ADAPTIVE SOCIAL PROTECTION IN SOUTHERN AFRICA ADAPTIVE SOCIAL PROTECTION IN SOUTHERN AFRICA © 2022 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, inter- pretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the infor- mation, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2625; e-mail: pubrights@worldbank.org. Cover image: © Munimara / Shutterstock. CONTENTS AC KN OW L E D G M E N TS V I I A B B R E V I AT I O N S I X 1. I N T R O D U CT I O N 1 B A C K G R O U N D A N D R AT I O N A L E 1 M E T H O D O L O G Y A N D A N A LY T I C A L F R A M E W O R K 4 S U B R EG I O NAL OVE RVI EW 9 2 . B OT S WA N A 29 OVE RVI EW O F R I S K AN D H O U S E H O L D VU L N E RAB I L ITY 29 AS S E S S M E N T O F A DA PT I V E S O C I A L P R OT E CT I O N 32 P R I O R I T Y I N T E R V E N T I O N S F O R B OT S WA N A 51 3 . E S W AT I N I 55 OVE RVI EW O F R I S K AN D H O U S E H O L D VU L N E RAB I L ITY 55 AS S E S S M E N T O F A DA PT I V E S O C I A L P R OT E CT I O N 57 P R I O R I T Y I N T E R V E N T I O N S F O R E S W AT I N I 74 4. L E S OT H O 79 OVE RVI EW O F R I S K AN D H O U S E H O L D VU L N E RAB I L ITY 79 AS S E S S M E N T O F A DA PT I V E S O C I A L P R OT E CT I O N 82 P R I O R I T Y I N T E RV E N T I O N S F O R L E S OT H O 1 07 iii iv ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA 5. SOUTH AFRICA 111 OVE RVI EW O F R I S K AN D H O U S E H O L D VU L N E RAB I L ITY 111 AS S E S S M E N T O F A DA PT I V E S O C I A L P R OT E CT I O N 114 PRIORITY INTERVENTIONS FOR SOUTH AFRICA 139 APPENDIX 143 G LO S S A RY 1 4 9 REFERENCES 153 B OX E S 1 .1 L E S OT H O ’S S O C I A L R E G I ST RY A N D S H O C K R E S P O N S E 20 3 . 1 T H E N AT I O N A L S O C I A L A S S I S TA N C E P O L I C Y 59 3 . 2 OVC C A S H T R A N S F E R P I LOT ( 2 0 1 6 –1 8 ) 65 3 . 3 F O O D R E S P O N S E P L A N F O R C OV I D -1 9 66 3 . 4 A C C E S S T O F I N A N C I A L S E R V I C E S A N D N AT I O N A L I D C A R D S I N E S W AT I N I 6 7 3 . 5 A C T I O N S P R O P O S E D T O I M P R O V E E S W AT I N I ’ S I N F O R M AT I O N S Y S T E M S 70 4 . 1 I N T E G R AT E D P O L I C Y F R A M E W O R K F O R A L I G N I N G T H E G O V E R N M E N T O F L E S O T H O ’ S C O O R D I N AT I O N S T R U C T U R E S 85 4 . 2 A C C E S S T O F I N A N C I A L S E R V I C E S A N D N AT I O N A L I D C A R D S I S S T I L L L I M I T E D A N D I N E Q U I I N L E S OT H O 93 4.3 L E S OT H O ’S D I S AST E R R I S K F I N A N C I N G I N ST R U M E N TS 105 4 .4 C O N T I N G E N CY P L A N N I N G F O R C G P S CA L E - U P S I N R E S P O N S E TO DROUGHTS 106 5 . 1 S O U T H A F R I C A’ S S O C I A L P R O T E C T I O N R E S P O N S E T O C O V I D - 1 9 126 5 . 2 M I C R O I N S U R A N C E A S A S H O C K- R E S P O N S I V E S O C I A L P R OT E CT I O N O PTI O N 128 FIGURES 1 .1 C O U N T R I E S A N D T E R R I TO R I E S U S I N G S O C I A L P R OT E CT I O N M E AS U R E S I N R E S P O N S E TO C OV I D -1 9 4 1 . 2 T H E F O U R B U I L D I N G B LO C KS F O R A DA PT I V E S O C I A L P R OT E CT I O N 5 1 . 3 D E V E L O P M E N T T R A J E C T O R Y O F S O C I A L P R O T E C T I O N I N F O R M AT I O N SYST E M S 1 8 2 . 1 B O T S W A N A’ S C U R R E N T D I S A S T E R R I S K M A N A G E M E N T O R G A N I Z AT I O N A L FRAMEWORK 36 2 . 2 T O TA L A N D P O O R P O P U L AT I O N S C O V E R E D B Y B O T S W A N A S O C I A L A S S I S TA N C E P R O G R A M S , 2 0 1 5 / 1 6 ( % ) 41 2 . 3 S O C I A L A S S I S TA N C E P R O G R A M E X P E N D I T U R E S A S P E R C E N TA G E O F B O T S W A N A’ S T O TA L B U D G E T 49 C ontents v 3 . 1 P O V E R T Y H E A D C O U N T R AT I O S I N S A C U C O U N T R I E S 55 3 . 2 D I S A S T E R R I S K P R O F I L E O F E S W AT I N I 57 3 .3 S O C I A L P R OT E CT I O N C OV E R AG E I N S AC U C O U N T R I E S : D I R E CT A N D I N D I R E CT B E N E F I C I A R I E S 64 3 . 4 S A C U S O C I A L A S S I S TA N C E S P E N D I N G A S A P E R C E N TA G E O F G D P 73 3 . 5 D I S T R I B U T I O N O F S O C I A L A S S I S TA N C E S P E N D I N G I N E S W AT I N I 73 4 . 1 P O V E R T Y H E A D C O U N T R AT I O S I N S A C U C O U N T R I E S 79 4.2 P OVE RTY I N M I D D L E- I N C O M E C O U NTR I E S 80 4 . 3 R U R A L A N D U R B A N W E L F A R E D I S T R I B U T I O N A N D P O V E R T Y B Y D I S T R I C T 8 0 4 . 4 P O P U L AT I O N A F F E C T E D B Y C L I M AT I C S H O C K S I N L E S O T H O , 1 9 6 8 – 2 0 2 0 8 1 4.5 F O O D I N S E C U R I T Y I N L E S OT H O, 2 01 4 – 2 0 81 4.6 S O C I A L P R OT E CT I O N C OV E R AG E I N S AC U C O U N T R I E S : D I R E CT A N D I N D I R E CT B E N E F I C I A R I E S 91 4 . 7 C O R R E L AT I O N B E T W E E N C G P C O V E R A G E A N D F O O D D E F I C I T S I N L E S OT H O, BY D I ST R I CT 91 4 . 