Constructing the gendered body? A critical discourse analysis of gender equality schemes in the health sector in England

Introduction

Gender equality strategies are not new. In recent decades an increasing number of institutions, across national, regional and international scales, have introduced principles of gender mainstreaming to their work. Gender mainstreaming, defined in 1997 as ‘the process of assessing the implications for women and men of any planned action, including legislation, policies or programmes, in all areas and at all levels’ (UN, 1997: 27), has been adopted by the United Nations, the World Bank, World Health Organisation and the European Union. At national level various countries, including Denmark, Canada and England, have introduced policies to address gender inequality (Kuhlmann, 2009). However, there is mixed evidence of the success of such policies, and critical analyses of the contribution of gender mainstreaming to gender equality and gender equity suggest progress has been slow (Daly, 2005; Mazey, 2002), particularly in relation to the health gap between women and men (Abdool et al., 2010; Briones-Vozmedian et al., 2012; Kuhlmann and Annandale, 2012).

Gender mainstreaming is seen by many as having developed into a technocratic exercise, often based on the administration of gender tools, leaving behind the original participatory ideal (Gideon, 2012; Squires, 2010) and leaving gender relations of power unchallenged (Sen and Östlin, 2008). In addition, the formal adoption of gender mainstreaming and the introduction of legislation can also mean that policy makers consider that the issue has been addressed and gender justice disappears from the agenda (Kenney, 2003). However, gender mainstreaming can also be seen as contributing to gender discourse by framing the problem in ways which further limit progress. Institutional approaches to gender structure social life through organisational practice, and even when there is a formal commitment to gender equality, gender discourse is embedded in cultural logic, resource distribution and ‘ideological andocentrism’ (Risman and Davis, 2013). ‘Gender mainstreaming’, ‘gender analysis’ and ‘gender equality’ are discursively constructed forms of social knowledge which produce political subjects, including those of ‘men’ and ‘women’. Equality policies are part of this productive process, in the representation of the problem of gender variations and gender inequality, and the solutions proposed (Bacchi, 2000; Bacchi and Eveline, 2010).

Analyses of gender mainstreaming which are framed in a realist perspective highlight the problems posed by a lack of resources, data and political will but are constrained in their ability to explain the persistence and resilience of the barriers they identify. This article presents a critical discourse analysis of gender equality strategies in the health sector in England, as a way of adding to our understanding of gender mainstreaming both in health policy and more broadly, and why gender mainstreaming as a paradigm has achieved only moderate success in practice. The article offers an alternative approach which focuses on how the problem of gender equality has been represented in policy making, and the implications of this representation for gender justice.

Theoretical framework: ‘What is the problem represented to be?’

Existing analyses of gender equality strategies in the context of health are limited by the way in which the problematisation of gender inequality in policy making is taken for granted. Traditional policy analysis adopts a rationalist approach which aims to uncover the best solutions to given problems, and in which the representation of the problem itself is unquestioned (Goodwin, 2012). While more recent critical perspectives in policy analysis have explored underlying political frameworks and the presumed neutrality of both policy problems and associated policy making, the characterisation of the problem itself is not explored. Newer anti-positivist, interpretive paradigms may have more to offer in relation to our understandings of problem faced by gender mainstreaming, and the limited success of gender equality strategies. Such approaches highlight the discursive character of policy and policy making and open up for scrutiny the ways in which policy discourses embody implicit and unspoken assumptions about the world (Fairclough, 2003). The systematic exploration of relationships between discursive practice, the production, distribution and consumption of text, and wider social relations, drawing on Foucauldian notions of power as embedded within social practices and as productive of social relations, helps to identify the ways in which such practices and texts are both shaped by, and act to maintain, power relations (Fairclough, 1993), including gender relations of power.

The work of Carol Bacchi builds on the idea that policy problems are socially constructed, arising at specific times and in specific policy spaces (Goodwin, 2012), particularly in her development of the ‘what is the problem represented to be’ (WPR) approach to policy analysis (Bacchi, 1999): ‘lives are lived in specific ways due to the shaping impact of proposals that create particular understandings of “problems” ’ (1999: 22, emphasis in original). The WPR approach to policy analysis aims to uncover normative assumptions embedded in representations of policy problems, and is particularly valuable in an appraisal of policy making and policy agendas where the problem itself seems obvious, and unproblematic. The WPR framework allows us to interrogate the taken-for-granted premises embedded in gender equality policies – such as the meaning of gender equality, the goals of gender mainstreaming in a health context, and the interpretation of significant gender differences in health. Gender equality strategies and their implementation are examples of ‘knowledge’ transfer where implicit and embedded ‘knowledges’ are not questioned within policy making, but are instead repeated and reiterated by the policies they produce.

