:51) followed the recommendation of the NBHW by demanding that routines be changed to single-embryo transfers, but did not comment further on the reports’ conclusions that the increased medical risks for IVF-conceived children could not be explained by multiple pregnancies alone (Ds 2000:51). As its general ethical standpoint, the inquiry declared that ‘[c]hildren born as the result of assisted reproduction of course should not be subjected to greater risks or have poorer preconditions than children born after natural conception’ (Ds 2000:51: 26), which suggests an interpretation of the BIP as absolute (Lagerberg, 2014). Yet, at the same time, the ministry stated that the ‘couple’s needs and desire to have a child must be balanced against the complications or risks for the child’ and that an assessment of ‘which risks can be regarded as acceptable’ must be made (Ds 2000:51: 26). The demand that these children should not be subjected to greater risks than other children was transformed into a demand that ‘everything is done to reduce the risks for children who are conceived through IVF treatment’ (Ds 2000:51: 36; see also Bill 2001/02:89: 23, 26). As several reports indicated that IVF-conceived children were subjected to, albeit slightly, greater medical risks, a zero-tolerance stance on greater medical risks for IVF-conceived children was no longer possible to maintain. Justification of the continued use of IVF required a discursive shift that represented the reduction of the medical risks as sufficient from the best interests of the child perspective.

A wide range of professional, political and confessional associations, university faculties, fertility clinics NGOs and interest groups were invited to submit referral statements on the Ministry of Health and Social Affairs’ memorandum, including BRIS, Rädda Barnen and Barnombudsmannen. In their referral statement, Rädda Barnen expressed hesitancy towards all forms of IVF treatment, including those that were already in use, and claimed that it had not been investigated properly whether ‘the higher risk of deformities and death’ for IVF-conceived children that was described in the 1998 and 2000 reports was properly addressed through a change in routines to single-embryo transfers (Dnr S2000/5967/HS Rädda Barnen). Barnombudsmannen and BRIS similarly criticised the inquiry for not having applied the BIP and for having focused on the needs and interests of adults alone (Dnr S2000/5967/HS Barnombudsmannen, Dnr S2000/5967/HS BRIS). Furthermore, Rädda Barnen specifically targeted the inquiry’s explicit balancing of the interests of the child and those of the childless adults and stated that ‘a child perspective left no room for doubt that the only consideration should be the best interests of the child’ (Dnr S2000/5967/HS Rädda Barnen). The criticism delivered by the three child rights advocates illustrates an interpretation of the BIP that clearly attributed more weight to the BIP than the ministry’s memorandum did. None of them referred to the UNCRC to support their criticism of the memorandum’s reasoning on IVF, which interestingly the government did do in the account of relevant legal documents in its subsequent bill presented to parliament (Bill 2001/02:89). In addition to Articles 3 and 7, it referred to Article 6, which was said to demand active protection of the child’s right to life and development, and Article 24, which demands actions to decrease infant mortality. Given the proposal to legalise the use of donor gametes in IVF treatment, the mobilisation of the references to these UNCRC articles must be interpreted as support mainly for the change in routines to single-embryo transfers.

What was at stake through the child rights advocates’ objections to IVF was the continued use of a technology that by the time of the publishing of the NBHW reports was legal and already widely accepted. Through the mobilisation of the BIP as an argument in support of single-embryo transfer instead, commitment to the best interests of the child could be demonstrated without questioning IVF as such. The analysis presented in the next section concerns a policy question, in relation to which different meanings of the BIP were mobilised and negotiated, and the manner in which those interpretations that were supported by references to the articles of the UNCRC gained primacy over other interpretations.

The primacy of UNCRC-supported child rights

The first motion in favour of legalising assisted reproduction for single women presented to the Swedish parliament (Motion 1984/85:22) was rejected by means of references to the best interests of the child and the claim that children should have two legal parents at birth (Bill 1984/85:2). The fact that the future child would have two parents, albeit not a mother and a father, 20 years later also constituted the main argument for the legalisation of assisted reproduction for lesbian couples (Bill 2004/05:137).

