As a complement to building fiscal and financial sector resilience, countries need to establish and adapt social safety nets to ensure they are flexible enough to respond to the increased and temporary need associated with economic downturns.
With World Bank support, progress has been made to establish and strengthen social protection delivery systems and social safety net programs that can serve as the foundation for a social safety net that can become more flexible in response to a shock. However, important challenges remain in terms of enhancing the coverage, the dynamism of systems, and the flexibility and scalability of such programs.
The World Bank has expanded its support for an “adaptive social protection” approach to build household resilience and enhance the shock responsiveness of safety nets and social protection delivery systems.
Social protection mechanisms are necessary to mitigate social impacts during severe economic downturns. Without social safety nets, economic crises can drive vulnerable households into poverty, make those who are already poor even worse off, cause irreversible losses to human capital and household assets, and stir social unrest. A major lesson from the food, fuel, and financial crises between 2007 and 2009 is that social protection arrangements—including intake and registration—must be put in place during normal times and that safety nets should be more adaptable and flexible in the face of major economic downturns.
This lesson has not been fully absorbed. To be sure, the World Bank has contributed to a better understanding of which safety net programs can be used to prepare and react to economic downturns. This requires building the basis for rudimentary systems in low-income countries and promoting safety net system harmonization in middle-income countries. Although World Bank support for social safety nets has concentrated on helping countries move from fragmented programs to systems, the primary focus has been on expanding access to chronically poor people, rather than on the ability of systems to respond in an economic downturn. In the case study countries, neither the governments nor the World Bank were specifically focused on the flexibility dimension of social safety net system performance, except in Jamaica.
In advanced economies, social insurance provides effective crisis response mechanisms (automatic stabilizers) and active labor market programs that can be rapidly deployed to address increased unemployment. However, such interventions are linked to formal sector employment, which is limited in developing countries. Social safety net programs provide consumption smoothing and protection against poverty and vulnerability and are an effective instrument to address the impact of a large crisis on poor and vulnerable people (Banerjee et al. 2017; World Bank 2010, 2011, 2012). As a result, social safety nets have become a de facto social protection crisis management instrument in the developing world, even though most programs have been designed to address chronic poverty.
During the evaluation period, the World Bank stepped up financial, technical assistance, and knowledge support for safety nets in all case study countries, some of them for the first time (Mozambique, Tajikistan). Assisting clients in building delivery systems for normal and crisis times became part of the World Bank’s longer-term agenda of support, especially in low-income countries. To this end, the World Bank made substantive knowledge investments, promoted strong global partnerships, and mobilized trust funds (box 4.1) for social safety net capacity building (Bangladesh, Benin, Jamaica, Mozambique, and Tajikistan).
The World Bank conducted diagnostic work to identify gaps and strengthen social safety net systems in client countries. Strengthening these systems was thought to increase the capacity of the safety net to respond in a crisis. But the focus of this work was from the point of view of supporting chronically poor people rather than preparedness to respond to a severe economic downturn that could exacerbate poverty and generate new poor people. For example, in Morocco and Ukraine, the World Bank’s policy advice was to reduce large energy subsidies as part of a fiscal consolidation at the same time as making subsidies more targeted to poor people. Ukraine’s 2018 Public Finance Review recommended rebalancing the social assistance system to gradually make the well-targeted but smaller guaranteed minimum income scheme the primary safety net for poor people, with the housing utility subsidy becoming a secondary safety net to protect against excessive utility costs. Similar, though less extensive, reforms were carried out in Tajikistan.
In Benin and Mozambique, where initial safety nets were extremely weak or nonexistent, the World Bank focused on foundational work to identify entry points to help build targeted safety systems for chronically poor people, even though the specific programs supported (cash transfers and public works) could be expanded to provide temporary support during a cyclical downturn. This work also revealed that in low-income countries, low domestic revenue mobilization remains a major constraint to effective safety net systems and their capacity to react in a crisis (as documented in the case studies for Benin, Mozambique, and Tajikistan).