Bacchi (2012) outlines six questions concerning the representation of ‘problems’ within policy making and policy proposals, which form the framework of the WPR approach to policy analysis: (1) What is the ‘problem’ represented to be in a specific policy? (2) What presuppositions or assumptions underpin this representation of the ‘problem’? (3) How has this representation of the ‘problem’ come about? (4) What is left unproblematic in this representation, that is, where are the silences? (5) What effects are produced by this representation of the ‘problem’? and (6) How and where has this representation of the ‘problem’ been produced, disseminated and defended, and how could it be disrupted?

These questions do more than simply identify the importance of the ways in which policies represent the problem they are aiming to address. The WPR framework also aims to elucidate implicit or unexplored assumptions, as well as what has been silenced by policy, while locating this exploration within the context of power relations and the discursive impact of policy’s representation of the problem.

This allows us to identify the ways in which policies produce not only problems but also the subjects of those policy solutions. However, it is important to note that, for Bacchi, representations of the ‘problem’ in policy are not the result of deliberate manipulation by policy makers, and the aim of WPR analysis is not to find the ‘real’ problem and the ‘right’ solution, but rather to interrogate how representations come about, and how these shape solutions and subjectivities (Bacchi, 2012).

In the context of gender mainstreaming and gender equalities such an approach allows us to go beyond documentation of barriers to gender justice, to explore the ways in which gender policies unwittingly recreate and reinforce the equalities they purport to address through their representation of the problem. It also allows a different interpretation of the failure of gender mainstreaming to challenge existing gender relations of power (Sen and Östlin, 2008). In this approach, representations of differences between women and men in health risks or health behaviours, for example, are not seen as strategic on the part of (male) vested interests but are to be understood in the context of discursive constructions of gender, gender equality and gender ‘knowledges’ which are embedded in policies, and in policy implementation.

This is not to argue that male vested interests and gender relations of power do not play a part in the representation of problems, and embedded assumptions, but rather that we need to be alert to how representations come about, what knowledges are seen as valued within this process, and which are silenced, and how these are defended and disseminated.

Gender variations in health

This article focuses on policy to address gender variations in health, and this section offers a brief background to key health differences between women and men. In the majority of countries women live longer than men, although the size of this advantage in life expectancy varies (WHO, 2012), and men are more likely than women to die from some causes, especially ischaemic heart disease, non-reproductive cancers and accidental and non-accidental injury (WHO, 2012). Data on morbidity show further variations: overall women appear to be more likely than men to experience poor health, but this differs according to location, age and stage of life course, how data are collected (whether they are based on use of services or self-report measures for example) and across specific diseases (Bird and Rieker, 2008; Payne, 2006). These variations reflect both biology, including genetic, reproductive and hormonal variations, and gender-linked factors, including access to resources, paid and unpaid work, health behaviours and health risks resulting from the behaviour of others (Bird and Rieker, 2008; Wizeman and Pardue, 2001). In particular, there is often no simple pattern of health advantage for either men or women when we look at the combination of ill-health, disability and mortality.

Health system influences, including how well and how appropriately services meet the needs of both men and women, also play a part. Financial resources are a significant constraint: health systems which rely on user fees limit access by poorer groups, and women are disproportionately affected by this (Nanda, 2002). Insurance-based systems also create barriers for some groups, especially women, who are less often in employment with health insurance, and are less able to afford private insurance (Wagner et al., 2013). Further barriers to service use arise from cultural factors which may prevent women from travelling alone or from consulting male practitioners (Majaj et al., 2013). Similarly the location of clinics, opening times and transport limitations can affect access (Tadesse et al., 2013).

Evidence also suggests that men under-consult for symptoms of illness, and present late, due to discourses of masculinity which mean men find it harder to admit to pain or anxiety over health symptoms (Courtenay, 2011; Wilkins et al., 2008), although more research on gender, health beliefs and consultation is needed (Annandale et al., 2007).