8 B E N E F I T A D E Q U A C Y O F A L L S O C I A L A S S I S TA N C E P R O G R A M S I N S A C U COUNTRIES 92 4 . 9 S O C I A L A S S I S TA N C E S P E N D I N G A S A P E R C E N TA G E O F G D P F O R S A C U A N D OT H E R C O U N T R I E S 103 5.1 S O U T H A F R I CA : F I N A N C I A L C O ST O F D I S AST E R S , 1 9 6 8 – 2 01 9 112 5 . 2 S O U T H A F R I C A’ S S O C I A L P R O T E C T I O N S Y S T E M 119 5 . 3 D I S T R I B U T I O N O F S O C I A L P R O T E C T I O N G R A N T S B Y W E A LT H Q U I N T I L E IN SOUTH AFRICA, 2014/15 122 MAPS 2 .1 B OT S WA N A 31 5 . 1 H A Z A R D L E V E L S F O R VA R I O U S T Y P E S O F F L O O D I N G I N S O U T H A F R I C A , 2021 113 5 . 2 S O U T H A F R I C A S TA N D A R D P R E C I P I TAT I O N I N D E X , 2 0 1 6 113 5 . 3 D AY S W H E N F I R E D A N G E R I S “ V E R Y L I K E LY ” I N S O U T H A F R I C A 114 TA B L E S 1 .1 S C O R E C A R D F O R T H E F O U R B U I L D I N G B LO C KS AC R O S S C O U N T R I E S 10 1 . 2 S C O R E C A R D F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S AC R O S S C O U N T R I E S 11 1 .3 S C O R E CA R D F O R P R O G R A M S A N D D E L I V E RY SYST E M S AC R O S S COUNTRIES 15 1 . 4 S C O R E C A R D F O R D ATA A N D I N F O R M AT I O N S Y S T E M S A C R O S S COUNTRIES 17 1 .5 U T I L I T Y O F A N D T RA D E- O F FS B E T W E E N B E N E F I C I A RY A N D S O C I A L R EG I STR I ES FO R S H O C K R ES PO N S E 19 vi ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA 1.6 S C O R E CAR D FO R F I NAN C E AC R O S S C O U N T R I E S 24 2 . 1 I N C I D E N C E A N D I M PA C T O F D I S A S T E R E V E N T S I N B O T S W A N A , 1965–2018 30 2 . 2 S C O R E C A R D F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S I N B OT S WA N A 33 2 . 3 S C O R E C A R D F O R P R O G R A M S A N D D E L I V E R Y SYS T E M S I N B OT S WA N A 38 2 . 4 S O C I A L A S S I S TA N C E P R O G R A M S I N B O T S W A N A 40 2 . 5 S C O R E C A R D F O R D ATA A N D I N F O R M AT I O N S Y S T E M S I N B O T S W A N A 46 2 . 6 S C O R E C A R D F O R F I N A N C E I N B OT S WA N A 49 2 . 7 S U M M A R Y O F A L L R E C O M M E N D AT I O N S B Y B U I L D I N G B L O C K F O R B OT S WA N A 5 3 3 . 1 I N C I D E N C E A N D I M PA C T O F D I S A S T E R E V E N T S I N E S W AT I N I , 1 9 8 1 – 2 0 1 9 5 6 3 . 2 S C O R E C A R D F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S I N E S W AT I N I 58 3 . 3 S C O R E C A R D F O R P R O G R A M S A N D D E L I V E R Y S Y S T E M S I N E S W AT I N I 62 3 . 4 M A I N S O C I A L A S S I S TA N C E P R O G R A M S I N E S W AT I N I 63 3 . 5 S C O R E C A R D F O R D ATA A N D I N F O R M AT I O N S Y S T E M S I N E S W AT I N I 69 3 . 6 S C O R E C A R D F O R F I N A N C E I N E S W AT I N I 72 3 . 7 S U M M A R Y O F A L L R E C O M M E N D AT I O N S B Y B U I L D I N G B L O C K F O R E S W AT I N I 7 6 4 . 1 S C O R E C A R D F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S I N L E S OT H O 82 4.2 S C O R E CA R D F O R P R O G R A M S A N D D E L I V E RY SYST E M S I N L E S OT H O 89 4 . 3 M A I N S O C I A L A S S I S TA N C E P R O G R A M S I N L E S O T H O 90 4 . 4 S C O R E C A R D F O R D ATA A N D I N F O R M AT I O N S Y S T E M S I N L E S O T H O 98 4 . 5 U T I L I T Y O F N I S S A D ATA F O R I N F O R M I N G S H O C K R E S P O N S E S 101 4.6 S C O R E CA R D F O R F I N A N C E I N L E S OT H O 103 4 . 7 S U M M A R Y O F A L L R E C O M M E N D AT I O N S B Y B U I L D I N G B L O C K F O R L E S OT H O 1 0 9 5 . 1 I N C I D E N C E A N D I M PA C T O F D I S A S T E R E V E N T S I N S O U T H A F R I C A , 1952–2019 112 5 . 2 S C O R E C A R D F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S IN SOUTH AFRICA 115 5.3 S C O R E CA R D F O R P R O G R A M S A N D D E L I V E RY SYST E M S I N S O U T H AFRICA 120 5 . 4 S O C I A L A S S I S TA N C E G R A N T S A N D P R O G R A M S I N S O U T H A F R I C A 121 5 . 5 S C O R E C A R D F O R D ATA A N D I N F O R M AT I O N S Y S T E M S I N S O U T H A F R I C A 1 2 9 5 . 6 E A R LY W A R N I N G S Y S T E M S I N P L A C E F O R M O S T F R E Q U E N T S H O C K S I N SOUTH AFRICA 132 5.7 S C O R E CA R D F O R F I N A N C E I N S O U T H A F R I CA 134 5 . 