Overview

The CSG is a publicly-funded means-tested monthly cash transfer, reaching more than 60 per cent of children, or over 12 million children each month (SASSA, 2017). The CSG was introduced in 1998 by the first democratic government of South Africa. Access to the grant has since been substantively scaled up and extended. The age threshold has been gradually raised from 7 to 18 years, the means test was revised in 2008 to be calculated at 10 times the grant amount (substantially extending eligibility), and the value of the grant, while modest, has risen marginally above inflation and is equivalent to USD26 (as of April 2017).

The CSG was designed to be gender-responsive and relevant to the realities and challenges of households in South Africa where mothers are not always able to live with the child, for instance due to the need to migrate for work, or high levels of HIV/AIDS leading to orphan-hood. The designated recipient of the CSG is the child's primary caregiver, who may be male or female and who is defined as the person that takes primary responsibility for meeting the daily care needs of the child.

The CSG is a rights-based entitlement and was designed to be redistributive. It was explicitly intended to contribute to the well-being of young children (primarily their food requirements) who were among the poorest and most vulnerable groups in the society (Lund, Reference Lund2008).

Is the CSG gender-redistributive?

The majority (98.1 per cent) of CSG recipients are women (mothers, grandmothers or other female caregivers) (SASSA, 2014) although, in terms of child beneficiaries, the numbers are evenly split between girls and boys (Brockerhoff, Reference Brockerhoff2013). Low engagement of men in care-giving is reflected in the low take-up of the CSG by male caregivers (Khan, Reference Khan2018, forthcoming; Patel, Reference Patel, Midgley, Dahl and Wright2017).

Social grants (especially the CSG) have had a substantial impact on the incidence and depth of poverty and inequality. The CSG effectively targets the poorest, and increases the incomes of the bottom income decile by almost four times (Inchauste et al., Reference Inchauste, Lustig, Maboshe, Purfield and Woolard2015). If social grants were not available, income inequality in 2010, as measured by the Gini coefficient, would have been substantially higher at 0.74 instead of 0.69. In particular, female-headed households would have experienced much higher levels of inequality in the absence of social grant income (Bhorat et al., Reference Bhorat, van der Westhuizen and Tseng2013b).

The receipt of social grant income has been effective in ensuring women's access to resources, especially in female-headed households (Patel, Reference Patel, Midgley, Dahl and Wright2017). Increased income through grants is also associated with higher likelihood of having bank accounts, life insurance and some form of savings (Delany et al., Reference Delany, Ismail, Graham and Ramkisson2008). However, this has not been sufficient to overcome the poverty gender gap and poverty still remains a gendered phenomenon (Posel and Rogan, Reference Posel and Rogan2012).

Only 13 per cent of CSG recipients have employment with a regular wage (Patel et al., Reference Patel, Knijn, Gorman-Smith, Hochfeld, Isserow, Garthe and Kgaphola2017). Despite structural barriers to women's employment, there are indications that the relatively small Child Support Grant is associated with a higher probability of female participation in the labour force (Eyal and Woolard, Reference Eyal and Woolard2011; Leibbrandt et al., Reference Leibbrandt, Lilenstein, Shenker and Woolard2013). The findings are in contrast with common fears that social grants create dependency (Williams, Reference Williams2007). While grants are primarily spent on food and children's basic needs, evidence also suggests that they can enable individuals to subsidise childcare costs; to finance and alleviate constraints to job search; to fund small enterprise creation; and improve productivity through better health, nutrition, and training (Devereux et al., Reference Devereux, Adato, Sabates-Wheeler, McConnell and Becker2011; Williams, Reference Williams2007).

The redistributive and poverty-reducing potential of the CSG has been limited by poor linkages with other strategies that can enhance the well-being of women and alleviate some of the burdens in the domestic sphere, such as increased access to quality child-care services and broader access to social insurance. Deliberate communication and administrative efforts could connect grant beneficiaries with existing poverty reduction programmes related to water, sanitation, electricity, food security and internet services (Patel, Reference Patel, Midgley, Surender and Alfers2018, forthcoming).

Is the CSG gender-transformative?

Dignity is a common theme in the narratives of beneficiaries, who feel validated in their personal, caring and social roles as a result of the CSG, demonstrating a close relationship between economic redistribution and human dignity in the lived experiences of grant recipients (Hochfeld and Plagerson, Reference Hochfeld and Plagerson2011). Beyond the confines of the household, the CSG has been associated with increased capabilities to fulfil societal roles, and to embrace active citizenship through participation in burial societies, savings groups, community forums and engagement with political processes (Plagerson and Ulriksen, Reference Plagerson and Ulriksen2015).