Gender mainstreaming in health

Analyses of gender mainstreaming strategies in a health context are generally based within a realist framework and aim to measure changes in, for example, health outcomes, diagnoses or service use by women and men. Such studies suggest that while gender mainstreaming strategies have been in place in international and national settings for some time, there is little evidence that these have led to significant changes in the health variations between women and men.

For example, the recent report by the UN Commission on the Status of Women on progress in gender mainstreaming since the adoption of the Platform for Action in 1995 found that many countries had introduced gender equality strategies (UN, 2010) and gender mainstreaming had helped to identify the health needs of women, and improve women’s access to health services and gender training for health professionals. However a number of barriers were also identified, including a lack of gender-disaggregated data, low skill levels and inadequate financial resources. In addition, gender differences in access to basic health services have worsened in many parts of the world as a result of neoliberal health reforms which have had a disproportionate impact on women (Gideon, 2012; UN, 2010).

A review of gender mainstreaming for the World Health Organisation’s Women and Gender Equity Knowledge Network (Ravindran and Kelkar-Khambete, 2007) also reported evidence of improvements in gender-based knowledge but gender mainstreaming in non-governmental health organisations was often limited to education and training rather than clinical treatment, and interventions accommodated difference rather than challenging gender relations of power. In addition, despite attempts to address the exclusion of women in medical research these have been less successful than might be hoped (Geller et al., 2011).

National studies of health systems also suggest strategies to address gender inequalities have made limited progress. Sweden, for example, has a long history of attention to gender in policy making, but research there highlights problems for women in respect of equality in service delivery and satisfaction with interventions (Jonsson et al., 2006), and a study of equality strategies introduced by health organisations in England also reported weaknesses in the ways gender had been addressed (EHRC, 2011).

Similarly, institutional gender mainstreaming within health organisations has been slow to produce results. A recent internal review by the World Health Organisation concluded that, despite a commitment to gender mainstreaming, fewer than a quarter of WHO publications used gender-disaggregated data, and fewer than one-third of public speeches by WHO senior management referred to gender. WHO concluded that ‘WHO has implemented a far-reaching gender mainstreaming program, but the impact on day-to-day work has been limited’ (WHO, 2011: 2).

However, while this literature suggests that international and national health policies have found it difficult to address gender variations in health, these analyses also frame the limitations of gender equality policy making within an objectivist realist position in which the representation of the problem by policy is treated as unproblematic and is unexamined. If instead we question the representation of the problem in policy solutions, then other conclusions emerge. The country-level reports on women’s health initiatives in the UN Commission’s review, for example, represent the problem as a need to address women, and especially women’s reproductive health. Such a representation constructs women as the problem to be addressed, rather than gender or gender relations of power, which remain obscured, unexamined and taken for granted, and also implicitly reinforces the significance of biological difference. Other reviews which highlight the failure of gender equality strategies to deliver similarly fail to question the underlying assumptions of policies which represent the problem as one of lack of data, or training and again the definition and assumed causes of gender inequality embedded in these policies remain unexplored.

Gender equality policies in England

This article focuses on gender equality policies in England. The period of the 1997–2010 Labour government has been described as a ‘window of opportunity’ (Annesley et al., 2010) for gendered policy making, reflecting both wider influences, including European Union directives, and the prominence of feminist activism within the internal political agenda. As such it offers a particularly interesting example of the development of gender equality policies across different sectors and policy scales, and the representation of policy problems this involves.

Health organisations in Britain have been required to address gender inequalities since 2007 following the 2006 Equalities Act and the introduction of the Gender Equality Duty which required all public sector bodies to eliminate gender discrimination and promote equality of opportunity between women and men. Under the legislation all health organisations in receipt of public funding were required to draw up and publish Gender Equality Schemes. The 2006 Act was followed by the 2010 Equalities Act which consolidated previous legislation and broadened the requirements for public sector organisations to address equality in respect of nine ‘protected characteristics’: gender, ethnicity, age, disability, transgender, sexual orientation, religion, pregnancy and maternity, and marriage and civil partnership. The ‘general equality duty’ created by the 2010 Act applied to England, Scotland and Wales, ¹ while specific duties are created by secondary legislation and vary between countries. The specific duties applying to England under the 2010 Act, for example, changed the requirements for transparency, allowing health organisations to determine for themselves how they would adhere to the duty and publish their strategies, and adopted the idea of proportionality in relation to such requirements.