During the subsequent policy process, the arguments that came to constitute the two main pillars in support of assisted reproduction for single women were the rights of single women to equal access to assisted reproduction and the rights of children to knowledge about their genetic origins. As many single women already went abroad for treatment, where donors were allowed to be anonymous, a change in policy that made it possible for single women to receive assisted reproduction within the Swedish healthcare system was claimed to strengthen the right of the child to knowledge about her or his genetic origins (2011/12: SoU26: 13). As a result, the proposed policy change was represented as not only compatible with the best interests of the child, but as actively furthering the child’s interests.

In 2013, the government decided to legalise assisted reproduction for single women. It appointed an inquiry, but pre-empted the inquiry’s report with the decision and therefore commissioned it to suggest the necessary policy changes, rather than to investigate whether assisted reproduction should be made available to single women (SOU 2014:29: 195). In its directives to the inquiry, the government stated that the ‘the best interests of the child’ should be made ‘the foremost consideration’ (p. 207) and that ‘the best interests of the child must be the focal point, including the right of the child to knowledge about her/his origins’ (SOU 2014:29: 197–198). However, as the decision to legalise assisted reproduction for single women was already made, the scope of the consideration of the BIP did not concern the question of the legalisation as such.

Rädda Barnen and Barnombudsmannen both supported the legalisation of assisted reproduction for single women and referred to it – with explicit references to UNCRC Article 7 – as a means to strengthen the rights of the child to knowledge about her or his origins (Ju2014/03407/L2 Rädda Barnen, Barnombudsmannen). It is therefore noteworthy that the inquiry itself stated that it could

hardly be argued that the motive for the introduction of the possibility of assisted reproduction for singles is in the best interests of the child. The reason for its introduction is instead a wish to assist more involuntarily childless persons to conceive a child in this manner. (SOU 2014:29: 105–106)

Thereby, it implicitly rejected the claim that the strengthening of the child’s right to information about the donor’s identity constituted a motive for a policy change. Instead, it regarded the government’s previous standpoint as still valid and claimed that ‘the main rule should continue to be that a child has a right to two legal parents’ (SOU 2014:29: 106).

Both Rädda Barnen and Barnombudsmannen also proposed more far-reaching policy changes in order to secure the rights of donor-conceived children to information about their genetic origins. Barnombudsmannen demanded that the donor’s name should be entered into the population registry because research indicated that many parents did not disclose to their children that they had been conceived with the assistance of a donor, which in effect hindered children from finding out who the donor was (Ju2014/03407/L2 Barnombudsmannen). None of these were implemented, however. Neither the government’s own explicit goal to ensure the right of the donor-conceived child to information about the donor’s identity nor the status of Rädda Barnen and Barnombudsmannen as child rights advocates served as sufficient rhetorical support for policy changes that would reduce parental control over the disclosure of information.

The argument that the child’s right to information about her or his genetic origins would be strengthened through the legalisation of assisted reproduction because women already received treatment abroad, where donor anonymity was allowed (Bill 2014/15:127), is of interest for two reasons. First, the argument served as a response to the critique that the proposal did not take into consideration the interests of the child, especially as it could be supported with reference to the UNCRC (Article 7). Second, because of Sweden’s early legal ban on donor anonymity and a long-standing tradition of arguing in support of donor disclosure in the name of the best interests of the child, such an interpretation of the BIP was able to rhetorically trump alternative interpretations that emphasised the child’s interest in having two parents (Bet. 2011/12: SoU26).