Yet the CSG is also associated with gender discrimination that has negative impacts on social recognition. Negative discourses are common in the media and within beneficiary communities that construct women as self-serving and a drain on resources, despite the evidence that show its poverty-alleviating effects (Surender et al., Reference Surender, Noble, Wright and Ntshongwana2010). These include concerns that CSGs incentivise childbirth, and more generally about a lack of responsibility amongst young mothers for their children and inappropriate use of the grants for personal consumption (Hochfeld and Plagerson, Reference Hochfeld, Plagerson, Patel and Ulriksen2017). However studies investigating these claims have found no evidence of widespread perverse incentive effects of social grants but show that the vast majority of mothers use the grant in ways that are beneficial for their children and households (Makiwane, Reference Makiwane2010; Rosenberg et al., Reference Rosenberg, Pettifor, Nguyen, Westreich, Bor, Bärnighausen and Kahn2015). Nonetheless, these discourses can undermine beneficiaries’ and children's rights to social assistance, and may stigmatise women for relying on grants, engendering a lack of cultural and social value for them despite their substantial social contributions via both informal productive labour and unpaid caring work (Hochfeld and Plagerson, Reference Hochfeld, Plagerson, Patel and Ulriksen2017).

As indicated in the previous section, the CSG does have multiple benefits that accrue directly to women both in their reproductive and productive roles. CSG recipients typically manage the grants and report relatively high levels of financial decision-making autonomy (Delany et al., Reference Delany, Ismail, Graham and Ramkisson2008; Patel et al., Reference Patel, Knijn and Van Wel2015). The CSG supports recipients’ economic independence within the household, enabling women to allocate financial resources for their preferred welfare-related and income-generating goals (Patel and Hochfeld, Reference Patel and Hochfeld2011; Patel et al., Reference Patel, Knijn and Van Wel2015).

Yet, a transformative analysis suggests that the feminisation of social grants does not automatically overcome the challenges that women face in bridging their double roles as poorly rewarded carers and productive economic agents (Plagerson and Ulriksen, Reference Plagerson and Ulriksen2015; Sabates-Wheeler and Roelen, Reference Sabates-Wheeler and Roelen2011). The grant recognises women's care-giving role but at the same time reinforces traditional divisions of labour. In practice, male caregivers struggle to access the grant and officials assume that it is for female caregivers (Khan, Reference Khan2018, forthcoming). Critics from a gender perspective also underline the instrumental way in which women beneficiaries are constructed as conduits for child well-being (Molyneux, Reference Molyneux2007). A lack of a fundamental shift in gender relations is confirmed by the changing patterns of childcare, redistributed between women within the extended household and across generations, rather than between men and women. Overall, the CSG has not served to significantly shift the boundaries and social expectations around care responsibilities. These expectations constrain women's ability to access and maintain their position in the labour market.

Overview

The Unemployment Insurance Fund (UIF) pays benefits in the event of unemployment, illness, maternity, the adoption of a child and death (by assisting family members of the deceased) (Mpedi and Fourie, Reference Mpedi, Fourie, Olivier and Kuhnle2008). Almost 1.6 million employers and 9 million employees are registered with the UIF (R.S.A., 2015). The UIF is financed by the monthly contributions of employers and employees (Bhorat et al., Reference Bhorat, Goga and Tseng2013a). The benefit amount is dependent on the claimant's wage level prior to applying for unemployment benefits and the previous length of time in employment. Maternity benefits can be claimed for a maximum period of 17.29 weeks. To gain maximum credit for the full period of maternity pay it is necessary to have worked continuously for four years (R.S.A., 2015).

The primary focus of the UIF is on the formal employment sector (where males are over-represented) and on the segment of the unemployed who have previously been employed (Bhorat et al., Reference Bhorat, Goga and Tseng2013a). The UIF excludes self-employed workers and employees who work less than 24 hours per month. In practice, less than 5 per cent of the unemployed receive unemployment benefits (NPC, 2011). Efforts to expand UIF coverage led to its extension in 2003 to domestic workers (including those working for multiple employers) and to seasonal workers (predominantly women).

Is the UIF gender-redistributive?