8 S U M M A R Y O F A L L R E C O M M E N D AT I O N S B Y B U I L D I N G B L O C K F O R S O UTH AF R I CA 141 AC KN OW L E D G M E N TS T his report was prepared by a team the Caribbean: Building Human Capital for that was led by Thomas Bowen Resilience” (Beazley and Williams 2021). (Social Protection Specialist) and In addition, this report was developed in included coauthors Rodolfo Beazley tandem with the World Bank report Stress (Consultant) and Catherine Fitzgibbon (Con- Testing Social Protection (2021g), benefit- sultant). The report’s extended team also ing substantively from a mutual exchange on included the social protection task team assessment approaches during the drafting leads for its assessment countries: Vic- process. Of note, the framework developed toria Monchuk (Senior Economist), Melis in Stress Testing Social Protection has since Guven (Senior Social Protection Economist), been finalized and will now be the World Emma Monsalve (Social Protection Spe- Bank’s standardized ASP country assess- cialist), and Indira Bongisa Lekezwa (Social ment tool. Protection Specialist). Barry Maher (Senior Financial Sector Specialist) provided over- Each country assessment of adaptive social sight and quality control for the sections on protection in this report also draws signifi- risk finance. cantly upon recently finalized assessments of social assistance in the four countries The report and its methodology have bene- covered. These include “Botswana Social fited greatly from several related reports and Protection Programs and Systems Review” initiatives that define and measure adaptive (World Bank 2022a); “Social Assistance Pro- social protection. First, the report applies grams and Household Welfare in Eswatini” the definitions and framework developed by (Raju and Younger 2021); “Lesotho: Social Bowen et al. (2020). More recently, several Protection Programs and System Review” studies have applied the framework outlined (World Bank 2021b); and “South Africa: by Bowen et al. to the analysis of country Social Assistance Programs and Systems systems, preparedness measures, and the Review” (World Bank 2021f). required investments for adaptive social protection, on which this report builds. The Finally, the sections on financing adaptive report also features the traffic light assess- social protection drew upon analysis recently ment approach developed by Beazley and undertaken in each country by the World Williams in “Adaptive Social Protection in Bank Group’s Finance, Competitiveness & vii viii ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Innovation Global Practice and presented in helped to sharpen the report and highlighted “Eswatini: Disaster Risk Finance Diagnostic,” important additional information and per- “Lesotho: Disaster Risk Financing Diagnos- spectives that enriched it. tic,” and “South Africa: Disaster Risk Finance Diagnostic” (World Bank 2022b, 2019b, and Lastly, the authors would like to sin- 2021e). cerely thank Lisa Ferraro Parmelee for her thorough, patient, and precise editorial assis- Overall guidance and quality control for this tance throughout the production process, as report were provided by Marie Françoise well as Nita Congress, who provided graphic Marie-Nelly (Country Director), Asmeen Khan design support and publication quality con- (Operations Manager), and Paolo Belli (Prac- trol to finalize the publication. tice Manager), while comments received during the peer review process from Evie Calcutt (Financial Sector Specialist), Sarah Coll-Black (Senior Economist), and Asha Wil- liams (Senior Social Protection Specialist) A B B R E V I AT I O N S ASP adaptive social protection NGO nongovernmental organization ATM automated teller machine NISSA National Information System for Social Assistance CGP Child Grants Program NISPIS National Integrated Social Protection DMA Disaster Management Agency Information System DMC disaster management committee NSAP National Social Assistance Policy DPMO Deputy Prime Minister’s Office NSPF National Social Protection Framework DRM disaster risk management NSPS National Social Protection Strategy E lilangeni (Eswatini currency) OVC orphans and vulnerable children ECHO Directorate-General for European P pula (Botswana currency) Civil Protection and Humanitarian Aid Operations PMT proxy means test GDP gross domestic product PPP purchasing power parity HEA Household Economy Approach R rand (South African currency) ID identity card SADC Southern Africa Development Community IPC Integrated Food Security Phase Classification SACU South African Customs Union LVAC Lesotho Vulnerability Assessment SASSA South Africa Social Security Agency Committee SOBERS Social Benefits Registration System M maloti (Lesotho currency) SOCPEN Social Security Pension System MIS management information system SRD Social Relief of Distress MLGRD Ministry of Local Government and SSR single social registry Rural Development TERS Temporary Employer-Employee Relief MOSD Ministry of Social Development Scheme NCDM National Committee on Disaster UN United Nations Management UNICEF United Nations Children’s Fund NDMA National Disaster Management Agency VAA vulnerability assessment and analysis NDMO National Disaster Management Office WFP World Food Programme NDP national development plan All dollar amounts are U.S. dollars unless otherwise indicated. ix I N T R O D U CT I O N 1 BAC KG R O U N D influence of climate change and the increas- ingly erratic weather patterns it is likely to A N D R AT I O N A L E bring. The countries of the Southern Africa Customs Climatic shocks such as these tend to affect Union (SACU)—Botswana, Eswatini, Leso- the poorest most, exacerbating inequalities tho, Namibia, and South Africa—are exposed and increasing poverty. While international to climatic shocks, especially drought, that poverty rates remain low in the SACU rela- pose a continual threat to lives and livelihoods tive to other Sub-Saharan African countries,1 across the subregion. The El Niño event of inequality is notably high and persistent. 