3.1 CESCR General Comment No. 19

The General Comment (at para. 2) defines the right to social security as follows:

This definition follows the ILO's Social Security (Minimum Standards) Convention, 1952 (No. 102) in linking income support to work interruptions based on nine contingencies. While the General Comment recognises that social security plays a role in poverty alleviation (at para. 3), it fails to move from the traditional work-related formulation of social security to a broader inclusion of causes of poverty, such as lack of housing and food, and natural disasters and emergencies (Langford and King, Reference Langford, King and Langford2008, p. 505). For millions of women who have no access to formal employment, this definition is problematic. The ILO has itself begun to define social security more broadly by including a fourth category in the definition – ‘general poverty and social exclusion’ (2010, p. 13). Accordingly, the General Comment lags behind recent advances in this area by remaining tied to a traditional definition that fails to acknowledge large sections of the population facing poverty, predominantly women.

In addition, by restricting itself to the ILO's nine contingencies, the General Comment fails to mention other contingencies that may be specific to working women, such as domestic violence, a frequent cause of work interruption for women. Violence and sexual harassment affect many women's earning capacity, requiring them to leave or move jobs or remain unemployed (McFerran, Reference Mcferran2011). Caring functions, often performed by women, are not listed as a contingency, since the focus is on the worker who cannot work due to disability (for example), rather than the carer of such a person, whose work may also be affected. Thus, the sickness, disability, injury and old age of those depending on the care of another person, whose income is affected by these obligations, are also contingencies against which the right should protect.

Despite the shortcomings in the definition, the General Comment pays specific attention to gender issues. It notes that the obligation to ensure the right to social security is ‘enjoyed without discrimination, and equally between all men and women … pervades the Covenant’ (para. 29). It devotes specific attention to gender equality (para. 31). First it cross-refers to the earlier CESCR General Comment No. 16 on the equal right of men and women,Footnote ⁴ which highlighted (at para. 26):

It then adds the following (at para. 32):

This paragraph raises a number of important issues. It gives further content to CESCR General Comment No. 16 by showing how equality in public and private pension schemes can be achieved. By recommending that States Parties remove factors that create unequal wage outcomes or address the impact of family responsibilities, presumably by providing, for example, child care or equal wage laws, the General Comment hints at more structural changes to the workplace, economy and society. This substantive and more far-reaching approach is needed if gender inequalities in social security are to be addressed. Similarly, by requiring States Parties to take account of care responsibilities in calculating benefits, the General Comment goes considerably further than most countries are doing at present and renders visible care work that is so often naturalised and ignored. The acknowledgment of women's poverty reflects a recognition that women generally earn less than men and face greater financial hardship. Again, women's unequal care burden is acknowledged. Elsewhere in the General Comment there is specific mention of domestic work and homework (para. 31), both women-dominated forms of work that have had limited visibility and protection in the past.

The paragraph does not, however, expressly note that women are often involved in unpaid subsistence labour, work in family enterprises, and in household and reproductive labour, which means they have no opportunities to access contributory social insurance. Their labour is not seen as work. The paragraph also makes no reference to violence against women, alongside other cultural measures that serve to restrict and control women's access to work and to social security.

While the detailed elaboration of the meaning of the right to social security by the CESCR in General Comment No. 19 is welcome, a more fully developed gender-based approach to the right is needed. Before looking at what might inform such an approach, this paper considers the circumstances in which gender shapes women's poverty, including their relationship to work and to social security.

4.1 Women's poverty in a global context

Women throughout the world, in developed and developing countries alike, face disproportionate poverty, un- or under-employment, poorer working conditions, and greater responsibilities for the care of others than their male counterparts. Recent decades have seen positive changes in certain development indicators relating to women and girls in many parts of the world, including educational enrolment and labour force participation (World Bank, 2011). However, despite their increased access to work and income opportunities, women predominate in the informal sector, in unpaid work in family enterprises, in less profitable agriculture and other business, and in the lowest paid jobs (UNRISD, 2010, pp. 111–119). As a consequence, women in all countries generally earn less than men (World Bank, 2011, pp. 16–17). While employment of women has increased, less than half of the world's women have income-producing work, as opposed to nearly four-fifths of the world's men (World Bank, 2012).