In a context of high unemployment, the UIF system has the potential to play a key role as the only public social insurance programme in South Africa that caters to the unemployed (Bhorat et al., Reference Bhorat, Goga and Tseng2013a). The Fund has a poverty alleviation mandate, and aims to alleviate the social and economic costs of unemployment (NPC, 2011). However, the redistributive role of the UIF is still confined to a small sub-section of the unemployed. This leaves many vulnerable groups excluded such as those who have never worked as formal employees before and many of those in the informal sector, in which women are over-represented.

Amendments introduced in 2003 were favourable to women since they provided for higher payments to lower income workers. However, this did not address the underlying issue of the under-representation of women as contributors and claimants. In the fourth quarter of 2017, more men (64 per cent of male employees) reported contributions to the UIF than women (54 per cent of female employees) (SSA, 2018). In relation to claimant behaviour, females are further under-represented. Women have become more disadvantaged with time since their representation ratio (ratio of female-to-male claimants including for maternity benefits over the ratio of female-to-male contributors) has declined from 0.99 in 2005 to 0.75 in 2011 (reaching its lowest point during the financial crisis in 2009–2010). The disparity between men and women in terms of access to the UIF and claim behaviour may result from women's over-representation in part-time, low-income, intermittent and precarious jobs, and thus fewer credit days with which to claim unemployment benefits (Bhorat et al., Reference Bhorat, Goga and Tseng2013a).

Both men and women contribute an equal proportion of their salaries to UIF, but men cannot claim maternity benefit. This implicitly acknowledges women's reproductive role in society. The UIF provision of maternity benefits has benefited a growing number of women in formal employment (and a small segment of women in informal sectors). Take-up of UIF maternity benefits has increased over time, from 81,000 claims in 2005/6 to 98,631 in 2016/17. In monetary value, maternity claims represent 11 per cent of all UIF benefits disbursed (R.S.A., 2017). Yet, available evidence indicates that the reality experienced by women remains precarious. Research has shown that the bureaucracy involved in accessing maternity benefits is complicated, with long delays in the disbursement of claimants’ maternity benefits. This drastically undermines the potential benefits of the money. If benefits are not received during the maternity period, mothers may not be able to remain on leave or may suffer as a result. Further financial pressures may occur as a result of the maternity benefit being smaller than a woman's salary, increasing the chance of mothers not taking the full four months to which they are entitled (Boswell and Boswell, Reference Boswell and Boswell2011). Recent changes to the law may improve these limitations: women may now apply eight weeks before the child is due and 12 months after birth; maternity benefits rates have been increased from a sliding scale of 38–60 per cent to a fixed rate of 66 per cent of the (female) contributor's earnings; maternity benefits have been de-linked and do not detract from a woman's ability to claim for other benefits and maternity benefits are available in the case of miscarriage in the third trimester (Department of Labour, 2016).

The proportions of (almost exclusively female) domestic workers covered by UIF contributions showed a large increase in the 18 months following their inclusion, albeit from a low base. Steady increases in the numbers of registrations of domestic employers have continued to occur, and a recent report showed that 667,300 domestic workers were registered with the UIF in March 2016 (66.1 per cent of domestic workers) (R.S.A., 2015). Nonetheless, many remain unregistered, primarily due to widespread non-compliance by employers (Malherbe, Reference Malherbe and Toit2013). Unemployment benefits payable upon dismissal are particularly necessary in the context of the job insecurity with which many domestic workers are faced. While there are provisions for domestic workers who lose one of their jobs to claim partial unemployment benefits, the low wages in this sector mean that the amount of benefit to which they are entitled may be too limited to provide any meaningful level of income security.

There are some provisions that complement maternity leave provisions, such as Child Support Grants for the children of caregivers earning less than the means test threshold. There is some provision for childcare through subsidies to the Early Childhood Development sector but, in general, the care of children under the age of three is considered the responsibility of parents and extended families, private institutions and non-governmental organisations (Mokomane, Reference Mokomane2011). In terms of social security as a whole, provision for domestic workers remains patchy, given the exclusion for example from access to the Compensation for Injuries and Diseases Act (COIDA). These complementary interventions remain fragmented and, as this section has shown, women face disproportionate challenges in accessing labour markets and social insurance.

Is the UIF gender-transformative?