2015/16 resulted in the worst drought there Based on the Gini coefficients of con- in 35 years (World Bank 2022c), which sumption (or income), Botswana, Eswatini, was followed in 2019 by another particu- Namibia, and South Africa are among the top larly severe drought—just prior to the onset 15 of the 164 countries in the World Bank’s of the COVID-19 pandemic. In Lesotho, the global poverty database. Despite recent government declared a national disaster, as improvements, Lesotho also ranks among a fourth of the country’s population faced the top 20 percent of the world’s most eco- severe food insecurity. Namibia meanwhile nomically unequal countries (World Bank suffered its worst drought in 90 years, caus- 2022c). At the same time, estimates sug- ing a third of the population to go hungry, gest the consumption losses experienced with a state of emergency declared for the by households from a climate shock can be second time in three years. In Botswana, the substantial among the SACU countries— rains that year were described as the spars- on average, the affected population suffers est the country had seen in three decades losses in per capita consumption of 11.7 per- (IFRC 2021). In line with global trends, data cent from a drought and 13.2 percent from a indicate interannual weather variability in flood (World Bank 2021d). In line with global the SACU countries has been increasing, findings, evidence also indicates that the and the droughts have been more intense and wider spread (see, for example, Ala- hacoon et al. 2022; Fauchereau et al. 2003). This trend toward more frequent and more 1 Poverty rates are measured according to the global poverty line of $1.90 per day in 2011 pur- severe climatic shocks will worsen under the chasing power parity (PPP) U.S. dollars. 1 2 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA impacts from climatic shocks are distributed Food insecurity—which is chronic in the unequally in the SACU countries, with poorer subregion and both a root cause of vulnera- households tending to be more exposed and bility to drought and an outcome of it—also vulnerable (World Bank 2021d). With limited increased as a result of impacts from the pan- assets to draw upon after a shock, poorer demic. In January 2020, before the onset of households often resort to negative coping COVID-19, the World Food Programme (WFP) mechanisms, including cutting consumption, estimated that as many as 45 million people selling productive assets, availing them- were food insecure across the SACU coun- selves of high-interest loans, and removing tries due to ongoing drought conditions, children from school (del Ninno, Coll-Black, with Eswatini, Lesotho, and Namibia found and Fallavier 2016; Hallegatte et al. 2016; to have been the hardest hit (WFP 2020c). Hill, Skoufias, and Maher 2019). While these With the additional effects of the pandemic, are short-term mitigation strategies driven food insecurity is expected to deteriorate by necessity, the loss of productivity and even further. For example, a March 2020 divestment in human capital they produce rapid assessment by the Lesotho Vulner- can result in slower recovery and long-term ability Assessment Committee found that harm. more than 90 percent of rural households in Lesotho had depleted their food stocks The pandemic has compounded these as a result of drought and were depending existing vulnerabilities. The World Bank largely on food purchased at the onset of the Macro Poverty Outlook recently projected pandemic. Indeed, around 26 percent of the increased poverty rates in each of the SACU country’s population (374,000 people) faced countries. Globally, the COVID-19 pandemic high food insecurity from July to Septem- is estimated to have pushed an additional ber 2020, and around 40 percent (582,000 97 million people into poverty in 2020 (World people) were projected to be in crisis (IPC Bank 2021h). The latest estimates for 2021 Phase 3) between October 2020 and March suggest this substantial increase had begun 2021, with pockets of highly vulnerable pop- to decline slowly (from 732 million in 2020 ulations in emergency (IPC Phase 4; WFP to 711 million in 2021), but has remained far percent 2020b).2 Similarly, in Eswatini, 37  above pre-pandemic poverty reduction esti- of the rural population (around 335,420 mates (631 million). The rising trend has held people) were expected to experience acute true as well across the SACU subregion. food insecurity between October 2020 and Inclusive of mitigation measures, in Leso- March 2021, with 280,201 classified as in tho the $1.90 poverty rate was expected to crisis and 55,219 as in emergency (WFP have increased from 27.7 percent in 2019 to 2020a). 