introduction

In 2012, the Daily Telegraph reported that two doctors working in a private medical practice were prepared to authorise an undercover journalist’s request for an abortion based on the sex of the foetus (Watt et al., 2012). Although transcripts of these conversations revealed that sex had been mentioned in relation to a genetic disorder—which can be sex-specific—in one case, the article omitted this significant detail. Over the ensuing months, the Daily Telegraph ran several more stories about what it termed ‘gendercide’ (Pell, 2013; Perry, 2014), a theme picked up by other newspapers (Connor, 2014).

The then Health Secretary Andrew Lansley (2012) condemned SSA as an ‘illegal’ and ‘morally repugnant’ practice and subsequently ordered the Care Quality Commission (CQC)—the independent regulator of health and social care in England and Wales—to inspect all abortion providers to ensure that legal requirements were being met. He also announced that he would be passing on the information about the two doctors from the original Daily Telegraph report (Watt et al., 2012) to the police with a view to prosecution. The CQC’s (2012) investigation found no evidence of illegally authorised abortions; in September 2013, the Director of Public Prosecutions announced that the two doctors would not face charges (CPS, 2013). Although the Christian Legal Centre brought private charges against the doctors, the Crown Prosecution Service succeeded in quashing the case.

The debate resurfaced in January 2014 when The Independent claimed that the practice of sex-selective abortion (SSA) was ‘commonplace’ in the UK (Connor, 2014). Drawing on terminology mirroring Amartya Sen’s (1990) critique of ‘missing girls’ in India, the article estimated that between 1,400 and 4,700 girls had been ‘lost’ because of this practice (Connor, 2014). In March 2014, the then Prime Minister David Cameron told the House of Commons that ‘It is a simply appalling practice, and in areas such as that, and female genital mutilation and forced marriage, we need to be absolutely clear about our values and the messages we send and about these practices being unacceptable’ (Hansard, 2014a, column 780). By aligning SSA with forced marriage and female genital mutilation (FGM), Cameron framed it as a ‘harmful traditional practice’ in communities whose values are different from ‘our’ more egalitarian ideals, thereby suggesting a binary distinction between the West and the ‘Third World’.

Cameron's statement implies that any violence against women that occurs in the West is against ‘our’ dominant value system and is an individual and idiosyncratic act in a context where gender equality is widely accepted and has been achieved, in contrast to the ‘Third World’—and ethnic minority communities in the UK—where such violence is part of ‘their’ tradition and value system. Such assumptions, including the United Nation’s conceptualisation of ‘harmful traditional practices’ (Winter et al., 2002), have been critiqued by postcolonial and transnational feminists (Mohanty, 1991; Narayan, 1997). Criticism relating to the neglect of the intersections between race/ethnicity and gender voiced in the 1990s (Afshar and Maynard, 1994) still hold; these limitations are evident in the insufficient attention given to political and economic issues in contrast to cultural debates. Recent events surrounding the media coverage and parliamentary debates on SSA point to the importance of unpacking these discourses around race/ethnicity, which continue to shape public and policy debates in critical ways.

In May 2014, the Department of Health (DOH) (2014) sought to clarify the law on SSA through guidelines for ‘all those responsible for commissioning, providing and managing service provision’ by stating that ‘abortion on the grounds of gender alone is illegal’. However, under current law—which was also applicable in 2014—it is lawful to abort a foetus when two registered medical practitioners form an opinion, in good faith, ‘that there is a substantial risk that if the child were born it would suffer from such physical or mental abnormalities as to be seriously handicapped’, further recognising that ‘some serious conditions are … gender-related’ (Abortion Act 1967, 1967).

However, Conservative MP Fiona Bruce (2014), the chair of the All Party Pro-Life Group, argued that interpretations of the law were not consistent and led a campaign to restrict SSA. Anti-abortion and religious groups such as The Christian Institute, Christian Concern, the ProLife Alliance and Society for the Protection of Unborn Children, as well as some organisations supporting South Asian women such as Jeena International and Karma Nirvana,Footnote 1 mobilised in support of the prohibition through the ‘Stop Gendercide’ campaign. Stop Gendercide likened SSA to FGM and forced marriage, and then used this parallel to call on the Conservative/Liberal Democrat-coalition government to show the same initiative it had taken on other forms of violence against women and girls (VAWG) within minority communities. The term ‘gendercide’ evokes the related term ‘femicide’, which was coined to describe ‘the killing of females by males because they are females’ (Russell, 2011) and has been used to draw attention to violence against women (Radford and Russell, 1992), and thereby serves to co-opt the language of feminism and frame this issue as a matter of women’s rights.

Later in 2014, Bruce introduced the Abortion (Sex-Selection) Bill 2014–15 (2014) as a Private Members’ Bill under the Ten Minute Rule;Footnote 2 the Bill, which aimed to clarify the law relating to abortion on the basis of sex-selection, received cross-party support. At the conclusion of its Committee Stage, Fiona Bruce also proposed an amendment to Section 65, Part 5 of the Serious Crime Bill 2014–15, which deals with crimes such as FGM that have a disproportionate effect on girls. This proposed amendment recommended barring the termination of pregnancy ‘on the grounds of the sex of the unborn child’ (Hansard, 2014b).