As a result of the negative impact of neoliberal economic policies on women in the developing world (Elson, Reference Elson, Molyneux and Razavi2002), the majority of women are located in precarious informal work, including migrant labour, which is inadequately protected (Standing, Reference Standing2011, pp. 60–63; Kabeer, Reference Kabeer2008, pp. 32–33). Globalisation has resulted in growing international migration flows that are increasingly feminised as women migrate to provide care and other labour elsewhere in the world or within states (Fudge, Reference Fudge2012). This work is often exploitative, dangerous and precarious. Women migrant workers are particularly vulnerable. Many take on domestic work in the unprotected confines of private homes, or engage in the sex trade and are trafficked (Ehrenreich and Hochschild, Reference Ehrenreich and Hochschild2002). Women migrate to richer countries to perform care work in part to meet ‘care deficits’ caused by the increasing entry into the labour market of women in the developed world (Fudge, Reference Fudge2011). The migration of these women workers is in turn causing ‘care deficits’ in their home countries, where they have left children and other dependants (Hassim, Reference Hassim2008). The growing ‘feminisation of labour’ has seen women take the bulk of casual and seasonal jobs and work in export processing zones without any expectation of an adequate wage or benefits (Standing, Reference Standing2011, p. 60). The increasing casualisation and informalisation of work ensures flexibility for employers and reduced social responsibility for corporations and states (Razavi, Arza, Brunstein, Cook and Goulding, Reference Razavi, Arza, Brunstein, Cook and Goulding2012, p. xxii).

These changes, accompanied by globalisation, have resulted in growing inequality, insecurity and the loss of workplace rights for workers (Standing, Reference Standing2011, p. 14). Unemployment, underemployment and low-earning self-employment and subsistence work is a reality for most workers in developing countries and many in the developed world (World Bank, 2012). In fact, the model worker around whom social security standards were designed (the formally employed, full-time, male breadwinner) (Lamarche, Reference Lamarche2002) is rapidly becoming a rarity in global terms. Women are particularly vulnerable to these forms of precarious work (Razavi et al., Reference Razavi, Arza, Brunstein, Cook and Goulding2012, p. xxii). There has been a ‘feminisation’ of work in terms of its gender composition alongside a ‘feminisation’ of working conditions amounting to deregulation, discrimination and reduced protection (Sabates-Wheeler and Kabeer, Reference Sabates-Wheeler and Kabeer2003). Women also engage in homework for multinationals and face harsh working conditions on the bottom rungs of their own societies’ labour markets. A further impact of globalisation occurs where land is sold to multinationals, affecting women's subsistence livelihoods (UN Women, 2013, pp. 3–4).

The increasing number of women in the labour market has not seen a reduction in their caring responsibilities, nor an adequate societal response to this unequal burden (Razavi and Hassim, Reference Razavi and Hassim2006, p. 7). In fact, increased life expectancy has meant that women are now caring for both children and elderly relatives while also needing to earn an income – a ‘triple burden’ (Standing, Reference Standing2011, p. 61). The global financial crisis since 2008 has added to the struggles of poor women. Higher food prices, job losses, austerity measures and cuts to development aid have deepened vulnerability in rich and poor countries (Ortiz and Cummins, Reference Ortiz and Cummins2012, pp. 4–7).

4.2 Discrimination underlying women's poverty

The gender dimensions of poverty are causally related to women's unequal position in all realms of society – family and community, the economic and the political. Feminist legal theorists have highlighted the way in which different roles are allocated to men and women, through the law, within the public and private spheres (for example Olsen, Reference Olsen1983; O'Donovan, Reference O'donovan1985; Boyd, Reference Boyd1997). The division of labour in the family restricts women's access to the labour market (Fineman, Reference Fineman1995, Reference Fineman2004). This means that women are generally poorer than men and have less power, status and influence both in the home and in society. The ‘feminisation of poverty’, a term coined to describe the growing number of women who comprise the poor, has been used to record this phenomenon at both the micro-level of home and family and at the macro-level within nations and globally (Chant, Reference Chant2006). Women are primarily responsible for caring in society, an activity which is devalued, usually unremunerated or poorly paid, and which restricts women's entry into other parts of the labour market (Williams, Reference Williams2000; Rhode, Reference Rhode1999; Mahony, Reference Mahony1995). Subsistence work is also women-dominated and, along with care-giving work, is usually poorly paid or unpaid. Feminists working in development and economics have pointed to the significant unpaid and invisible contribution that women make to the economy through such work (Waring, Reference Waring1999; Kabeer, Reference Kabeer2008, pp. 28–29).

