Gender-sensitive social protection: A critical component of the COVID-19 response in low- and middle-income countries

T o grapple with the unfolding economic crisis triggered by COVID-19, governments are increasingly turning to social protection to mitigate against widespread economic downturn and to support vulnerable populations through times of health risk and economic scarcity. According to the World Bank, as of April 17, 133 countries1 had adapted or introduced approximately 564 social protection initiatives.2 Understandably, the primary concern of these social protection responses has been to provide rapid economic assistance; gender considerations have not been at the forefront. This is unsurprising, as most existing social protection programs in lowand middle-income countries (LMICs) are either gender-blind or gender-neutral at best.3 However, such programs may inadvertently exacerbate gender inequalities.4 In the context of the COVID-19 response, this is particularly worrying, given that the pandemic is likely to pose or intensify different risks for men and women, including potentially widening existing gender inequalities.5 Examples of risks include health risks (e.g., high risk of infection due to larger shares of women in the global health workforce, reproductive health risks, and maternal mortality due to overburdened health systems); loss of jobs and livelihoods (e.g., economic impacts are likely to be more severe for informal workers and women already earning comparatively less than men); increased, disproportionate burden of care among women; and protection risks for women and girls (e.g., increased school drop-outs, pregnancy, and early marriage among adolescent girls; increased risk of gender-based violence).6 Men and women are also likely to respond to or benefit from social protection mitigation measures in different ways. Thus gender considerations in the social protection response to COVID-19 are important, given both intrinsic concerns for the holistic well-being of populations and the fact that longer-term economic development in LMICs is closely linked to women’s empowerment.7 Excellent resources with general guidance for responding to COVID-19 through social protection have already been released.8 However, more concerted efforts are needed to take gender inequalities into account. Designing gender-sensitive programming can be complex even in settings with strong social protection systems, and doing so is made more challenging by fast-moving timelines and COVID-19 mitigation approaches that complicate delivery and challenge the functioning of social protection systems. However, there are also relatively simple design and implementation adaptations that can make social protection systems more gender-responsive. SUMMARY


GENDER-SENSITIVITY AND THE EARLY COVID-19 SOCIAL PROTECTION RESPONSE
While the gender-sensitivity of social protection has been defined in different ways, we broadly consider a continuum whereby programs and systems can be considered more gender sensitive if they "recognize specific needs and priorities of women and men, and seek to purposefully and proactively tackle gender inequalities by questioning and challenging the structures, institutions and norms on which these inequalities are based." 9 A rapid assessment of the gender sensitivity of initial COVID-19 social protection responses (as of April 3, 2020) shows that, out of 418 social protection initiatives, only about 11 percent show some (but limited) gender sensitivity. 10 For example: • Six programs specifically target pregnant women or women receiving maternity benefits (Argentina, Armenia, El Salvador, Hungary, Russia, and Sri Lanka).
• Eight programs target women specifically, due to various criteria, including nutritional risk, lack of spouse, women leaders, pre-existing female beneficiaries, or top-ups to programs for women (Argentina, Brazil, Colombia, Egypt, India, Italy, and Pakistan).
• Sixteen programs specifically take into account childcare duties or provide benefits related to childcare (Austria, Cook Islands, Czech Republic, France, Germany, Italy, Norway, Poland, Romania, Serbia, Slovenia, South Korea, Spain, and the United States) • Two programs are targeted specifically to healthcare workers (who are primarily women), 11 including covering exposure or injury-related costs and compensation for infection (Philippines) and higher levels of childcare vouchers as compared to the rest of the population (Italy).
• Sixteen programs target informal workers, who are likely to be disproportionately women, through instruments including vouchers for skills training (Indonesia), wage subsidies (Australia), utility subsides (Vietnam), public works for those who lost livelihoods (Philippines), food vouchers (Jordan), and cash transfers (Argentina, Australia, Cabo Verde, Colombia, Ecuador, Morocco, Namibia, North Macedonia, Peru, Philippines, and Tunisia).

LESSONS AND CONSIDERATIONS
While there is no "one-size-fits-all" approach, based on existing evidence, we summarize key lessons, considerations, and guidance across five areas: 1 ) Adapting existing schemes and social protection modality choice, 2) targeting, 3) benefit level and frequency, 4) delivery mechanisms and operational features, and 5) complementary programming.

Adapting existing schemes and social protection modality choice
Adapting existing schemes to be contagion-safe is a likely

Targeting
Which households should be targeted for benefits and who in those households should be "named" as recipients are critical questions. Given the practicality of leveraging existing social protection programs, it makes sense to retain the original individual-level targeting of many such benefits (e.g., unemployment insurance). However, many vulnerable people will be excluded by such targeting; for example, unemployment insurance typically does not cover informal workers, including the majority of women, who primarily work in the informal economy. 22 Providing universal "house- en's empowerment when transfers were given to women rather than men. 25 In addition, while no regional statistics exist showing the current sex-disaggregation of social protection benefits, recent reviews of the evidence agree that programs that target women have led to increases in women's well-being across multiple domains. 26 Although there are fears that targeting cash transfers to women may lead to male backlash and greater risk to women, this is largely not borne out in the development literature. 27  Gender-sensitive delivery options should be considered for existing feeding and care programs -such as school meals and early child development centers -if they can be implemented within recommended safety guidelines for COVID-19. In addition, essential workers require safe childcare options, and existing child development centers may be well placed to fill this gap. However, government-run schools and early child development centers are often the first to be shut down in response to the pandemic.
Local and national NGOs have stepped in to fill some of these gaps. Several local NGOs in India, for example, have started community kitchens to deliver rations and cooked meals to the most vulnerable households and also to quarantined migrants who recently returned from urban centers as a result of the nationwide lockdown. 42 If operationally successful, these effects can pave the way for utilizing existing platforms for other services, including "school feeding" at home, while serving the dual objective of supporting local women's groups. The main channels for providing information and support services that require the least amount of physical contact and travel during the COVID-19 crisis are via telephone, internet, television, and radio. While television and the internet are widespread and have higher functionality in high-income countries, phone and radio are likely to be the best options in LMICs. For example, one-on-one support services for maternal health or mental health could be delivered as "televisits" through online platforms, phone calls from experts, hotlines for women or men to call in and speak with an expert, or live-messaging (such as WhatsApp). 44 For more generalized messages, television, radio, SMS or voice messages are another way to reach people at scale; however, studies of digital technology across different interventions and settings have produced mixed results. 45 For example, while mobile phones are ubiquitous and a promising way to continue to provide information and support services, a recent evaluation of a mobile health and nutrition program in Ghana and Tanzania (mNutrition) identifies several challenges to the mobile-phone messaging approach that should be considered. 46 First, while most households have a cell phone, women have less access to it than men.

Complementary programming
Second, women may not be able to pay for the service, so waiving any associated fees will be critical. Third, women may be illiterate and not able to read text messages; thus voice messages may be preferable to text messages and phone calls may be preferable to live-messaging. Lastly, where possible, it is important to consider the best time of day for women to receive messages via phone (or radio), given the multiple demands on women's time and that they may not always be carrying a mobile phone (or near a radio).
Thus, while mobile approaches may be ideal during the COVID-19 crisis, they must be carefully implemented and combined with broader community-level approaches that include radio and television, which may be better suited to reaching a greater number of individuals.

CONCLUDING THOUGHTS
As social protection programs and systems adapt to miti-

FURTHER GUIDANCE:
Gender-sensitive, shock-responsive social protection and emergency cash transfers

INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
A world free of hunger and malnutrition