Changes in legislation have mainly been in the direction of greater inclusivity, with some recognition of the implications for legislation of gender equality considerations. The right to maternity leave and maternity benefits is integral to the intersection of women's care and economic roles in society and has important implications for the social and economic well-being of women. Relevant legislation and the availability of maternity benefits for mothers registered with the UIF represent an important recognition of working mothers, and their combined productive and reproductive roles. While maternity benefits are institutionally problematic (such as delays and difficulties in claiming) and do not extend to cover nearly enough women (as they often have not contributed enough to receive full benefits), it is an important mechanism which can have symbolic and substantial value for women. Yet, these provisions remain confined within a narrow paradigm of assistance to women. They are provided as a social right to women in employment rather than as a universal right to all women and are primarily designed to protect the health of the mother (and child) rather than to enhance women's social rights through a holistic approach to work/family balance and gender equality considerations. For example, the amendment introduced in 2016 includes a clause that women must have worked for 13 weeks prior to submitting a maternity benefits claim. In a context of increasing feminisation of the labour market, and despite broad acceptance of women entering the workplace, traditional attitudes towards gender role division within the household persist and are echoed by current policy approaches (Gordon et al., Reference Gordon, Roberts and Struwig2012).

The inclusion of domestic workers in the UIF is progressive and provides a precedent for extending insurance to the informal sector (Dinkelman et al., Reference Dinkelman, Ranchhod and Hofmeyr2014). However, the data also shows that, for transformative outcomes to be achieved, complementary interventions such as improved access to quality child care are necessary as well as efforts to reduce administrative barriers, to counter discrimination against women in the workplace and to increase employer compliance with legislation (on contracts, wages and conditions of employment) (Malherbe, Reference Malherbe and Toit2013).

Overview

In policy terms, the EPWP has an important role of ‘contributing to the social security net and to community development, and possibly as a stepping stone to formal employment or self-employment’ (Vaughan, Reference Vaughan2016: 12). The EPWP in South Africa operates in four distinct sectors, of which the social sector is one. The social sector (EPWP-SS) was mandated specifically to provide job opportunities for women, and provide an important service that had women and children as its primary beneficiaries (Parenzee and Budlender, Reference Parenzee, Budlender, Parenzee and Budlender2015; Skosana, Reference Skosana2013). The work opportunities that the EPWP-SS provides are in the delivery of social development and community protection services, with a primary emphasis on Early Childhood Development (ECD), such as pre-school and day care facilities for young children, and Home and Community Based Care (HCBC), which is the delivery of home-based basic health services within communities, such as administering medication and assisting with the personal hygiene of persons with illness or disabilities.

The inclusion of the social sector as a key division of the EPWP is unusual and innovative globally (Antonopoulos, Reference Antonopoulos2009; Parenzee and Budlender, Reference Parenzee, Budlender, Parenzee and Budlender2015) as public works internationally are largely dominated by infrastructure projects, especially road construction and maintenance. Its conception recognises that care work is not only confined to the private sphere of the home but also contributes to filling public infrastructural gaps and mitigating service delivery failures in the health, social welfare and education sectors (Razavi, Reference Razavi2011). Below, we argue that, as far as we can tell by using the only available data on the EPWP which is severely limited and of poor quality (EPRI, 2015), the redistributive and transformative gendered outcomes have been weak.

Is the EPWP-SS redistributive (for women)?

As the EPWP-SS specifically engages with the gendered expectations of care work as a means to engage and benefit women, it has been successful in including women as participants. In 2012, 85 per cent of the participants in the social sector were female (Vetten, Reference Vetten, Parenzee and Budlender2015). This is a higher percentage than in any of the other sectors, and higher than the quota for women participants, originally set at 60 per cent but subsequently lowered. However, some have criticised the reinforcement of gendered care roles in this sector and the use of an already female-dominated field to increase women's EPWP participation (Parenzee and Budlender, Reference Parenzee, Budlender, Parenzee and Budlender2015).

One area of moderate success is that, overall, by targeting people living in poverty including women, youth and people with disabilities, the EPWP-SS stipend has provided some poverty relief to participant households. Researchers have estimated that the programme is associated with a reduction in EPWP-SS participants living below the food poverty line from 55 to 40 per cent (EPRI, 2015). In the same study, participants indicated that their stipends ‘put food on the table’ (p.39) This is an important achievement as the EPWP in South Africa has been designed, at its most fundamental, as a programme to protect people from the worst consequences of structural unemployment (Vaughan, Reference Vaughan2016).