Concluding discussion

The focal point of this analysis has been the mobilisation of children’s rights and the BIP as rhetorical resources in the governance of assisted reproduction. The fact that Swedish governmental inquiry reports and bills on ART policies do contain numerous references to the BIP in itself says something about the status of the BIP as an argumentative resource. The explicit commitment to the BIP in early Swedish ART policies and Sweden’s general official commitment to the protection of children’s rights created what could be referred to as a rhetorical path dependency (Grube, 2016) in subsequent ART discourses and arguments. Liljestrand (1995) has described the best interests of the child argument as a smokescreen that served as a cover up for other purposes than the protection of the welfare of the child. Given the status of the BIP as a powerful rhetorical resource, its mobilisation may also be interpreted as compliance with a rhetorical imperative.

There are, however, two noteworthy limits to the scope of the mobilisation of the BIP in government argumentation on assisted reproduction. First, the BIP in government inquiry reports, bills and memoranda is mobilised primarily in relation to donor treatment. Hence, the assessment of the capacity as parents of candidates is seen as necessary in treatment with donor gametes, but not in treatment forms that will result in full genetic bonds between the intended parents and child. Sweden’s legal ban on donor anonymity in 1984, explicitly motivated by the child’s right to information about her or his genetic origins, earned Sweden an early reputation for making the interests of the intended child a primary consideration in the regulation of assisted reproduction. The Committee on the Rights of the Child’s interpretation of Article 7 as supporting donor disclosure contributed to further justifying the legal ban on donor anonymity as a legitimate policy consequence of making the best interests of the child a primary consideration in ART. Hence, when two different meanings of the best interests of the child – the interests of the child in having two legal parents and the interests of the child in having access to information about the donor’s identity – were mobilised as arguments against and in support of the legalisation of assisted reproduction for single women, respectively, the rhetorical power of the latter trumped that of the former. Not only had the ‘child’s right to information about the donor’ argument the support of the Committee on the Rights of the Child, but it also served as an argument for the policy outcome that had already been decided upon by the government, that is, to legalise assisted reproduction for single women.

Second, when an absolute interpretation of the BIP threatened the policy outcomes that were preferred by the government, the mobilisation of the BIP shifted to a relative interpretation. The analysis demonstrated that when the BIP was mobilised by child rights advocates such as BRIS, Rädda Barnen and Barnombudsmannen in the early 2000s in order to question the continued use of IVF, a technology that was already in widespread use, the government’s official commitment to the best interests of the child constituted an argumentative predicament. A zero-tolerance stance on greater medical risks for IVF-conceived children compared to other children threatened the justification of the continued use of IVF. Such a justification instead required the interpretation of the BIP as a relative principle, and a shift from the zero-tolerance stance to the representation of the goal as a reduction, rather than the elimination, of the relative risk increase for IVF-conceived children.

By focusing on assisted reproduction, a policy area in which the interests of intended children are in potential conflict with the reproductive rights of adults, the limits of governance in the name of the rights of the child can be traced. In adoption, a practice that concerns existing children and adults, who can invoke neither parental nor reproductive rights, the best interests of the child is referred to as not merely a primary, but as the paramount consideration (UNCRC, Article 21). At the other end of the spectrum, that is, in unassisted reproduction, by contrast, it is impossible to mobilise the BIP as an argument (Areschoug, 2005; Lind, 2008). It is politically impossible to demand the vetting of all parents, as well as restriction of the reproductive liberty of adults who can conceive without assistance. Assisted reproduction constitutes a middle ground, a policy area where the scales are clearly tipping in favour of neither the interests of the child nor the interests of the adults. The BIP is not the only consideration for policy-makers, yet at the same time the status of the BIP as a rhetorical imperative, in my examples from Sweden, demands explicit acceptance of the best interests of the child as a primary consideration in the practice of child rights governance. When the BIP is mobilised, for example, by child rights advocates, in a manner that threatens the intended policy direction of the government, either the weight, or the meaning of the best interests of the child or its scope of relevance is manoeuvred around. This analysis demonstrates, I suggest, the hegemonic status of the best interests of the child argument, as well as the limits of its influence on governance.