This article focuses on health organisations in England in the period following the 2006 Act, during which the 300+ bodies which made up the publicly funded health sector at the time published schemes which outlined how they proposed to address the Gender Equality Duty.

The public sector duty in practice is implemented by the different organisations which make up the NHS in England, including Primary Care Trusts (PCTs) (now replaced by Clinical Commissioning Groups), other NHS trusts and NHS England. Following the 2010 Act example the Department of Health agreed a Framework for Action with the Equalities and Human Rights Commission (EHRC), the independent body which oversees the equalities legislation. In 2009 the NHS Management Board created the Equality and Diversity Council, chaired by the Chief Executive of NHS England, to lead on the introduction of an Equality Delivery System (EDS), a voluntary ‘tool kit’ to enable health organisations to meet the public sector requirement. Gender equality is therefore one of many equalities which now need to be addressed in health policy, and implementation has followed technocratic solutions, including gender tools such as Equality Impact Assessments, rather than ‘participatory’ engagement with activists and grassroots organisations.

Method

This article uses a critical discourse analysis of documentation on gender equality strategies in the health sector in England, focusing on Primary Care Trusts (PCTs), the health organisations which until April 2013 held responsibility for commissioning health services from various providers on behalf of their population. Under the legislation PCTs were required to comply with the equalities duties and develop systems of ensuring that those they commission services from were acting in accordance with the requirements of the Act, which placed PCTs in an interesting position in terms of their surveillance of health providers and the use of contracting processes to deliver equality objectives.

Fifteen PCTs were purposively sampled out of the 150 in existence at the time (September 2011) to ensure the sample included rural, urban and inner city PCTs, various geographical locations and different population demographics. All of the PCTs equality schemes and supporting documents which were publicly available at the time were identified for each PCT and included in this study. In addition the Department of Health’s website was scrutinised for relevant documentation, and the DH guidance for NHS organisations on how to prepare a gender equality scheme (DH, 2007) was selected for inclusion due to the role this text plays for PCTs when they develop their equality schemes. The article therefore focuses on policy statements, and although the implementation of such policies would be of interest, given the part played by equalities officers in interpreting such texts, this is beyond the scope of this study.

The analysis is based on a thematic review of the equality schemes, DH guidance and associated documents which drew on Bacchi’s WPR framework, particularly three of Bacchi’s questions for policy analysis: the representation of the ‘problem’, the assumptions underpinning this representation, and what is left unproblematic or silent in this representation. This framework was used initially to identify key search terms within the texts. Texts were read for key terms including sex, gender, gender (in)equality/ies, men, women, descriptions of health and health influences and differences between men and women, and discussion of the specific aspects of the gender equality duty. Following the first reading a number of discourses around gender equality and the policy duty, were identified. The initial review was then followed by a second close reading to develop further understanding of these discourses across the documents. In addition, the reading aimed to identify gaps or ‘silences’ in the texts concerning the representation of the problem, drawing on other representations of health variations between women and men, such as those found in medical, epidemiological and feminist literature.

Overview of the schemes

The PCTs varied but there were a number of similarities in their approach. Rather than writing separate schemes for gender and other equalities, the PCTs all developed Single Equality Schemes (SES) to address all of the equalities strands (race, gender and disability) under the 2006 legislation. In addition, all of the schemes identified at least one other form of inequality in their equalities plans, including areas also subject to legislation (sexual orientation in employment for example) and aspects of diversity and inequality which are outside legislation, notably deprivation and social exclusion.

The schemes covered similar time frames, designed to run for three years between 2008/2009 and 2011/2012, and were written for an audience of senior management and equalities officers within the organisation, as well as the general public. They varied in length between 8 and 68 pages, with an average of 37 pages. Within the Single Equality Schemes and other documents there were some variations in how gender was addressed, although for all documents the most frequent reference to gender was as part of general statements which listed all of the equalities groups being addressed. There were minor variations in the proportion of the overall document which referred specifically to gender alone: for the majority the gender section comprised fewer than two pages, while one PCT included 3.5 pages out of 51. One PCT document included only one paragraph on sex/gender, which simply referred to the proportion of men and women in the PCT area.

The next section explores discourses around gender and gender equality embedded in these documents, and their implications for broader discourses on gender and health, and health variations between men and women.