These debates on SSA were examined by Ellie Lee (2017), who unpacked the construction and framing of the core arguments. Lee charted the chronological development of claims that constructed SSA as a problem by scrutinising three key issues: the vilification of abortion providers, the role of statistics, and the significance of personal testimony. Our paper draws attention to a connected but different area of inquiry: the role of ideology in policymaking. This helps to explore the original cross-party support for restricting SSA, and why this consensus could not be sustained in the second set of parliamentary debates on the issue. The data underpinning our study comprises the two sets of parliamentary debates as recorded in Hansard, along with the related submissions made to Parliament by expert communities and networks. While we recognise the medicalisation discourse that prevails in the UK and has informed debates on abortion, including SSA (Sheldon, 1997; Lee, 2004, 2017), we situate our analysis in the context of the existing scholarship on the shift towards ‘women-protective’ framings in abortion discourses (Siegel, 2014 [2012]; Jesudason and Weitz, 2015) and feminist analyses of policymaking more generally. Our contribution focuses on racialised discourses on abortion and how these relate to broader culturally essentialist framings of VAWG in minority ethnic communities (Mohanty, 1991; Narayan, 1997).

problematising sex-selective abortion in the UK

SSA is defined as the targeted abortion of female foetuses. While the biological norm is 105 male births to 100 female births (Jha et al., 2006), sex ratios are considerably skewed in favour of male children in certain parts of the developing world. When combined with greater male survival rates throughout the life cycle, this skewed ratio causes a female ‘deficit’ in the sex ratios of countries such as India and China (Bhaskar, 2011). Anartya Sen (1990) suggests this deficit is directly attributable to human interventions including SSA, the gender-biased allocation of resources (leading to lower female survival rates), and the influence of gender-iniquitous social relations. Global estimates of ‘missing women’ (ibid.) have ranged from 60 to 101 million; estimates for South Asia suggest that there are currently 40 million missing women in India, 3.8 million in Bangladesh, and 4.9 million in Pakistan (Jha et al., 2006). Both the practice of SSA and prenatal screening to determine the sex of the foetus have been criminalised in India since 1994; the law, however, is poorly enforced.

The technology to enable SSA is widely available and information about the sex of the foetus is commonly provided to expectant parents at prenatal check-ups in the UK. Both research and the anecdotal experience of practitioners who work with South Asian women in the diaspora indicate a widespread preference for sons, leading to pressure on women to produce male progeny (Purewal, 2003), with a failure to do so sometimes resulting in violence (Anitha, 2011).

Dubuc and Coleman (2007) used data from the UK registration system for births to compare the child sex ratios of immigrant and UK-born mothers. The study indicates that, while the sex ratio of births to mothers born in India was relatively even in the 1970s and 1980s, it increased in favour of boys and diverged significantly from the general pattern thereafter. Dubuc and Coleman’s (ibid.) research data suggests that, between 1969 and 1979, the sex ratio of births to Indian-born women in England and Wales rose from 103 boys to 100 girls, to 114.4 boys to 100 girls between 2000 and 2005. However, subsequent studies have shown no evidence of differences between British- and Indian-born women, suggesting that the scale of SSA may be too low to skew overall ratios. Thus, on the whole, quantitative data on birth ratios suggests that SSA is ‘relatively little taken up and fails to reach a threshold of visibility’ (Bubeck, 2002 cited in Greasley, 2016, p. 563).

Recording of the self-reported ethnicity of women having abortions was introduced in 2002 in the UK. In 2015, among women with ethnicity recorded, 75 per cent of those having an abortion were White, 9 per cent were Asian or Asian British, and 8 per cent were Black or Black British (Harker, 2016, p. 5). These figures suggest that abortion rates are proportionally higher for non-white ethnic groups, including—but not limited to—South Asian communities. The DOH has undertaken detailed analysis to investigate whether the relevant ratios in the UK vary beyond the range that might be expected to occur naturally. Their analysis, first published in May 2013 and updated in 2014 (DOH, 2013), concluded that no ethnic group was statistically different from any other in terms of the sex ratios at birth that would be expected to occur naturally. The DOH has stated that it will continue to repeat this analysis annually as new birth data becomes available.

analysing policymaking

Michel Foucault’s (2000, p. 118) concept of ‘problematisation’ furthers our understanding of the role of ideology in policymaking, directing attention to the ‘development of a given into a question’ that ‘transform[s] a group of obstacles and difficulties into problems to which the diverse solutions will attempt to produce a response’. Carol Lee Bacchi (2009) takes a Foucauldian approach, grounded in discourse analysis, to offer a theoretical framework for policy analysis: the ‘What’s the problem represented to be?’ (WPR) approach. Starting with the assumption that a problem is not a given but rather a social construct, she draws attention to how particular conditions come to be characterised as social problems and the effects this has on them, including in relation to the assumptions and presumptions underlying the representation of the problem and also what is left unproblematised. Such an approach allows us to unpick the processes whereby SSA is constructed as a social problem in policy debates and the contours of these processes.

From 1997 to 2010, the Labour government advocated ‘inclusive’ approaches to policymaking and parliamentary analysis (Blunkett, 2000) drawing on problem-solving paradigms rooted in evidence, including that provided by external actors such as practitioner organisations, charities and professional bodies (Brazier et al., 2008). The positivist, rationalist assumptions involved have been critiqued widely by scholars who draw attention to the role of ideology and its discursive character in the inception of policy, in policymaking, and in policy analysis (Bacchi, 2009). For example, Nutley and Webb (2000) examine the role of values and structural factors in policymaking and argue for greater appreciation of the ways in which language can highlight how policy discourses emerge and frame the construction of problems and agendas. Taking a social-constructionist stance, Bacchi (2009) offers a critique of approaches that treat policymaking as a neutral, technical process, and by examining how policymakers are active in creating or producing ‘problems’, she seeks to shift the focus from problem-solving to problem-questioning.