Box 4.1. The World Bank’s Global Social Safety Network
The World Bank promoted strong partnerships with donor and United Nations agencies, nongovernmental organizations, and others to build country social safety net systems with technical and financial assistance; to leverage knowledge exchange; and to improve interagency coordination. Key initiatives include the following:
- The Rapid Social Response (RSR) Program to provide catalytic resources to assist low-income countries in their response to the economic turbulence of the past decade and boost capacity for future response. By March 2013, the RSR’s $34 million fund had leveraged $3.1 billion from International Development Association for 45 projects in 34 countries, mainly in Africa (World Bank 2013). Following the global financial and economic crisis, the RSR shifted to supporting scalable social protection systems. As of December 2018, the RSR had leveraged more than $9.6 billion for social protection projects in more than 100 International Development Association–eligible countries (World Bank 2019c).
- The Social Protection Inter-Agency Cooperation Board, co-chaired by the World Bank and the International Labor Organization, and created in 2012 by the Group of Twenty countries to support countries in improving social protection systems and to promote the exchange of knowledge, policy experience, good practices, and data and information. A major achievement was the development of Inter-agency Social Protection Assessment tools, including a core social protection system diagnostic (16 countries completed; 6 in progress), identification (12 countries completed, including Morocco; 5 in progress), and social protection payments (13 countries completed, including Jamaica; 5 in progress).
- Collaboration with the International Labor Organization and the World Food Programme on the production of the Atlas of Social Protection Indicators of Resilience and Equity (ASPIRE) to benchmark the scope and performance of social safety nets, social insurance, and active labor market programs. As of May 2020, ASPIRE had data for more than 120 countries based on national administrative and household survey data (including data for all evaluation case study countries).
- The World Bank–wide Identification for Development Initiative to help countries build robust and inclusive digital identification systems through analytics, assessments, and financing. These systems are foundational to the development of social safety net registries and targeting systems. (Initiative work is ongoing in Bangladesh, Benin, Morocco, and Mozambique.)
- Promoting South-South knowledge exchange on social safety nets and systems, including cash transfer communities of practice in Africa, Asia, and Latin America and the Caribbean, and arranging knowledge forums on key topics, such as designing and delivering social protection systems for urban areas (2015) and adaptive social protection (2018).
Source: ILO n.d.; World Bank 2019c, n.d.b; World Bank ASPIRE database (http://datatopics.worldbank.org/aspire).
Safety Net Crisis Preparedness at the Program, System, and Policy Level
In the case study countries, World Bank support for safety nets has focused on interventions with the potential to play a significant role in mitigating the impact of a crisis. Program-level preparedness requires a mix of programs suitable for combating chronic poverty that can be “adapted” and quickly scaled up in response to an economic downturn. To this end, it is essential to have at least one well-established program with expansive coverage that is appropriate to address the social impacts of an economic crisis (Grosh et al. 2011). Flexible program designs allow quick adaptation to increased demand by adjusting targeting criteria; increasing beneficiaries or benefit levels; relaxing conditionalities; streamlining grievance and redress processes to manage targeting errors (which may be critical to manage social tensions during a crisis); increasing public communications so that program changes are explained; and setting “sunset clauses” to scale back programs once a crisis recedes. In Jamaica, the World Bank supported the development of a graduation, or exit, strategy for beneficiaries of its flagship cash transfer program. In Benin, Mozambique, and Tajikistan, World Bank support has been in the form of establishing cash transfers and public works programs. In Morocco, a social protection investment project was approved in 2017 to expand coverage of a unique identifying number for the Moroccan population and foreign residents and to improve targeting of social safety nets.
Support to build social protection systems has been a major World Bank priority in all case study countries. Having these systems in place provides a preparedness infrastructure on which specific programs could be adjusted or expanded. These “building blocks” enable programs to function in either normal or crisis times so that “the right people receive the right transfer amounts at the right time” at the lowest possible cost. These include outreach; intake and registration; assessment of needs and conditions; eligibility and enrollment; payment of benefits or provision of services (if any); and monitoring and management of beneficiaries, including program exit (George and Lindert 2017).