Men, women and difference

One of the most significant discourses across academic research, epidemiology and policy literature is that of the binary ‘gendered body’ (Connell, 2012). Annandale and Hammarström (2011), in an analysis of the rise of gender-specific medicine in the last decade, highlight the binary narrative of the gender-specific body, which is ‘brimming with difference’, and which aligns with broader cultural stereotypes about women’s and men’s health. Gendered discourses in the PCT equality schemes documentation and the DH guidance on gender equality reflect and repeat this story (DH, 2007). DH guidance, for example, gives figures for male and female life expectancy in the opening ‘context’ section, while also noting that women are likely to experience more years than men with disability or poor health. Eight of the PCT schemes also referred to the lower life expectancy experienced by men in comparison with women, as part of the background and explanation for the policy.

Schemes also described the main differences between women and men in causes of ill-health and death, largely with reference to male mortality. The documents describe men as being more likely to develop and die from cancer and as having higher rates of suicide, while women are described as experiencing more depression, anxiety, arthritis and rheumatism.

However, schemes do not mention similarities between women and men in cause of death – coronary heart disease is the leading cause of death for both women and men among the over-65s for example but this is not made explicit. No document describes congruence between women and men in patterns and causes of mortality and morbidity and the ways in which women and men share health risks is thus obscured, while the discourse of difference is extended. In addition, neither the DH guidance nor the PCTs discuss complexities in these experiences across the life course:

Men are twice as likely as women to both develop, and die from, the ten most common cancers affecting both sexes; Incidence and mortality for heart disease are much higher in men; 60% of sudden infant deaths occur in boys; Men are 3 times as likely to take their own lives; 78% of drug-related deaths occur in men. (DH, 2007: 14)

Causes of difference are also largely not described. The guidance describes ‘the vast differences in susceptibility to different conditions between men and women’ (2007: 11) but does not take this further in terms of the explanations of this susceptibility, and few of the PCT documents discuss reasons for the variations. One PCT refers briefly to biological influences in the context of specific diseases:

Women’s immune systems make them better able to resist some kinds of infection (such as tuberculosis), than men. (North Somerset PCT SES: 31)

Another describes sex as biological, and gender as socially constructed, but refers to inequalities in life expectancy as gender variations, with no discussion of causes, rendering the difference in longevity between women and men as social in origin.

Thus the discourse framed by this presentation is one in which women and men are essentially different and separate (Connell, 2012). The discourse presents ‘men’ as dying early and from specific diseases, in contrast to ‘women’. The juxtaposition of men’s premature mortality and causes of death also constructs men as disadvantaged in comparison with women, in the same way gender-specific medicine constructs men as disadvantaged (Annandale and Hammarström, 2011). These differences remain unexplored but when placed next to the discourse on gender differences in health behaviours, the underlying narrative is one in which masculinity and men’s behaviour play a key role in explaining such variations.

Men behaving badly

In addition to the presentation of men and women as experiencing essentially different health risks, the schemes and documents identify gender variations in behaviour, including health behaviours and help seeking, such as consultation and screening uptake.

The focus on health behaviours in the schemes is primarily in relation to men. Men are described as being more likely to smoke, use alcohol and illicit drugs and have a poor diet. Schemes also describe men as being much less likely to consult a GP or dentist in comparison with women, being less likely to use schemes such as smoking cessation or weight loss programmes, more likely to present late and to have their cancers diagnosed late:

Men in our Borough do not take up services as readily as women and this is having a detrimental impact on their health. (Heywood Middleton and Rochdale PCT SES: 24)

Similarly, DH guidance states that:

… men in general do not access primary health care as often as women; they tend to wait until symptoms are serious or can no longer be ignored. This leads to late diagnosis, and poor-health, and in the long run costs the health service money. (DH, 2007: 11)

A smaller number of schemes discuss some aspects of women’s health behaviours as problematic. Two raise concerns over an increase in problematic use of alcohol among women; one refers to increasing smoking among young women and another to low levels of physical activity among women. Three other PCTs only discuss women’s behaviours in the context of other equalities, noting smoking and the use of recreational drugs and alcohol among young lesbian women; low levels of exercise among Pakistani and Bangladeshi women; and lower levels of smoking among Black and Minority Ethnic women in comparison with Black and Minority Ethnic men.