In addition to general disadvantage experienced by women as a result of their gender, specific groups of women face heightened inequality as a result of further forms of discrimination based on factors such as race, disability, age, religion, ethnic and indigenous status, or geographical location. This discrimination often translates into economic disadvantage, affecting vulnerable groups of women's access to resources, and results in greater poverty. Violence against women, prevalent worldwide (WHO, 2014), is sometimes used to control women's access to property and often contributes to the impoverishment that follows when women leave abusers. Legal and cultural barriers prevent women in many parts of the world from owning land and other property (Pruitt, Reference Pruitt2009). It is estimated that just 1 per cent of the world's women own land (UN Women, 2013, p. 3). Even where women do own resources or bring in income they may not be able to fully access or control these in patriarchal settings where men are designated household heads (UN Women, 2013, p. 7). Women also encounter stigma and stereotyping in a range of contexts that have bearing on their economic access (Cook and Cusack, Reference Cook and Cusack2010, p. 22). For example, women applying to rent property or for bank loans may be turned away because they are seen as incapable of managing money. Women also face exclusion from political representation, decision-making and full participation in many parts of the world, which has an impact on their life chances and material position.Footnote ⁵ The many facets of discrimination against women, discussed here in brief, contribute to gendered poverty across the globe (Fredman, Reference Fredman, Goldblatt and McLean2011a).

4.3 Gender and social security

As noted, social security is one of the means to address women's poverty and disadvantage. The absence of adequate or appropriate social security adds to the poverty burden faced by women. Where social security does exist, it sometimes discriminates directly against women, but more usually does so indirectly (Sepúlveda and Nyst, Reference Sepúlveda and Nyst2012, pp. 32–33). Direct discrimination may occur where women are deliberately excluded from participating in a scheme. Indirect discrimination may occur where, for example, social assistance payments are provided to household heads. Since these are usually men because of patriarchal assumptions in families and the wider society, such an approach may have a discriminatory impact. Men are less likely than women to use the income for the benefit of the whole household (Bradshaw, Reference Bradshaw2008). Access to social security is in some cases more difficult for women who face the danger of violence when collecting payments, or within households where men attempt to attach or control women's income (Goldblatt, Reference Goldblatt2005). Women's caring responsibilities for children, the elderly and sick are often unremunerated or unacknowledged in social security provision and very little has been done through social security measures to challenge the lack of involvement of men in the care and support of families (Razavi, Reference Razavi2011b).

Even where women are in the workforce, they often work part-time, are poorly paid, have smaller social security entitlements and hence still rely on men for support. Welfare states in the developed world and under state socialism have proved generally deficient (to varying degrees) in addressing sex/gender discrimination in relation to pay and the status of women's work, and in relation to their reproductive and caring functions in those societies (Razavi and Hassim, Reference Razavi and Hassim2006, p. 7; Fraser, Reference Fraser1989, Reference Fraser1994). While some European countries have tried to address the impact of the gender wage gap and work interruptions faced by women assuming caring functions (Luckhaus, Reference Luckhaus2000, pp. 168–169), many countries perpetuate gender inequalities through gender ‘neutrality’ or indirectly target women through cutbacks to existing programmes. In some countries, welfare provision has become increasingly punitive, especially for single mothers, with recipients stigmatised, stereotyped and burdened with conditions (Fraser and Gordon, Reference Fraser and Gordon1994; Williams, Reference Williams1994). Welfare restructuring and cutbacks since the 1980s have been accompanied by a corrosive discourse that casts women recipients of welfare as irresponsible, immoral and lazy, and that sees dependency as deviant (Fineman, Reference Fineman1995). Recent austerity measures since the global financial crisis of 2008, often involving reductions to single-parent payments (Goldblatt and Lamarche, Reference Goldblatt and Lamarche2014), have also been accompanied by negative discourse.