30.3 percent in 2021. In Botswana, based on the upper-middle-income country poverty Social safety net programs can help poor line ($5.50), the poverty rate was projected and vulnerable households manage the risks percent in 2020 (roughly to reach 57.6  1.5 million people)—a record for the country. In South Africa, around 2 million people may 2 The Integrated Food Security Phase Classifi- have fallen into poverty in 2021, bringing cation (IPC) is a multipartner initiative using the $5.50 poverty rate there to 59 percent internationally recognized scientific standards to classify countries’ food security across five (World Bank 2021h). phases: 1: food security, 2: stressed, 3: crisis, 4: emergency, and 5: catastrophe. 1 :   I ntroduction 3 they face from shocks, helping to mitigate households, backed by an e-payment the impacts on poverty and food insecurity, system, early warning information, and risk but their effectiveness can be constrained in financing. In Mauritania, as part of the Sahel several ways. Impact evaluations of safety Adaptive Social Protection Program, a cash net programs in six African countries have transfer program called Tekavoul is comple- described “unambiguous” increases in the mented by a dedicated emergency response food security of beneficiary households program, Elmouna, that provides the flex- (Asfaw and Davis 2018). In Ethiopia, the Pro- ibility to reach beyond the core safety net ductive Safety Net Program reduced the caseload. Elmouna was developed in part- initial impact of a drought on beneficiaries by nership with humanitarian actors and is 57  percent, eliminating the adverse effects informed by the subregional food insecurity on food security within two years (Knippen- assessment system, assessment system, the berg and Hoddinott 2017). Furthermore, a Cadre Harmonize. In Senegal, in addition to recent study found that, while household enabling it to respond to lean season fluc- food insecurity in Ethiopia increased by tuations, ex ante investments have prepared 11.7 percentage points following the onset the social protection system to respond to of the COVID-19 pandemic, participation in flooding in Dakar. The response has been the Productive Safety Net Program reduced made possible through contingency plan- the likelihood of becoming food insecure ning and standard operating procedures, to only 2.4 percent (Abay et al. 2020). Evi- as well as preprepared household damage dence has repeatedly shown that social assessments linked to the social registry. assistance can also help reduce the need to resort to negative coping strategies that trig- Fundamental gaps in social protection cov- ger longer-term detrimental effects (Barca erage and rigidity in delivery systems in et al. 2015; Dammert et al. 2018; Hill, Skou- SACU countries constrain the ability to pro- fias, and Maher 2019). As has been vividly vide social protection to persons affected highlighted in many countries since the start by shocks. Despite progress in the devel- of the pandemic, however, sizable gaps opment of social protection systems in in social protection coverage often exist, southern Africa in recent years, many are which, when combined with rigid delivery still underdeveloped, with only limited forms systems and limited information systems, of social protection offered, inadequate prevent social protection programs from coverage of the poor, low benefit amounts, reaching nonbeneficiaries with assistance and some examples of manual adminis- after a shock hits. trative mechanisms. Moreover, as noted in a recent World Bank report on inequality Several governments in Sub-Saharan Africa, among SACU countries, social protection especially those of countries that are highly programs as currently designed cover only exposed to climatic shocks, are increas- a small fraction of climate-vulnerable house- ingly undertaking ex ante investments to holds in the subregion (World Bank 2022c). overcome these constraints. In Kenya, for Relatedly, the WFP has emphasized that the example, the Hunger Safety Net Program, effect of recurrent droughts in the subregion which operates in drought-prone dis- highlights the need for more adaptive social tricts, was developed with