Gendered differences in behaviour and health risk are also embedded in the discourse around the threat to health of gender-based violence mentioned by four of the PCTs. Within this discourse it is mainly women who are seen as at risk, while men are the implicit perpetrators, although one PCT notes that a minority of men also experience such violence. Female genital mutilation is also identified as a health risk for women in the context of maternity services by one PCT.

Reproductive women

One aspect of gendered discourses of health is the association between reproduction and women’s health, and the various ways in which women’s physical and mental health has been seen as ‘other’ and problematic as a result of their reproductive nature (Annandale, 2009). Historically, public health strategies have addressed women in their role as mothers, particularly in smoking cessation interventions which encourage women to stop smoking on behalf of their unborn child (Graham, 1987; Jacobson, 1981) while the exclusion of women from medical research and pharmaceutical trials reproduces the notion of the problematic female reproductive body, in need of protection but which also complicates the interpretation of findings.

While increasing knowledge about biological influences on health has helped to deepen this discourse (Annandale and Hammarström, 2011; Wizeman and Pardue, 2001), adding understanding of the significance of genetic variations as well as hormonal and reproductive influences, gender-specific medicine also serves to reinforce the notion of men and women as fundamentally different, obscuring the similarities between them while underscoring the significance of women’s reproductive nature.

While most of the PCTs did not identify women’s reproductive health needs as of particular concern, for the four that did these were constructed as relating to pregnancy and childbirth, and services for contraception or abortion were not mentioned. The reference to reproduction reflected medical discourse in relation to the delivery of maternity care and the role of women as mothers:

The foundations for virtually every aspect of development – physical, intellectual and emotional – are laid during pregnancy and the first years of life. It is essential that mothers and children receive a quality of care from pre-pregnancy, through pregnancy and childbirth, to the early days and years of life. (Ashton Leigh and Wigan PCT: Population Report: 15)

Within such discursive turns women as reproductive bodies are seen as responsible for the health and wellbeing of their (unborn) child(ren).

External risk factors

There is a virtual silence on the part played in health differences between women and men by external factors such as paid and unpaid work, poverty, exclusion or the environment. One PCT cited data showing that men were more likely to be in full-time employment than women, connecting this with men’s poorer health, and data showing that women were more likely than men to be in privately rented accommodation, in relation to women’s health risks, but the connections between these risk factors and health were not discussed. For all the other PCTs included here, there is no mention of such influences. This invisibility of structural and economic factors limits and constrains the discourse embedded in gender equality documents to essential differences between women and men, obscuring the significance of wider influences.

Gendered service needs

Research on whether women and men use services differently, in what ways and with what effect suggests that in high income countries with universal coverage men under-utilise services (Courtenay, 2011; Farrimond, 2012), although as Annandale et al. (2007) observe, this over-simplifies complex patterns and calls into question how we measure use–need ratios.

The gender equality schemes explored here discursively construct women and men as requiring different services, delivered differently. Men in particular are identified within the schemes as needing services targeted through male spaces which include workplaces, football and social clubs, and shops used more often by men. There is less discussion of women’s service needs or the use of female spaces to enable or encourage women to consult. Where such a need is identified it is often in the context of women’s mental health or the needs of specific groups, including for example women-only exercise sessions for minority ethnic women, and cervical cancer screening programmes to increase uptake among lesbians and bisexual women.

Four PCTs discussed both women and men in terms of service use. One suggested that both men and women might be deterred from using services because of childcare needs or where the service might feel unsafe or unwelcoming, while another suggested increased flexibility around appointments and opening times is important for both men and women.

The relative lack of discussion of women’s service use suggests that women already consult at the level required for optimum health, and that their needs are being met. Again men are implicitly compared with women, who are framed as the norm, and as advantaged. For example, the use of the word ‘disproportionate’ here suggests an over-use by women:

There is disproportionate take-up of some services related to gender including a reduced take-up by men of some screening and preventive services. (Heywood Middleton and Rochdale PCT SES: 3)

DH guidance also presents women and men as needing services to be delivered differently:

Consideration could be given to providing information through workplaces, pubs, clubs and shops popular with men. (DH, 2007: 11)

Consideration should be given to tailoring appointment times so that both men and women can access the service easily … (men are more than twice as likely to be in full time work). (DH, 2007: 11)

NHS organisations are asked to ‘examine why services are under used by men’, and ‘design health services accordingly’ (p. 12), with the example of breast cancer screening, currently only offered to women:

For instance, although breast cancer screening can affect men as well as women, preventive and screening services are generally targeted at women only. (DH, 2007: 12)

The implication that these services should be extended to men contrasts with the fact that of the 55,000 new cases of breast cancer diagnosed each year in the UK, only 400 are male (Breast Cancer Care, 2012) and again helps to construct men as disadvantaged by service provision.