The issues underlying these critiques have been articulated through the related concepts of ‘framing’ and ‘frame analysis’ (Goffman, 1974) in scholarship focused on how social issues are constructed in the process of making and conveying public policy. According to Rein and Schön (1993, p. 146), this process enables the shaping of ‘a perspective from which […] a situation can be made sense of and acted on’. The way communications are framed through highlighting some aspects of a subject, while omitting others, influences public opinions, beliefs and behaviours. Benford and Snow (2000) argue that framing is an active process of meaning-making, whereby actors are engaged in a ‘politics of signification’ (Hall, 1982) that entails struggle over the production of mobilising and counter-mobilising ideas. Nelson and Oxley (1999, p. 1041) define these tussles over meaning as ‘issue framing’: groups use ‘alternative definitions, constructions or depictions of a policy problem’ to help them characterise issues in terms that will convince a variety of actors (e.g., the press or general public) to support their position. For Chong and Druckman (2007, p. 104), framing effects ‘occur when (often small) changes in the presentation of an issue or event produce (sometimes large) changes of opinion’. They note the presence of framing when politicians seek to mobilise support by selectively highlighting features of a policy, such as its likely effects or its relationship to important social values (ibid., p. 106). The power of frames to influence people’s opinions has been studied in the context of sexting, sexual grooming on the internet, revenge porn and assisted suicide; frame analysis demonstrates the strategic use of discursive devices to convince others of the correctness of seeing an issue as a problem and the importance of adopting the concurrently proposed solution (Benford and Snow, 2000).

Hulst and Yanow (2016, p. 93) argue that such analysis relies on a static conceptualisation of ‘frames-as-objects’ that stakeholders strategically deploy for political aims; instead, the concept of framing-as-process allows an exploration of how people arrive at a particular understanding of public policymaking, recognising the interactive and intersubjective processes through which frames are constructed. They identify the components of framing as sense-making, selecting, categorising and naming. These, along with the broader narrative linking these processes through storytelling about the issue, shape social issues into a problem with a potential solution. Hulst and Yanow (ibid.) draw on Erving Goffman’s (1959) earlier work, emphasising the often shifting and unconscious positions actors adopt as they communicate: where Goffman’s later work Frame Analysis: An Essay on the Organisation of the Experience (1974) focuses on game-playing, Hulst and Yanow (2016) challenge the conception of policymakers as strategic actors who consciously, intentionally and cognitively develop and alter their positions to enhance the possibility of alliance or coalition, arguing that this approach ignores the complexity of the policymaking process. While this attention to process and unexpected outcomes is important, in much of these debates about framing, there is little attention to issues relating to power inequalities and the politics behind the policy process, which forecloses some options and foregrounds others (Bacchi, 2009).

In summary, the concept of problematisation allows us to scrutinise the process that policymakers use to delineate a situation—here, SSA—as policy-relevant and to construct a favoured interpretation of its meaning (Hulst and Yanow, 2016). Through close examination of the 2014 to 2015 parliamentary debates on SSA, we explore the processes of sense-making and naming (including categorising and selecting) through which SSA has become constructed as a social problem requiring a particular policy response. Through our analysis of the policymaking process on SSA, we draw attention to ways in which dominant discourses about gender and race—in particular, protectionist racialised discourses about ‘saving brown women’ and postcolonial feminist discourses that draw attention to the socio-economic contexts of gendered oppression and disadvantage—have been deployed in this particular set of debates.

unlikely bedfellows: marshalling against ‘gendercide’

In the House of Commons, the Ten Minute Rule offers a way for backbench MPs (i.e. Private Members) to propose legislation. The process is often used to test Parliament’s opinion and raise the profile of an issue, rather than as a serious attempt to pass a bill: few bills introduced this way become law. Fiona Bruce’s Private Members’ Abortion (Sex-Selection) Bill 2014–15 states:

Sex-selective abortions are happening in the UK, and there is widespread confusion over the law, which is why this Bill is needed. The Bill is extremely straightforward, merely clarifying that nothing in section 1 of the Abortion Act 1967 allows a pregnancy to be terminated on the grounds of the sex of the unborn child. (Hansard, 2014b, column 677)

No arguments were put forward opposing the motion and only one MP voted against it on 14 November 2014; turnout was 29.4 per cent.

Given the time constraints on bills proposed under the Ten Minute Rule, arguments are usually principled and philosophical rather than detailed discussions or critiques of existing legislation and its implementation. We have identified three main arguments that were presented by Bruce as she sought to problematise SSA and suggest corresponding solutions in support of her Bill.

i) the existing law is not clear

Although abortion on the grounds of sex alone is illegal, this prohibition is not clearly articulated: the law allows for diverse interpretations by the medical practitioners tasked with implementing it. It is useful to clarify here that the ‘grounds’ on which an abortion can be authorised according to Section 1 of the Abortion Act 1967 (1967) include the most common one: that continuing the pregnancy poses a greater risk to the woman’s health than would a termination. The other grounds relate to more serious risks to the woman’s life, her mental or physical health, or a substantial risk of serious foetal abnormality. Although sex selection is not itself a ground for abortion under existing law, the sex of a foetus can have implications for one or more of the other existing grounds: for example, in cases of an identified sex-related foetal condition. Bruce argued that the sex of a foetus should not be treated as a justification for termination under any of the existing grounds set out in Section 1.