Nonetheless, stipends in the EPWP-SS remain unacceptably low. In addition to the consequent daily survival struggles, the building up of social or financial assets which will have positive longer term effects for participants is impracticable at this level of pay. The stipends across the EPWP vary from more than the minimum wage of R88/ day (USD6.5) stipulated in the EPWP ministerial dispensation (Mahlakoana, Reference Mahlakoana2017), to far below this minimum, indicating a persistent lack of compliance by implementing departments and their partner organisations on conditions of employment in the EPWP (EPRI, 2015). The lowest EPWP wages are in the social sector, followed by the non-state sector, which includes NGO care services. This contrasts with the infrastructural sector wages, paid at far over minimum levels (Vetten, Reference Vetten, Parenzee and Budlender2015).

The poor redistributive outcomes are compounded by serious underfunding and weak resourcing administration, with direct gendered effects. Payments made to NGOs or ECD centres are often delayed by months, creating financial stress for the participants. Precisely because the sector participants are largely female, the major implementation problems disproportionately affect women and the families they support. Similarly, NGOs are regularly not given sufficient notice about future funding. This results in a perpetual state of uncertainty, leading to institutional insecurity which has knock-on effects for participants and potentially lowers service quality due to a demoralised workforce. While EPWP-SS funding was intended to cover trainee or new workers, the problem of substitution has been noted in that posts previously paid a salary were migrated to EPWP to reduce costs to NGOs (Vetten, Reference Vetten, Parenzee and Budlender2015). These and other widely reported administration problems are a critical weakness in the EPWP-SS and appear to be qualitatively worse than in the other sectors (EPRI, 2015; Vaughan, Reference Vaughan2016; Vetten, Reference Vetten, Parenzee and Budlender2015).

The likelihood of finding secure, decently paid employment post-EPWP-SS participation is slim (van der Westhuizen, Reference van der Westhuizen, Parenzee and Budlender2015). This is due to the presence of similarly trained workers in an underfunded field with few vacancies; a challenge that all sectors of the EPWP face but one that has been particularly noted in the ECD sector. Even where employment is found, the insecure working conditions and low wages on the open labour market in the ECD, HCBC, and other social sector fields, almost guarantee that those working join the ranks of the working poor (Vetten, Reference Vetten2016). There are few existing care organisations in rural areas, where the needs are greatest and where there is a huge lack of employment opportunities for women (Jacobs et al., Reference Jacobs, Baiphethi, Ngcobo and Hart2010).

With regard to cross-policy linkages, EPWP workers are in theory eligible for unemployment insurance but, in 2015, the rate of compliance with UIF was estimated at only 50 per cent (Butler and Cartwright, Reference Butler, Cartwright and Vaughan2016) and therefore half of EPWP workers are excluded from benefits as no contributions have been made on their behalf. EPWP workers are eligible for public health services, state Old Age Grants, and the CSG yet, if they meet means-test criteria, their participation in the public works programme does not smooth access or reduce the need for negotiating administrative hurdles, even where, for example, means testing is repeated by different agencies.

The redistributive gendered effects of the EPWP-SS are therefore minimal. The low stipend for the period of participation and some training in ECD and HCBC are the only redistributive benefits. No mechanisms are in place to offer women more income security, a chance to build financial or productive assets over time, or the possibility of finding work post-participation.

Is the EPWP-SS gender-transformative?

The EPWP-SS is inherently gendered, as the care services it provides are traditionally located in women's sphere of activity. The participants themselves are primarily women, and the services to users mostly benefit women as they shoulder the burden of private and community care in the absence of public or publicly subsidised care services (Razavi, Reference Razavi2011). The simultaneous creation of work and the delivery of social services, frequently in the same communities, is an example of integrative social policy in action. The potential to doubly benefit women as workers and beneficiaries has been heralded as having the potential to contribute significantly to improving gender equality (Antonopoulos, Reference Antonopoulos2009).

The programme does increase the visibility of care work, performed largely by women. Improved and specific training delivers symbolic benefits, as it recognises the care work skills required from EPWP-SS participants. This recognition confers dignity and gravitas to a traditionally gendered and undervalued area of work. Changes in legislation have sought to define minimum stipends, working hours and leave days, though levels of compliance remain low (Vaughan, Reference Vaughan2016). The professionalisation of the care work skill-set via ECD training offers participants a sense of dignity and improved sense of self-worth (EPRI, 2015), and is thus a small step towards transformative forms of social protection (Sabates-Wheeler and Roelen, Reference Sabates-Wheeler and Roelen2011).

However, both the design and the implementation of the EPWP-SS raise important concerns from a transformative perspective. Firstly, some have argued that the inclusion of the social sector in the EPWP is opportunistic (Budlender, Reference Budlender2009), as many women were already doing HCBC and ECD work, and that the deepening of caring skills among women will not challenge traditional gender relations (Parenzee and Budlender, Reference Parenzee, Budlender, Parenzee and Budlender2015).