To be prepared for an economic downturn, it is essential that operational systems have the ability to quickly identify and target the newly vulnerable; that payments can be delivered quickly, predictably, and with low transaction costs to a greater number of beneficiaries; and that management information systems for beneficiary monitoring and management can ensure basic fiduciary standards, as well as interaction between individuals and program agencies along the delivery chain. Technological innovation in the use of big data, biometrics, and digital payments has greatly enabled the development of social protection systems in the past decade.
With World Bank support, Benin, Mozambique, and Tajikistan have started to put in place basic social protection systems as part of setting up cash transfer or public works programs with the longer-term goal of providing national platforms for other safety net programs. In case study countries with existing programs and institutional capacity, the World Bank is supporting the overhaul of specific system components such as intake and registration, payments, and beneficiary management (Bangladesh), and harmonization of programs and systems to exploit potential synergies and improve coverage, targeting and efficiency (Bangladesh, Morocco). In Jamaica and Ukraine, countries with better-established systems and expansive program coverage, World Bank support is leveraging the integration of core business processes for program delivery to serve as integrated platforms for multiple social protection programs and beyond, such as health insurance, scholarships, utility subsidies, and disaster risk management.
Adaptive Social Protection
Although this work is incipient in case study countries, the World Bank is supporting an increasing number of countries to incorporate an “adaptive social protection” approach. This is forming around two areas related to crisis preparedness, drawing from synergies among social protection, climate adaptation, and disaster risk management: (i) building household resilience ex ante so that poor and near-poor households are better able to withstand large shocks and (ii) enhancing the capability of social safety nets to respond to large shocks by introducing more flexible and scalable program designs and more dynamic delivery systems, including the use of early warning systems and defined triggers for activating the response and release of resources from program contingencies or additional funds (World Bank 2018b).
Overall, this evaluation finds that significant progress has been made to set up and strengthen social protection delivery systems and social safety net programs. However, social safety net preparedness across the seven case study countries varies significantly at program, system, and policy levels. Jamaica, which has received longer-term World Bank social protection support, shows the strongest adaptability at all levels, thanks in part to strong and sustained government leadership since the early 2000s. In Bangladesh, Morocco, and Ukraine, policy, systems, and programs are of moderate strength with differing weaknesses, including in terms of institutional coordination and political economy constraints, hindering implementation progress. These dimensions of flexibility are the weakest in Benin, Mozambique, and Tajikistan, where social protection systems are only nascent (table E.1).
The World Bank’s experience in the case study countries reveals several difficult challenges, especially in low institutional capacity and resource-constrained settings. Although by 2015 all countries had at least one safety net program (World Bank 2015), social safety net coverage remains limited for poor and vulnerable people, especially in low-income countries. For example, in Benin, public works and cash transfer programs cover about 3.5 percent and less than 4 percent of extremely poor households, respectively (World Bank 2017). Safety net coverage of the poorest quintile stands at close to 10 percent in Tajikistan and less than 2 percent in Mozambique (ASPIRE database, last year available). In contrast, safety net coverage of the poorest quintile in Jamaica and Ukraine stands at 72 and 47 percent, respectively (ASPIRE database, last year available).
The development of social safety net registries or databases with household- and individual-level socioeconomic information, operational data, and information to capture vulnerability is a major challenge in most case study countries. In Morocco, for example, the development of a national identification system so that people can assess their identity is necessary to build a robust social registry. The usefulness of a registry, as a basis to determine eligibility for one or more programs in a large shock situation, depends on (i) coverage of current and potential beneficiaries, (ii) relevance of the information that is collected and stored in the registry to detect newly vulnerable people, (iii) inclusion of operationally relevant information on potential beneficiaries to quickly deliver benefits in a shock situation, and (iv) having up-to-date information to represent current household circumstances.
More than 30 developing countries are integrating beneficiary registries of safety net programs, and a similar number are in the process of doing so (Barca 2017). By 2018, more than 26 African countries were using social registries and another 16 were developing them, many with World Bank support. However, most of these registries require significant expansion, strengthening, and adoption by programs to become a meaningful instrument for responding to economic downturns. Benin and Mozambique’s registries cover close to 12 and 0.3 percent of the population, respectively (Beegle et al. 2018).