In the developing world, the lack of adequate welfare safety nets has a significant impact on women's poverty (UNRISD, 2010, pp. 107–134, 185–206). In the past decade some of the middle-income developing countries have introduced social assistance programmes that have had marked impacts on human development (UNDP, 2011). While traditional social security systems were often designed around male-dominated formal employment, more recent programmes in developing countries have tried to address this imbalance by specifically targeting women (as citizens rather than as workers). Social assistance programmes targeted at women in some developing countries have had positive results, but many gaps in provision remain and certain problems have emerged. The development literature illustrates how policies can reinforce existing gender inequalities based on patriarchal assumptions about work, family and the economy (Chant, Reference Chant2008; Bradshaw, Reference Bradshaw2008). Some of the programmes targeted at women have the effect of perpetuating gender divisions by making women responsible for care, yet addressing women's disproportionate responsibility for care in these societies has been highlighted as a critical consideration for social protection policy (UNRISD, 2010, pp. 185–206; Williams, Reference Williams2010; Razavi, Reference Razavi2011b). Conditions that are attached to such assistance often impose additional burdens on women (Molyneux, Reference Molyneux2008; Lund, Noble, Barnes and Wright, Reference Lund, Noble, Barnes and Wright2009).

At the global level, women worker migration results in a range of challenges around social security provision and rights (Lister et al., Reference Lister, Williams, Anttonen, Bussemaker, Gerhard, Heinen, Johansson and Leira2007) – the social insurance and other rights of the migrant workers and the complexities of claiming rights as non-citizens; the adequacy of the social security systems in destination countries in meeting the rights of women workers and their dependents; and the rights of women and their dependants in the countries from which they originate.

This overview has highlighted the nature and causes of women's poverty and the gender issues implicated in social security systems in both developed and developing countries. Feminists working within the social policy and development fields have proposed a variety of approaches to ensure that social security addresses women's poverty while also promoting gender equality. Many of these writers place value on a human rights approach to underpin their proposals (Waring, Mukherjee, Reid and Shivdas, Reference Waring, Mukherjee, Reid and Shivdas2013; Holmes and Jones, Reference Holmes and Jones2013; Sen, Reference Sen2011; Razavi, Reference Razavi2011a; Sabates-Wheeler and Kabeer, Reference Sabates-Wheeler and Kabeer2003). The next section proposes a gender-based approach to the right to social security and a set of principles for ensuring that the right is interpreted and applied following such an approach.

5.1 Rethinking ‘work’ in the right to social security

While the ILO's Social Protection Floor Recommendation and wider international use of the term ‘social protection’ is drawing poverty into understandings of the scope of social security, the narrower worker-oriented definition of social security still has prominence in human rights law, as discussed above in relation to CESCR General Comment No. 19. ‘Work’ is central to the meaning of social security in this definition since it is the absence of ‘work-related income’ that prompts the operation of the right. Thus, it is essential to deconstruct the meaning of ‘work’ as it is used in human rights law, to consider what work means for the world's women, and to redefine this core concept to better reflect these realities. This is important for two reasons: first, it means that where social security is linked to work, greater numbers of women workers should be given access to social insurance benefits by being brought within the definition of worker; and second, it suggests the concurrent need to ensure that social security, delinked from work, is also available to all who need it (often women) as a citizenship entitlement (in the form of universal schemes and social assistance).

In order to appreciate the different dimensions of women's labourFootnote ⁶ and other activity, the following categories are discussed: reproductive work (unpaid and paid); productive work (formal and informal that is paid and unpaid); and non-productive activity.

Reproductive work is both unpaid work in the home and paid domestic work. It can be formalised or informal. Because unpaid reproductive work (household labour and care activities) is located in the home it is deemed private and is not recognized as work. It is seen as the ‘natural’ responsibility of women and separate from the public world of work, where labour requires remuneration and (sometimes) social security.