the capability to protection (ASP) and anticipatory actions expand its caseload rapidly after a drought to ensure adequate coverage of vulnera- by up to an additional 302,000 pre-targeted ble and potentially vulnerable households 4 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA (WFP 2019). Social protection systems gen- were effective in substantially increas- erally have not, however, been designed or ing the incomes of the poor and that the prepared to respond to climatic shocks in grants themselves were pro-poor (Köhler most of the SACU countries, despite the and Bhorat 2020), indicating a significant frequency and predictability of their occur- mitigation of the impact of the crisis on pov- rence and people’s vulnerability to them. erty. These COVID-19 responses and the challenges encountered in their mobiliza- The mobilization of social protection in tion provide an entry point for examining the response to COVID-19 and the challenges social protection systems more closely and that have emerged to that mobilization have considering the ways in which they may be strengthened the case for investments in better prepared to respond to future crises preparedness ahead of future shocks. Glob- through ASP. ally, between March 20, 2020, and May 14, 2021, a total of 3,333 social protection mea- sures were planned or implemented in 222 countries or territories (figure 1.1), of which M E T H O D O LO GY 1,841 were social assistance programs. A N D A N A LY T I C A L Among these, 734 cash-based programs FRAMEWORK were planned or implemented in 186 coun- tries (Gentilini et al. 2021). In line with this Adaptive social protection refers to an global social protection response, each of agenda for preparing social protection sys- the countries in southern Africa mobilized tems to improve their response to shocks social protection responses to COVID-19 in and to build the resilience of poor and vulner- one form or another, in an attempt to ame- able households. The World Bank defines liorate the impact of the crisis on poverty ASP in the following way: “Adaptive social and food insecurity. The response in South protection helps to build the resilience of Africa was notably large, with initial evi- poor and vulnerable households by invest- dence finding that social assistance grants ing in their capacity to prepare for, cope F I G U R E 1 .1   Countries and territories using social protection measures in response to COVID-19 250 3,333 3,600 222 Number of countries/territories 212 215 3,100 200 195 200 190 181 2,600 171 Number of measures 159 151 2,100 150 126 133 1,414 1,600 106 100 1,024 1,055 1,179 84 870 937 1,100 752 803 564 685 418 505 600 50 45 283 103 100 0 −400 20 Mar 27 Mar 3 Apr, 10 Apr 17 Apr 24 Apr 1 May 8 May 15 May 22 May 12 Jun 10 Jul 18 Sep 11 Dec 14 May 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2021 Source: Gentilini et al. 2021. 1 :   I ntroduction 5 with, and adapt to shocks, protecting their FIGURE 1.2  The four building blocks for adaptive social wellbeing and ensuring that they do not protection fall into poverty or become trapped in pov- erty as a result of the impacts” (Bowen et al. 2020). While the prescription for the devel- opment of ASP is highly context specific, Data and depending on the risk profile of the country information and its social protection system, prepared- ness investments tend to be concentrated in four key areas, represented here as build- ing blocks (figure 1.2). These four areas are institutional arrangements and partnerships, Finance Programs programs and their delivery systems, data and information systems, and finance. This report takes stock of ASP in four of the Institutional five SACU countries—Botswana, Eswatini, arrangements and Lesotho, and South Africa—and provides tar- partnerships geted recommendations for each country’s development.3 A “traffic light” scorecard is used to summarize the status of each ASP Source: Bowen et al. 2020. building block for each of the four countries. Some of the dimensions assessed relate to what can be considered foundational social protection systems and, occasionally, foun- and measures represent the context within dational disaster risk management (DRM) which ASP could be developed and imple- systems. These dimensions represent the mented and offer insights into the extent to foundations that underpin the general which this context constrains or enables this capacity of a country to respond to shocks development. Additionally, the assessment with social protection, but which the country reviews specific adaptive dimensions and is not necessarily investing in or conceiv- preparedness measures. These are more ing of for that specific purpose. Examples specific inputs—going beyond “business might include a social registry or electronic as usual” investments in