A related challenge has been that traditional poverty targeting methods for social safety net programs, including proxy means tests, lack the ability to detect sudden changes in household welfare levels. In many countries, registries on which targeting mechanisms are built tend to use a mass registration approach conducted in geographic areas with high poverty once every three to five years. For crisis purposes, an approach that is on demand and on a rolling basis, where people can register at any time close to where they live, would be more appropriate. Of the case study countries, only Jamaica and Ukraine have more open registration. Although more developing countries are moving toward such a dynamic approach (such as Brazil, Chile, Georgia, and Turkey), most countries tend to have static registries due to (i) limited fiscal space, (ii) difficulty removing those who no longer qualify, (iii) rigid eligibility criteria, and (iv) limited administrative capacity, including lack of access points across the country for people to apply. Even when registries cover a large share of the population and serve many programs (Dominican Republic, Indonesia, Pakistan, Philippines), data quickly become outdated, which reduces usefulness for cyclical purposes (Leite et al. 2017).
In some countries, the lack of formal identification (foundational ID systems) inhibits the ability of people to access a range of services and social safety net transfers because they are unable to prove they are who they say they are. In Benin, the national ID registration rate stands at about 68 percent (World Bank 2017), and in Mozambique, fewer than half of births are registered (Jeronimo 2019; see box 4.1). This is a pervasive problem compounding the issue of low coverage in low-income countries.
Although an area of rapid innovation, payment systems can represent another bottleneck area since social safety net programs in many countries use a variety of delivery methods depending on the accessibility and availability of private providers (cash-in-envelope, banking system, smart cards, debit cards, prepaid cards, cell phones). For example, in Mozambique, payments are manual, and in Morocco, the post office is the authorized agency for social protection payments (Leite et al. 2017). Nevertheless, digital government-to-person payments are being increasingly adopted and have the potential for more efficient, transparent, and reliable delivery, even in low-capacity environments. In 2013, the World Bank’s Universal Financial Access 2020 initiative included Bangladesh, Morocco, and Mozambique among the 25 target countries for heightened financial inclusion engagement (ISPA 2015). In Bangladesh, a government-to-person payment system pilot is enabling the beneficiaries of three safety net programs to select among seven payment providers (Baur-Yazbeck et al. 2019).
Lack of safety net financing is another major issue, especially in the low-income case study countries. In Bangladesh, Benin, Mozambique, and Tajikistan, lack of fiscal space and/or low domestic revenue mobilization prevent safety net programs and systems from reaching sufficient scale and coverage to become a meaningful tool for crisis response. In Morocco, resource reallocation may be an option to boost financing for targeted safety net programs (World Bank 2017). Some countries are experimenting with (Jamaica) or exploring (Bangladesh, Benin, Mozambique) different strategies and instruments to ensure financing for social safety net interventions (albeit to respond to climate shocks), such as contingency funds, contingent credit lines, disaster insurance, and index-based triggers.
- For example, the Ethiopia Productive Safety Net Program has been adapted to scale up. During the Horn of Africa drought of 2011, the program supported 3.1 million additional households for three months and extended the duration of transfers to more than 85 percent of its 7.6 million beneficiaries. The Kenya Hunger Safety Net Program (HSNP) preregistered all households (374,000) in the poorest northern counties of the country, opening bank accounts and issuing debit cards for them. Only 27 percent of these households are regular HSNP beneficiaries; the others receive a one-time transfer only if they are deemed to be “at risk,” as determined by a vegetation condition index derived from satellite data. Thus, the program may temporarily scale up to 50–75 percent of the area’s population (Watson 2016, cited by Beegle 2018). As a result of the preregistration and automatic trigger, the HSNP took only two weeks between the shock and the first payment in the 2015 and 2016 droughts (Bowen et al. 2020).
- The World Bank has started to experiment with the use of remote sensing, GIS and cell phone data, machine learning, and high-resolution satellite imagery to predict poverty, which could help address some of these constraints (Bowen et al. 2020).