Productive work can be formal work that is paidFootnote ⁷ (recognising that this in itself is gender defined, i.e. certain jobs are reserved for women at lower status and pay) and informal work that is paid but unregulated. It can also include work that is unpaid, such as subsistence work or work in family enterprises where women receive no income for their labour. As with unpaid reproductive work, this type of labour is ‘privatised’ and seen as falling outside the employment contract.

Note that some of these categories are concurrent, such as unpaid reproductive and unpaid or paid productive labour. Women may perform different types of work simultaneously, for example by looking after children while engaging in homework for income and also undertaking unpaid subsistence work such as collecting water or looking after livestock. Paid care work is both reproductive and productive work. In fact, care work highlights the blurring of the boundaries between these different types of work and the fact that these categories are used to justify hierarchies of reward.

Non-productive activity must also be acknowledged as a dimension of human experience. Not every person is able to engage in reproductive or productive labour due to their age, disability or illness. Yet such people are part of society and have equal entitlement to the fundamental preconditions for a dignified life (Fineman, Reference Fineman2010; Nussbaum, Reference Nussbaum2006). Non-productive activity requires recognition in the conceptualisation of social security as a social good that meets the needs of all people, regardless of their capacity to produce. In many cases it is only a person's initial location within paid productive labour that allows for the recognition of their inability to work that in turn entitles them to social security of any sort. Yet structural unemployment, together with discrimination, mean that for millions of women a paid job is not attainable in their lifetimes. Social security should be available as of right for those who have no access to the labour market in the first place – predominantly women – as well as for those who are unable to work.

Social security is generally linked to formal work which is just one of the categories of women's work. Even then, the benefits attaching to formal work do not always extend to forms of formal work such as part-time or casual work, in which women predominate. The small formal sector in the developing world and the shrinking formal sector in the developed world mean that social security linked to work is limited in most countries of the world. The past decade has seen important growth in social assistance in Latin America, Asia and Africa, often directed at women (UNDP, 2011). While social assistance is usually the major form of income support where it exists, it is often minimalist in developing countries (UNRISD, 2010, pp. 107–134) and facing attrition in many developed countries (Goldblatt and Lamarche, Reference Goldblatt and Lamarche2014). In addition, social assistance, without efforts to recognise women's existing contribution through unpaid work, to bring women into work, or to formalise informal work with attendant benefits (Heintz and Lund, Reference Heintz and Lund2012), is not enough.Footnote ⁸

This requires two responses. First, the work that women already do (reproductive and productive) should be recognised as work that attracts social insurance rights. This should occur alongside the generation of additional opportunities for work and livelihood for women, and the restructuring of society's approach to employment and care-giving work (Fraser, Reference Fraser1994). Second, and at the same time, social security must be understood as an entitlement, not related to an individual's location within the labour market, that is provided through universal schemes or as social assistance, in a manner that promotes gender equality.

5.2 Application of the right to social security beyond the national level

Women's work in its multiple forms is crossing borders in a range of new ways, with profound implications for their working conditions, family responsibilities and entitlements. Globalisation has led to a situation where a woman from country A might be employed by a company located in country B in a workplace in country C with her family remaining in country A or accompanying her to country C. The issues of transnational commerce, global work flows, and multiple levels of labour regulation (or lack thereof) require a right to social security that has the flexibility to accommodate work (and its absence) across national boundaries.

Writing in the field of legal geography has employed the geographical concept of ‘scale’ used in mapping to focus on the impact of law and justice at different levels of spatial and legal experience (Pruitt, Reference Pruitt2008, p. 383).Footnote ⁹ Nancy Fraser has drawn on this geographical idea of scale in combination with the metaphor of justice as the balancing of scales to think about ways of applying her theory of justice in a world that is no longer operating purely on the basis of territorial delimitations (2008, Reference Fraser2010, Reference Fraser2011). Fraser (Reference Fraser2008) has noted an historical focus on the territorial state as the site of citizenship entitlements and a growing contemporary awareness that social and economic issues ‘routinely overflow national borders’ (p. 13). She argues that political space has been ‘misframed’ to exclude certain groups such as the global poor (what she terms the ‘transnational precariat’) from justice claims. Ideas of justice are often framed as citizenship entitlements that are restricted to members of nation-states without addressing what she terms ‘transborder injustices’ (p. 2). Her concern is with the question of who ‘counts as a bona fide subject of justice’ (p. 5) and how justice can be achieved rather than just what the idea of justice contains. For Fraser, ‘scale’ is a conceptual device or lens to ensure that justice (in all its dimensions) takes account of transnational realities in this ‘politics of framing’.