social protection cash transfer payment system, which can (or DRM)—that are essential for explicitly identify beneficiaries and deliver benefits to preparing social protection to respond to them in normal times as well as in response shocks. to shocks, or a national early warning system for the dissemination of risk information While the full suite of social protection instru- and the mobilization of governmentwide ments includes social assistance, social shock responses. Such foundational inputs insurance, and labor market programs—all of which can help households manage climate risk to varying extents—this assessment 3 Namibia was outside the scope of this report concentrates on social assistance programs. due to budget limitations, but assessment can This focus reflects the prominent role played easily be extended there in the future for a com- plete understanding of the SACU subregion. by social assistance programs (such as cash 6 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA and in-kind transfers and public works pro- backed by constitutional and legal instru- grams) in response to both the COVID-19 ments, that acknowledge the role of social pandemic and a growing number of cli- protection in responding to shocks. These mate shocks within southern Africa, as well policies, structures, and mechanisms can as more widely. It also reflects the compar- create an environment that enables the nec- ative advantage of these programs in terms essary leadership and coordination among of their ability to reach the poorest and most social protection programs, the DRM sector, vulnerable in times of crisis by virtue of their the actors operationalizing climate change objectives and eligibility criteria. Finally, this adaptation initiatives, and humanitarian narrowing of scope allows this report to actors. build upon social assistance assessments recently completed by the World Bank in The following are the foundational and each of the four countries, expanding their adaptive indicators assessed under Build- findings specifically with regard to the shock ing Block 1, with examples of the questions responsiveness of those same social assis- associated with them. tance programs. F O U N D AT I O N A L The specific foundational and adaptive A S S E S S M E N T I N D I C AT O R S measures to be assessed within each build- ing block for each of the four case study ● Social protection policy and legislative countries are discussed briefly below. Fur- frameworks. Do these exist? Are they thermore, the appendix outlines the detailed relevant and up to date? scoring criteria for each, according to ● Social protection institutional capacity “nascent,” “emerging,” and “established” cat- and coordination. What is the adminis- egorizations. It must be emphasized that this trative (human, physical, and financial) is a qualitative assessment designed to bring capacity of social protection institutions structure to the analysis, and the ratings at different levels? What is the extent of assigned throughout should be regarded as their ability and influence to coordinate indicative and instructive but not definitive. action with others? ● DRM policy and legislative frameworks. BUILDING BLOCK 1: Do these exist? Are they relevant and up INSTITUTIONAL to date? ARRANGEMENTS AND ● DRM institutional capacity and coordina- PA R T N E R S H I P S tion. What is the administrative (human, physical, and financial) capacity of DRM The dimensions assessed under institu- institutions at different levels? What is tional arrangements and partnerships reflect the extent of their ability and influence to the fact that ASP is an inherently multisec- coordinate action with others? toral undertaking. Coordination is required across a broad range of actors that are ADAPTIVE AS S E S S M E NT engaged in building the resilience of vulner- I N D I C AT O R S able households and the myriad actors that are mobilized after shocks hit. To be effec- ● Clear mandate for ASP. To what extent is tive, ASP requires the adoption of policies, the use of social protection in response 1 :   I ntroduction 7 to/in advance of shocks explicitly sup- shocks, meaning they may require less sup- ported by national social protection, port after a shock hits. To do so, appropriate DRM, and climate change adaptation types of assistance need to reach those policies and strategies? identified as being the most vulnerable to ● Multisectoral coordination for shock shocks in a timely and predictable manner. response. To what extent are DRM, social protection, or other coordination The following are the foundational and adap- mechanisms in place to support effec- tive indicators under Building Block 2. tive vertical and horizontal coordination of ASP actions?4 F O U N D AT I O N A L ● Partnerships with nongovernmental A S S E S S