Bruce’s framing of the law as unclear on this issue rested partly on conflating legal grounds and factual explanations for abortion, which Kate Greasley (2016) argues is central to understanding the structure of Britain’s abortion law; Greasley contends that although neither pregnancy due to rape nor the sex of the foetus is an explicit ground for an abortion, both can be lawful reasons when the physical or mental health ground for abortion in Section 1 is invoked and is attested to in good faith by two medical professionals. However, both Greasley (ibid.) and Sally Sheldon (2016) argue that, while the pressure to bear a male child may indeed be part of the reason a particular pregnancy may put a woman’s physical or mental health at risk thus creating a legal ground for abortion, such coercive contexts may also invalidate her consent. The offence of assault already addresses contexts where a woman may be coerced to undergo medical procedures, including abortion. Hence, Greasley (2016) and Sheldon (2016) argue that the criminalisation of SSA to protect women who are subject to direct coercion is unnecessary.

By stating that her Bill was ‘merely clarifying’ the law, Bruce sought to minimise the significance of her proposed clarification, whilst simultaneously claiming the need for it. Thus Bruce framed her Bill as designed to enable more effective and consistent implementation of existing legislation, mitigating against concerns that it would alter the right to abortion granted by the Abortion Act 1967, an alteration pro-choice groups would refuse to back. Here, it is worth scrutinising the selection of certain aspects of the proposed policy—a selection that is a contingent, political act (Hurst and Yanow, 2016, p. 99)—while others are elided. Bruce’s selection focused on the existing legal grounds for abortion (i.e. no new restrictions are being proposed) while the categorisation of the problem frames it not as an issue with the law on abortion per se but as the ‘misinterpretation’ of its ambiguities, thus simultaneously employing naming as a framing device. This frames the ‘problem’ in terms of aspects on which policymakers are able to act.

ii) sex-selective abortions are prevalent in the UK

Eden et al. (2006) draw attention to the role of scientific boundary-work in policymaking, whereby an argument’s credibility is bolstered by claims drawing on scientific evidence. As illustrated earlier, evidence of the prevalence of SSA in the UK is contradictory and inconclusive. However, Bruce’s second argument relied on claims that this practice is widespread among the UK’s South Asian communities. She drew on the anecdotal experiences of charities working with South Asian women to argue that SSA and domestic violence related to the births of girls were common, but she did not present systematic evidence on the prevalence of SSA in the UK. This omission could reflect the format of the Ten Minute Rule, which does not provide scope for detailed scrutiny of the evidence or the inconclusive nature of the evidence. Despite the lack of conclusive evidence on the widespread prevalence of SSA in the UK and in the context of the media representation of SSA as a problem in some minority ethnic communities, Bruce instead drew upon these anecdotal accounts in a process that selectively framed them as scientific evidence.

iii) SSA is a form of VAWG in minority communities

A feminist analysis of policymaking invokes a wider set of discourses beyond the legal status of SSA, including the need to examine how specific policies on reproductive rights operate (i.e. whether they do or do not create the societal change intended). Bruce named SSA as one particular manifestation of a broader problem of VAWG in minority communities, whereby, in a context of son preference, women are denied their reproductive rights and coerced into undergoing termination of female foetuses. In the absence of systematic scientific evidence on SSA, Bruce supported her argument with statements from Jeena International and presented the case of Rupinder who ‘decided to abort her third child as she was expecting a girl’ (Hansard, 2014b, column 677). This decision was based on Rupinder’s memories of the disappointment that greeted the birth of her own sisters, and her consequent fear that ‘giving birth to a girl meant disappointment, betrayal and lowered status within the family and the community’ (ibid., columns 677–678). However, this example does not suggest Rupinder had been subject to any explicit threats or coercive expectations from her husband or his family.

Nevertheless, Bruce argued that SSA-related coercion is direct and explicit, with individual women facing pressure to abort female foetuses under the genuine threat of domestic violence: ‘I find it deplorable that anyone would be satisfied to provide a sex-selective abortion to a woman who, after she has had it, is then sent back to an abusive partner. What needs to be addressed in those dire circumstances is the abuse itself. Those women need help, and that is one aim of the Bill’ (ibid., column 677). When examined from a perspective that centres women’s autonomy, this aim was clearly questionable because the ‘help’ provided by the Bill entailed denying permission to abort when women are coerced into it rather than addressing the circumstances of the abuse itself.

This framing draws attention to a set of circumstances—explicit and direct coercion—within which most pro-choice feminists would problematise abortion. Here the process of selection and categorisation can be clearly observed. In tandem with the selective focus on the ambiguities of existing legislation and absence of statistical evidence on the prevalence of SSA in the UK, Bruce selected a set of accounts from a few individuals and agencies to construct a story that categorised SSA as a social problem. Framing and problematising SSA as a coercive mechanism imposed on South Asian women whose agency is constrained by an abusive culture, however, has the impact of placing greater scrutiny on abortion decisions made by women in minority communities as a means of curbing SSA. Thus, we can observe what Hurst and Yanow (2016, p. 98) see as the role of sense-making work in enabling a normative leap from what ‘is’ to what ‘ought to be’.