Fraser's notion of ‘scale’ or ‘framing’ within her political theory of justice has value in developing a right to social security that has relevance and application to multinationals with workers in many countries, migrants who are often denied rights in the destination countries, international bodies, arrangements between states, and so on. It encourages us to look not just at the content of the right and what it promises but also at the contextual realities of the subjects of the right and the agents that hold the power to realise or frustrate it. These are important insights in using the right to social security to address global poverty at a range of levels and not to look only to states to accommodate the needs of their own workers and local poor.

Susan Williams (Reference Williams, Rubenstein and Young2014) has suggested that Fraser's concept of scale can also be extended to the local or subnational level. This additional level also has relevance for the development of the right to social security. If the right is to have value across contexts it should be able to operate at the village or community level where women are subject to (often male) traditional leaders, local authorities and customary rules. These bodies play a role in allocating benefits and work and defining social entitlements. Private provision of social security through burial and loan societies, micro-insurance, trade union and worker association funds, and so on, should also be covered by the right. In addition, internal migration, affecting many workers in countries such as China (Li, Reference Li, Goldblatt and Lamarche2014), should not lead to reduced social security rights for women.

International human rights law recognises that richer countries will need to assist poorer countries through international assistance and co-operation.Footnote ¹⁰ Agreements between countries and international standards also need to ensure that women's rights are protected and met.Footnote ¹¹ Recent work on transnational or extra-territorial obligations proposes that rights should have a wider purview in global distributions, including in relation to non-state actors (De Schutter, Eide, Khalfan, Orellana, Salomon and Seiderman, Reference De Schutter, Eide, Khalfan, Orellana, Salomon and Seiderman2012). Ensuring that these obligations include a clear conceptualisation of the gender dimensions of the right to social security is essential.

5.3 The right to social security and substantive equality

As noted in the discussion of the international law above, equality and non-discrimination against women are central human rights, closely related to the realisation of other rights, including the right to social security. Equality, understood substantively, offers a valuable means of developing the right to social security from a gender perspective. Substantive equality can be used transformatively to address structural inequalities and achieve far-reaching social change. This paper uses Sandra Fredman's four-dimensional concept of substantive equality (Reference Fredman2011b), which includes the following aims (Reference Fredman2011c, p. 577):

This substantive equality framework assists in understanding the multidimensional nature of inequality as it affects women's experiences of social security. Thus, for example, obtaining a smaller pension on retirement because of child-rearing responsibilities and lower income over a lifetime denies women their equal rights to social security. Discrimination both causes this situation and results in economic disadvantage, lower social status and less control over women's life choices. The right to social security, interpreted in relation to substantive equality, should address the material dimensions of women's unequal status in the economy that result in women earning less and consequently suffering disadvantage in relation to social security (redistribution). It should also tackle issues such as stereotyping that prevent women from accessing certain forms of work or controlling social security payments within households (recognition). It should be a right that does not simply extend existing male-oriented social security models to women but requires a fundamental restructuring of such systems (transformation). Last, it should be a right that requires the involvement of women in the design and management of appropriate forms of social security and encourages their inclusion in society (participation).

It is also important to note that gender equality and the discrimination it seeks to address cannot be understood with reference to the experience of one group of women alone. Gender discrimination occurs alongside many other forms of discrimination such as race, age and disability. Forms of gender discrimination may affect women of one economic class differently from another and may take very different forms in different cultural contexts. The development of the right to social security, informed by substantive equality, requires careful focus on the particular context in which women find themselves. Close attention must be given to diversity, vulnerability and the complexity of discrimination.