Gendering governance tools

Research frequently highlights the lack of gender-disaggregated data (Theobald et al., 2005) and the schemes reflected this concern by reiterating the need for accurate information about women’s and men’s health, service use, outcomes, satisfaction and workforce data. Schemes also described impact assessment tools used as part of their strategy. These typically have two stages: initial screening to determine if a policy is likely to directly or indirectly discriminate against the groups covered by the legislation, and a second more detailed assessment of impact which is completed for policies identified as needing further analysis. The tools vary – the Policy Screening Checklist used by Trafford, for example, requires yes/no answers in relation to whether a policy might exclude people as service users or employees, or indirectly disadvantage specific groups, while Lambeth PCT uses an impact assessment tool which allows narrative answers. But however they are framed, such tools are by definition limited to policies already written (Bacchi and Eveline, 2010) and can only answer questions about the problem as it has already been represented. As such the tools define the boundaries of concern about gender equality and frame the problem in specific and limited ways, as exclusion from participation as a service user or employee. They are unable to open up questions about how the problem has been defined, or the unexamined and normative assumptions embedded in the policy.

The gendered workplace

The NHS in the UK has a relatively long history of anti-discrimination policies as an employer. The schemes explored here outlined key differences in their workforces, particularly the high proportion of female employees and the disproportionately low number of women in senior positions. The schemes included generalised comments about the need for ‘fair’ recruitment and career development opportunities, family friendly policies and childcare support, and the need to monitor gender in the workforce and in recruitment processes. Terms such as ‘family friendly’ and references to childcare were accompanied by an argument that these should ‘benefit both genders’ and that both women and men are parents. Similarly DH guidance suggests that:

Women are frequently put at a disadvantage by policies that are not family friendly whilst men can also be disadvantaged when family friendly policies are aimed just at women, and assume that men have no parenting responsibilities. (DH, 2007: 3)

Policy as discourse – discussion

The representation of ‘the problem’ of gender in these equality schemes drawn up by primary care organisations created a particular depiction of gender and health. Men and women are portrayed as different through the use of specific data on mortality and morbidity, and an absence of reference to structural, cultural and social factors. In addition, biological factors, when referred to, are those relating to reproduction, and female, while other explanations are not explored at all. In this depiction, men are consistently portrayed in contrast to women. This discourse also frames men and women as behaving differently, with consequences for their health. Men take more risks with their health, they smoke, they consume alcohol to excess, they eat badly, they are overweight and they do not use services. The data are either presented only for men, and the comparison with women is implicit, in relation to health behaviours, or differences between women and men are made explicit, as with service use. While women are sometimes visible, this is largely as specific groups of women – black and minority women, for example. The policy problem is thus represented as one in which women and men are different in their health behaviours, with women constructed as both the norm against which men are failing, and as advantaged in comparison with men, in that they take more care of their health and they derive more benefit from health services.

The schemes are also framed around specific forms of governance, particularly the use of information and technocratic solutions, and again this presents the problem in limited ways. Data which are organised around difference promotes the importance of gender-specific health concerns and obscures similarities between women and men (Risberg et al., 2009). Technocratic ‘fixes’ such as impact assessment tools increase reliance on what can be measured (Kuhlmann, 2009), limiting the problem of gender inequality to a narrow range of solutions. Schemes also shape the discourse around variations between groups of men and groups of women, adopting an additive approach to inequality in which individuals are defined in terms of their membership of various groups. Assumptions embedded in these representations are not discussed, effectively reinforcing differences between these groups.

There are also significant silences in these schemes on the wider influences on the health of women and men including those which are beyond the health care system, and on causes for differential behaviours. Scheme documents do not discuss why men consult less, but the solutions presented create a discourse in which men’s under-consultation is due to how services are delivered: services are not located in the right places, and they are not open at the right times. The implication that services are located appropriately for women and that women consult at the right level are taken-for-granted beliefs that are not challenged (Risman and Davis, 2013).