Through this strategy, Parliament was presented with the prospect of protecting not only minority women’s rights by preventing coerced abortions but also the right to life of the girls who would otherwise be aborted. Paradoxically, restricting women’s reproductive rights was proposed as a means of achieving greater gender equality in communities where this is not seen as part of their socio-cultural heritage. This can be understood as an example of what feminist postcolonial critiques would describe as an impulse to ‘save brown women from brown men’ (Spivak, 1988, p. 93). One of the most trenchant insights of postcolonial feminist theory has been that homogenised depictions of Third World women essentialise the Third World as if it were a singular locale. Such homogenisation slips all too easily into the exoticisation of the foreign ‘Other’, tinged with ethnocentric pity for the condition of ‘their’ forever oppressed women. This reads in a manner similar to colonial texts proclaiming the need to save and protect the oppressed from themselves (Mani, 1998). For example, Gayatri Chakravorty Spivak (1988, pp. 285–287) provides a rigorous critique of the paternalistic feminism that informed the British colonial ban on sati in India. In British accounts of the practice in which a widow would immolate herself on her husband’s funeral pyre, the voice of women who practised sati was always absent (ibid.). Postcolonial feminists writing on migration and identity in diasporic contexts have offered similar critiques of the disabling paradigms within which South Asian women are constructed, including the hyper-visibility of (im)migrant women in recent media and policy debates that emphasise the ‘problematic private sphere’ and religio-cultural basis of violence against minority ethnic women (Ahmad, 2003; Puwar and Raghuram, 2003).

Understanding the processes of sense-making at work in the parliamentary debates on SSA entails unpacking the assumptions and presuppositions that inform this representation of the problem, as well as questioning the silences therein—namely, what is left unproblematised (Bacchi, 2009). The particular assumptions and presuppositions revolve around the construction of South Asian women as a homogeneous category who, according to Fauszia Ahmad (2003, p. 43), are predominantly passive and wholly determined by their (repressive) culture: in other words, they are ‘essentialised oppressed figures of victimhood and despair, in need of rescuing from their men’. Ahmad argues that such representations are limited in scope and offer little possibility of recognising women’s struggles and processes of empowerment through social and/or political agency. Postcolonial feminists (Spivak, 1988; Mohanty, 1991) have long been critical of the universalising tendencies of what Chandra Talpade Mohanty (1991) has called ‘white feminism’, which has sought ‘to represent women from the Global South as being in need of liberation, not in terms of their own herstory and needs but into the “progressive” social mores and customs of the metropolitan West’ (Carby, 1982, p. 216).

Bacchi’s (2009) call to examine what is rendered unproblematic in any representation of a problem reveals that this framing of SSA is premised on a silence regarding the coercive contexts that underpin many of the reproductive decisions all women make regarding pregnancy. The silent assumption is that abortion decisions in white communities are free from coercive socio-economic and cultural constraints and, thus, these women can make free choices, unmediated by their social context. Through a process of selection, Bruce’s Bill framed a particular set of socio-economic contexts within minority communities as coercive, eliding a variety of other constraints in both minority and majority communities: coercion by a partner or parent, gendered power imbalances, poverty, the disproportionate impact of austerity measures on single mothers, and social expectations casting mothers as primary carers. It is within these constraints that all women make decisions to continue with or terminate pregnancies. This rhetoric of choice frames white women as free agents when making reproductive decisions and so differentiates them from their British Asian counterparts. Instead of suggesting a response in the form of bolstering consent procedures, Bruce advocated restricting abortion rights for the latter.

Towards the end of her speech, Bruce stated, ‘We can no longer ignore the fact that sex-selective abortion is a reality in the UK. Lest anyone think that this is an issue that applies only in certain communities, they should consider the tragic fact that the words “family balancing” are heard with increasing frequency and understanding across the country’ (Hansard, 2014b, column 679). The argument around family balancing, however, was peripheral in terms of the framing of the Bill; Bruce did not refer to it again, nor was it mentioned in media reports (see, for instance, Watt et al., 2012; Connor, 2014). This silence on a possible wider understanding of the ‘problem’ can be read as a attempt to avoid muddying her project of sense-making: to follow through with this argument would have undermined the project of problematising SSA as a manifestation of VAWG in only certain communities. As Sheelagh McGuiness (2013) argues in the context of disability as a ground for abortion, the distinction between ‘deserving’ and ‘undeserving’ abortion is problematic in that it is premised on the eugenic ideology of valuing people differently on the basis of their (dis)ability. In countries with high levels of SSA of female foetuses, similar arguments have been made in relation to the perceived lower worth of such foetuses. However, ‘family balancing’ cases do not involve an inherent devaluation of a foetus on the basis of its sex, but are determined by the context of the sex of existing children. Thus, by focusing on ‘protecting vulnerable women’ within minority communities, the Bill categorised some abortions as problematic because they constituted a ‘discriminatory practice and the first and most fundamental form of violence against women and girls’ (Hansard, 2014b, column 679) in particular communities. This entailed naming SSA within minority communities as a problem with resonance for feminists, in particular, and policymakers, more broadly.