Furthermore, the lower stipends in sectors dominated by women suggests that, just as care work, which is highly gendered, is considered of low value in the broader economy, so it is devalued and poorly recognised in the EPWP. This is often to do with the ill-informed assumption that care work is unskilled (van der Westhuizen, Reference van der Westhuizen, Parenzee and Budlender2015). Therefore the poor working conditions and remuneration levels have served to entrench women's exploitation and institutionalise the undervaluation of care work. Truly recognising the contribution of this sector would require the provision of permanent jobs at decent wages rather than the inclusion of this work in the EPWP.

Overall, the linking of job creation needs to service delivery needs is a commendable policy innovation, with great potential for an integrated approach to women's reproductive and productive roles. Yet, several design, institutional and implementation factors hinder the transformative potential of the EPWP-SS, particularly for women. The lack of synergy across interventions and the underfunding of ECD services also limits the transformative potential of the EPWP-SS.

Towards a synergistic approach to social security and gender

We conclude with a reflection on what lessons can be learned that are relevant to other low- and middle-income countries. First, we note that social assistance has had a significant redistributive effect, with high returns to investment that have helped to narrow the gender poverty gap. The sustained and well documented impacts of the CSG over two decades are testament to the role that a relatively small, well-targeted transfer, incrementally expanded and now delivered at scale, has played in reducing the effects of poverty and unemployment, especially for women.

In a context of stagnated economic and employment growth, which significantly disadvantages women, the South African case study shows that the extension of insurance and public employment has acknowledged the gendered nature of social provision. The instruments have potential but enormous challenges in their realisation. Social insurance in South Africa remains rigid and underdeveloped. Public employment programmes have been considerably scaled up and have injected small amounts of cash into large numbers of poor households. However, institutional structures, administrative weaknesses and weak monitoring systems have limited their ability to produce and conclusively show positive outcomes for women. Research funding and advocacy to inform further analysis are needed to bring a social justice and gender-sensitive lens to the analysis of under-reported social security instruments.

From a transformative perspective, the social security programmes have made contributions to the recognition and dignity of women, by acknowledging their roles in private and public spheres, by increasing the visibility of care work and by providing important material relief for households as well as increased opportunities for women to engage in economic activities, resulting in improved confidence and skills. However, these mechanisms have concurrently reinforced rather than relieved women's social reproductive responsibilities and buttressed assumptions about the type of work (and remuneration levels) appropriate for women in the labour market. The analysis has highlighted the structural and cultural challenges to redistributing care responsibilities in more equitable ways. Across the board, the surveyed social security instruments have struggled to facilitate the economic inclusion of women while recognising, valuing and redistributing unpaid care and domestic work.

Finally, cognisant of these challenges, it is important to consider which linkages between social security instruments and across other redistributive interventions (including basic, social and health services) can contribute towards the realisation of a comprehensive social security framework. In such a framework, each programme acts as a component of a system that supports women across lifecycle stages (childhood, job-seeking, employment, pregnancy, child-bearing and child-rearing, retirement); across multiple roles (productive and reproductive) and across employment status (unemployed, informal sectors, formal sectors, full-time/part-time). Examples of such strategies range from increasing access to quality child-care services to smoothing access to other available welfare services (including parenting support and mental health), as well as access to basic education and health services (including water, sanitation, electricity, nutrition and internet). The well-established networks of social security access points would be well placed to promote such linkages. Gender-sensitive marketing of the CSG programme to advocate the engagement of men in care work could serve to counter societal norms around gender divisions of labour within the home (Khan, Reference Khan2018, forthcoming). Given the evidence that the CSG can assist with job searches, linkages between the CSG and economic empowerment programmes could be developed, together with initiatives to increase compliance and extend social insurance to self-employed and informally employed workers. Examples from other countries demonstrate that childcare policy, negotiated between private and public stakeholders, is crucial for re-orienting welfare policy in a gender-sensitive direction (Yerkes, Reference Yerkes2014). While the CSG, EPWP-SS and UIF are all core features of a relatively generous social protection landscape, there is scope going forward for continued efforts to design and implement synergies between social assistance, social insurance, public employment schemes and other social protection measures, in order to overcome cultural barriers, system failures and unemployment insecurity for the achievement of gender redistributive and transformative outcomes.