There is a further silence regarding male health behaviours. The schemes do not address why men behave badly, in terms of smoking, alcohol and diet and obesity, or why they are less likely to use interventions to change health behaviours. Male health behaviours are represented as normal masculinity, and policy solutions to deliver services to men frame the problem in ways which also ultimately support and reinforce such gender discourses.

We can also speculate on the silence inherent in the seven schemes that do not discuss differences between women and men in relation to health needs and the gender equality duty. This silence is complete, in that these schemes do not instead highlight similarities, or raise more complex issues about why gender is significant or how we might interpret gender and sex as health influences. Gender is simply mentioned as one of the list of equalities addressed by a generic policy or strategy. These are schemes which focus on process not difference, and present data needs, technocratic tools and reporting and evaluation systems as the solution to a problem which in turn is represented as one of insufficient data, the lack of tools and weak governance and which further distances the implementation of gender mainstreaming from its origins in participatory ideals and activism (Squires, 2010). However gender bias arises both when systems fail to take into account biological influences on the health of both men and women, and when they assume differences when there are none (Risberg et al., 2009).

Schemes are also silent on the question of gender stereotypes in medicine and the role of bias in health professionals (Annandale, 2009; Govender and Penn-Kekana, 2007). For example studies reveal complex differences in how women and men are treated by physicians including under-diagnosis reflecting stereotypes about ‘candidacy’ for particular conditions (Courtenay, 2011; Wilkins et al., 2008). Health professionals ‘do gender’ in their consultation practices, including the questions they ask of women and men (Risberg et al., 2009).

However the representation of the policy problem in these schemes does not include these influences: while they all cover equalities training for the whole workforce, there is no reference to gender bias in medical discourse and training. Two PCTs and the DH guidance described the need to address under-diagnosis of heart disease among women and depression among men, but did not discuss why these patterns exist or how to challenge such stereotypes. Another PCT notes the existence of ‘stereotypes around the treatment needs of people based on their race, age and disability’, but does not mention gender. While gendering medical education and knowledge is outside the scope of primary care organisations, framing the solution without reference to such bias obscures the significance of stereotypes in gender inequity in health interventions at the same time as gender-medicine – and the focus on biological at the cost of gender – is becoming increasingly significant in medical practice and education (Annandale and Hammarström, 2011; Risberg et al., 2009).

Throughout the discourses identified above, gender relations of power are not identified. This is perhaps not surprising in that the schemes analysed here are intended to be practical plans to institute systems that address inequality in everyday policy making. But the silence on these relations shapes the resulting discourse in specific and problematic ways, as well as implying that gender justice can be achieved without challenging such wider differences, a failing also observed by Sen and Östlin (2008) in the context of gender mainstreaming in international institutions.

Conclusion

Adopting a ‘what is the problem represented to be’ approach offers new lines of enquiry and new ways of thinking about the progress made by gender mainstreaming strategies, particularly in the ways in which taken-for-granted assumptions embedded in the representation of policy problems can be unpicked. In the example explored here, the policies developed by health organisations in England in response to the requirement to eliminate gender discrimination and promote equality of opportunity between men and women can be seen as contributing to discourses of gender difference. The ‘problem’ as represented in these policies, and in the solutions, reinforces ideas about women and men as being essentially (and biologically) different, obscuring similarities between women and men and the significance of wider structural influences on health and gender relations of power which underpin gender inequalities in complex ways. The problem is also represented in terms of technocratic solutions, in which data on differences between women and men are privileged. These solutions reflect the appeal of the concrete rather than abstract formulations (Lee-Gosselin et al., 2013) but help to construct the problem around men’s needs given statistics on both health outcomes and service use.

It is also important that these policies render men as the disadvantaged gender, in terms of health care, and this has significant implications for health organisations and broader policy discourse, particularly in the context of the emerging discipline of gender-specific medicine and tensions between the men’s health and women’s health movements.

But what does the WPR approach to gender equality policies in health offer policy makers? One direction might be to develop wider understanding of the productive capacity of policies by those who are charged with making and implementing them. Through articulation of the ways in which policies represent problems we might open up spaces for those engaged in policy making to engage with the production of ‘knowledge’ and the silences embedded in policy texts and other discursive practices. While this might seem unrealistic, given the vulnerability of gender mainstreaming and equality policy making to resource cuts and shortages, and changes in political leadership, the alternative – to continue to assert the value of gender equality strategies while lamenting their limited achievements – perhaps brings greater risks of the abandonment of the project altogether.