In articulating her case for restricting SSA, Bruce reconstructed the pro-life position as a pro-woman one, arguing that restricting minority women’s rights to abortion constitutes a defence of their rights. Rahila Gupta (2014) draws attention to critiques of this position by transnational feminists, arguing for the need to take account of the nature of coercion for poor and marginalised women in low-income countries and where such women often face pressure to abort as a form of birth control. In seeming to adopt a similar pro-woman discourse, Bruce distanced her framing of SSA from pro-life discourses that set the personhood of the foetus against women’s reproductive rights: a common theme in recent anti-abortion rhetoric that utilises women-protective discourses to frame their pro-life positions by co-opting feminist language of ‘women’s rights’ and ‘choice’ (Siegel, 2014 [2012]; Jesudason and Weitz, 2015). Thus, an argument apparently promoting the autonomy of minority women invokes socio-economic analysis to suggest that their autonomy is invalidated by coercion and, as such, the available choices must be limited for their own protection, thereby reducing their reproductive choices rather than advancing them.

Bruce invoked the issue of violence against (existing) women and (future) girls when she cast intervention as a step towards achieving gender equality—the traditional terrain of feminist and pro-choice groups. She argued that ‘This month, for the first time, the UK has dropped out of the gender equality top 20. It is a further damning indictment of our commitment to female parity that we allow national institutions to contradict the Government on an illegal practice that predominantly affects girls’ (Hansard, 2014b, column 679). By suggesting that the UK’s falling position on the gender equality table was linked to its failure to curb SSA, Bruce co-opted the discourse of gender equality—a discourse that has long been used in defence of women’s reproductive rights—to make a case for restricting such rights.

Thus, Bruce’s sense-making and naming drew together features of an intractable policy situation, rendering them more coherent and graspable, while also diverting attention from their ambiguities and uncertainties. This process entailed selection from a range of possible features warranting attention: explicit and direct coercion, but silence on how this is already criminalised; indirect coercion such as the presence of coercive cultural contexts in minority communities (primarily the devaluation of women and son preferences), but not coercive socio-economic contexts in majority communities; naming the issue as the abortion of female foetus by South Asian women, but largely ignoring possible abortion based on sex by women of other ethnicities for ‘family balancing’ reasons; and the categorisation of the problem as a form of VAWG. Through such processes, policy actors ‘draw disparate elements together in a pattern, selecting some things as relevant or important and discarding, backgrounding or ignoring others, occluding other ways of seeing (and acting), and thereby silencing them in policy discourse and ensuing action’ (Hulst and Yanow, 2016, p. 99).

The arguments presented in support of the Bill help explain the cross-party support for it in a context where parliamentary debates on abortion had hitherto been sharply polarised (Weale et al., 2012). The campaign to end ‘gendercide’ resulted in an unexpected alliance between self-proclaimed pro-choice organisations, such as Karma Nirvana and Jeena International, and pro-life organisations, whose motivations and ideological stances they had previously opposed. These disparate organisations and groupings perceived a common cause because of Bruce’s problematisation of SSA as a symptom of patriarchy within particular minority communities.

feminist arguments on bodily autonomy in abortion debates

Bruce’s attempt to frame concerns about SSA in terms that could appeal to both pro-life and pro-choice groups can be seen as symptomatic of the complexities of the discourse around abortion and reproductive rights. SSA in particular remains a divisive issue for feminists because it raises complex issues of ethics and agency in relation to women’s control over their bodies. It also poses a challenge to feminist support for a woman’s right to bodily self-determination, as its practice in a context where male births are viewed as more desirable amounts to discrimination against women. Nivedita Menon (2012) argues that this dilemma arises because feminists seem to be counterposing the rights of (future) women to be born against the rights of (present) women to exercise control over their bodies.

However, a feminist perspective requires a recognition and critique of broader socio-economic and cultural factors. These include the relationship between gender and poverty, young and poor women’s lack of access to contraception, the coercive nature of sexual interactions that prevent contraceptive use and/or access, the stigma and material consequences of illegitimacy in particular countries, and a lack of facilities for childcare that places a disproportionate burden on women. Criminalising SSA constructs a particular set of limitations—related to the culture of son preference—as eroding choice, while ignoring other, normalised social constructs that are, thus, rendered invisible. This presents a fundamental challenge for feminists: to reshape the socio-economic culture that informs son preference and daughter aversion rather than simply advocating for measures clarifying the illegality of SSA. While access to abortion does not resolve any of the underlying issues, the prevailing framing of abortion politics shifts attention away from these problems and, thus, renders them less likely to be the subject of measures to address them.

For many non-white, non-middle-class women who might otherwise identify with the mainstream feminist movement, access to prenatal care, contraceptive services and freedom from coerced sterilisation may be much more pressing reproductive issues in relation to autonomy than abortion. However, governments often only reveal a serious interest in women’s health when they happen to be pregnant, highlighting the lack of value attached to women’s bodies and their health in other circumstances. Catherine MacKinnon (1989) articulates what is at stake in the defence of reproductive rights; she believes that it is the relationship between abortion and oppressive sexual relations that makes speaking of abortion in feminist terms both risky and difficult (see also Sheldon, 1997). Addressing abortion in the broader context of its meaning in women’s lives requires us to frame it not simply as a medical issue but also as one of sexuality and reproduction. Taking this approach means recognising that no amount of neutral sex education, medical information or government involvement will change the desire for abortion so long as it remains that woman cannot exercise control over how they choose to enter into sexual and reproductive relations with others. Indeed, even if women could exercise such control, the desire for abortion would remain, because contraceptives can fail and circumstances and minds can change.

in defence of women’s reproductive rights: the discourse shifts

Following near unanimous support for her Bill, Bruce proposed an amendment to Section 65, Part 5 of the Serious Crime Bill 2014–15, arguing that the ‘New Clause 1’ should state that ‘nothing in section 1 of the Abortion Act 1967 is to be interpreted as allowing a pregnancy to be terminated on the grounds of the sex of the unborn child’ (Hansard, 2014b, column 677). However, Ann Coffey MP, Sarah Wollaston MP and Jenny Willott MP tabled a counter-amendment (New Clause 25), requiring the collection and review of evidence on the issue and calling for a strategic plan to address concerns about the prevalence of SSA in England, Scotland and Wales.

The arguments supporting Bruce’s proposed amendment reiterated those presented in defence of the Ten Minute Rule Bill, gaining similar support from Bruce’s fellow Conservative MPs as well as sympathetic submissions from organisations that had supported the earlier Bill. However, a number of organisations and experts (including the Royal College of Midwives, the Royal College of Obstetrics and Gynaecologists, the British Medical Association, the Trades Union Congress, the End Violence Against Women Coalition, Genetic Alliance, IMKAAN and Southall Black Sisters) highlighted concerns about the disproportionate impact of austerity measures on black and minority ethnic women, including the closure of refuges, counselling and support services, and the broader impact of the withdrawal of the legal aid support needed to access protection and justice through the courts.

Southall Black Sisters (SBS) (2015),Footnote 3 for example, argued that rather than focusing on services that might support women in leaving violent relationships or empower them in exercising meaningful choices in their lives, Bruce’s amendment represented yet another instance of the government’s ‘resource neutral’ solutions aimed at (ostensibly) protecting black and minority women. These organisations, along with other groups representing health practitioners, argued that restricting SSA was unnecessary and could be detrimental to the women it sought to help.Footnote 4 These contributions were crucial in shaping the course of the debates on the proposed amendments.

Unlike the debate on Bruce’s previous Bill, the new debate was polarised along party lines, following a Labour whip urging members to vote against the amendment. On 23 February 2015, it was defeated by 292 votes to 201. Later that evening, Parliament voted in support of the counter-amendment (New Clause 25) by 491 votes to 2. Following a brief mention that abortion on the grounds of sex was already illegal, the first set of arguments made by opponents centred broadly on examining the paucity of evidence that SSA is practised in the UK. In response, Bruce quoted Rani Bilku (of Jeena International): ‘“I have been supporting women dealing with sex-selective abortions … for almost a decade. Saying that there is no evidence is tantamount to saying that the women we work with are lying”’ (Hansard, 2015, column 116). Opponents rejected this statement, citing an analysis by the DOH (2013). In response, Fiona Mactaggart MP stated, ‘I feel that I have been pulled along by a Trojan horse … I am therefore concerned that we are using anecdote from an unreliable source to make legislation on the hoof’ (Hansard, 2015, column 121).

The second set of arguments against the Bill framed the proposed amendment as an attempt to restrict women’s reproductive rights. The ideological basis of the problematisation of SSA was explicitly discussed, with the amendment now framed as ‘pro-life’ rather than ‘pro-woman’. Sarah Wollaston MP raised the implications of the use of the term ‘unborn child’ in the proposed amendment, arguing that it confers personhood on the foetus in a way that would change the meaning of the term in the Abortion Act 1967 (ibid., column 114), and which can be understood as an attempt to secure a legal definition of pregnancy that pits the ‘rights of the unborn child’—independent of the pregnant woman—against the reproductive rights of women. Reinforcing her unpicking of the ideological basis of the amendment, Wollaston, who originally voted for the Ten Minute Rule Bill, quoted the head of a US-based organisation supporting the All Party Pro-Life Group chaired by Bruce as saying, ‘“By formally protecting all female foetuses from abortion on the ground of their sex, we would plant in the law the proposition that the developing child is a being whose claims on us should not depend on their sex [...] the right to abortion is fundamental to women’s emancipation, but many will recoil at the thought of aborting their unborn sisters”’ (Hansard, 2015, column 130).

This unmasks the strategic game-playing that regularly happens around various notions of ‘problematic abortions’ based on disability, the age of the foetus and gender, which distract from the question of women’s rights to abortion. The recent widening of the ambit of the discourse on ‘problematic abortions’ to cover sex is part of a well-rehearsed strategy pro-life groups have used since the passing of the Abortion Act 1967. In 1990, the limit was reduced to 24 weeks, with further parliamentary motions unsuccessfully attempting to bring it down to 20 weeks (BBC News, 2008).

By drawing attention to the political and ideological character of the construction of SSA as a social problem, the second set of debates unmasked the processes whereby the problematisation of SSA as violence against minority ethnic women was constructed and, in so doing, made the framing explicit. Whereas the assumptions underlying the Ten Minute Rule Bill were implicit, the unmasking of these assumptions and their framing in connection with the proposed amendment to Section 1 of the Abortion Act 1967 played a crucial role in disrupting the previous consensus.

Ann Coffey MP raised another issue relating to the counter-amendment: the unintended consequences of a bill that would require particular scrutiny of South Asian women’s abortion decisions might risk a return ‘to the days of the botched backstreet abortions that […] have been the resort of desperate women’ (Hansard, 2015, column 119). Thus, she recast restricting access to abortion for particular categories of women as a source of potential legislative harm ‘for the very women it purports to protect’ (ibid.). This discourse was indeed a shift from the earlier debates framing the restriction of abortion for South Asian women as a woman-centred defence of reproductive rights (to have children) as well as the right of future female children to exist.

Bruce argued that her proposed law would not only signal the unacceptability of SSA but also would enable women to use the ‘legislation as a bargaining tool to negotiate … so a young woman could say, “You do realise this is a criminal offence?”’ (ibid., column 116). Coffey countered this argument made by Bruce, drawing attention to the complex nature of coercion for pregnant women in ‘very different circumstances’ who are ‘subject to different pressures—economic, familial and community—that can all influence a pregnant woman’s state of mind and her attitude to continuing her pregnancy’ (ibid.). Basing their arguments on both a liberal feminist defence of abortion rights and conceptions of autonomy, as well as more nuanced postcolonial and transnational feminist approaches that informed the interventions of organisations like SBS, the opponents of the Bruce amendment advocated the need to locate protection against SSA within a safeguarding framework that was focused on the available support services (SBS, 2015).

Echoing the position articulated by groups such as SBS, the positions of the opponents of the amendment seem to invoke both a critique of Western radical feminist modes that fixate on ‘saving’ non-white women from their men as well as of Western liberal feminist perspectives that prioritise autonomy over socio-economic analysis of the contexts that curtail freedom or choice.

conclusion

By singling out South Asian women and their reproductive practices for state intervention, Bruce’s proposed explicit criminalisation of SSA sought a ‘quick fix’ aimed at ending some forms of coercion in abortion decisions by curtailing abortion rights for particular categories of women. The initial support for the Bill indicates the continuing traction for protectionist responses to VAWG in minority communities. This paper demonstrates that the construction of SSA as VAWG draws on racialised discourses on reproductive rights, whereby a focus on abortion to the exclusion of other reproductive issues can be seen to benefit white women more than minority women: the protectionist turn in these discourses in relation to SSA reveals a selective concern for socio-cultural contexts and constraints on minority women, ignoring the factors that affect all women in favour of training a spotlight on minority communities. Examining this framing demonstrates the efficacy of deploying feminist language for non-feminist purposes. When presented with the opportunity to ‘save brown women from brown men’, parliamentarians who would otherwise frame their position as (liberal) feminist united in order to give Bruce’s original Bill overwhelming cross-party support. In contrast to this stands the position articulated by transnational, postcolonial feminists whose politics seek solidarity between women based on a recognition of the multiple, overlapping and discrete forms of oppression that create continuity between gendered coercive contexts across socio-cultural groups and nations, while remaining cognisant of the fact that they may appear in specific forms within particular communities.

Alongside highlighting the lack of statistical evidence that SSA is a problem in the UK, either generally or in minority communities, the second set of debates on the Bill explored the fact that criminalisation was not likely to make women safer and that safeguarding approaches, recognising the socio-economic factors impacting all pregnant women, were a more appropriate framework within which to legislate. These arguments were framed within a critique of the attempt to rename ‘the foetus’ as ‘the unborn child’ in an attempt to revitalise existing discourses on ‘problematic abortions’ and locate the attempt to criminalise SSA within a broader pro-life—rather than pro-woman—protectionist perspective. The contributions of expert communities in these debates played a crucial role in bringing about a shift in the problematisation of SSA, indicating the reach, capacity and success of women’s organisations in framing the legislative agenda and making their voices heard in Parliament. This highlights the potential for more open participation in policymaking to enhance the democratic process (Brazier et al., 2008), not only in terms of informing parliamentary debate but also of contributing to public discussions.

Paying close attention to the sense-making embedded in framing devices and how stories about problematised social issues are told in policy-oriented narratives alerts us to the political dynamics of framing processes. Our analysis of the debates on SSA contributes to existing debates on policy analysis by drawing attention to the importance of ideology—in the form of both implicit discourses on race and gender as well as more explicitly articulated political positions on abortion rights—to the construction and deployment of frames. A close examination of the text of the parliamentary debates on SSA illuminates how selections (and omissions) of issues and of emphasis are made, how naming is used strategically, and how categories are created. This helps to unpack the strategic deployment of frames and also the negotiations that shape the dynamic process of sense-making within these debates. It is our contention that these processes opened up new ideological alignments and alliances and a shift in the framing of SSA.

Our analysis enables a deeper understanding of both the relevant issue (SSA) and the politics of the framing processes. The first contribution of this study on the problematisation of SSA draws attention to the framing of social issues as an important political process in itself that produces ideological and material consequences; our analysis thus makes politics visible in the process of policy analysis. The second conceptual contribution of this paper is to transnational feminist debates on gender and race. Through a feminist analysis of law and policymaking, we demonstrate how media and policy debates on abortion, including on SSA, remain amenable to culturally essentialist discourses that prevail in particular Western feminist perspectives. The initial success of Bruce’s attempt to co-opt feminist language where the focus is on ‘other’ women illustrates how British abortion discourses have been insufficiently attentive to its racialised dimensions.