M E N T I N D I C AT O R S actors. What mechanisms and agree- ● Coverage of main programs. What shares ments are in place for coordination with of the population generally and of the humanitarian actors, development part- poor specifically are already covered by ners, and civil society? What have recent social protection programs? partnership experiences been like? ● Benefit adequacy. Do core and reg- ular transfers address households’ BUILDING BLOCK 2: consumption needs, thereby building PROGRAM DESIGN AND their resilience to shocks? D E L IVE RY ● Eligibility criteria and targeting. Are there Dimensions assessed under Building Block 2 clear eligibility criteria and processes for represent the fundamental aspects of the selecting and registering beneficiaries? design and delivery of existing social protec- ● Financial inclusion and national ID. What tion programs and the extent to which they is the coverage of these key enablers for have successfully responded to shocks in registration, payment delivery, and other the past. Programs may have responded to processes, and how sophisticated are previous shocks through vertical expansion they? (increased benefit amounts) and/or horizon- ● Social protection payment modalities. tal expansion (increased coverage). Often How effective are modalities to deliver countries have learned by doing, and these cash (or other) transfers to beneficiaries? experiences have resulted in adjustments to ● Delivery processes and systems. What social protection program design and sys- guidance, protocols, manuals, and so on, tems and/or new programs to fill gaps in exist to administer and operate the social coverage and enable additional beneficia- protection system? ries to be reached in times of crisis. Over the longer term, social protection programs can also build the resilience of beneficiaries to ADAPTIVE AS S E S S M E NT I N D I C AT O R S ● Experience of vertical expansion 4 Vertical coordination refers to coordination (increased benefits). Have any social among actors operating at the national level protection programs previously provided and actors at the subnational level. Horizontal additional temporary support during or coordination is that among the different actors operating at the national level. after shocks to existing beneficiaries? 8 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA How effectively was this done, including F O U N D AT I O N A L in terms of timeliness and adequacy of A S S E S S M E N T I N D I C AT O R S the response? ● Beneficiary registries. What registries or ● Experience of horizontal expansion databases exist for current social protec- (increased coverage). Have any social tion programs? Are they paper based or protection programs previously provided digital? Are digital registries integrated/ additional temporary support during or interoperable across social protection after shocks to additional beneficiaries? programs? How valid, accurate, relevant How effectively was this done, including (for ASP), and current are the data in the in terms of timeliness and adequacy of registries? the response? ● Social registries. Are social protec- ● Protocols and contingency plans. Do any tion databases integrated into social protocols or plans exist setting out when registries that include details on wider and how social protection programs populations or those receiving other gov- should respond to shocks? ernment services? ● Data and information on disaster risk BUILDING BLOCK 3: and household vulnerability. What early D ATA A N D warning and disaster monitoring sys- I N F O R M AT I O N S Y S T E M S tems exist? How well are the data they gather analyzed and used to forecast The strengthening of data and information or mitigate shocks? What data exist on systems is central to the ASP agenda, with an household-level poverty, vulnerability, improved understanding of risk and house- and exposure to shocks? How valid and hold vulnerability central to preparing and current are the data? mobilizing effective responses. Dimensions 3 fall into assessed under Building Block  ADAPTIVE AS S E S S M E NT two broad categories. The first concerns the I N D I C AT O R S quality of management information system (MIS) data held on existing social protection, ● Use of preexisting data to inform shocks. as well as potential ASP, clients. Although How have early warning data and his- interest in social registries has been grow- torical weather information been used ing, few countries in Africa have developed to inform ASP or other preparedness mature systems, and levels of digitization measures and shock responses? Have vary widely. The second key area concerns social protection databases been used to the quality of data and information on house- inform ASP? hold vulnerability and resilience to shocks. ● Post-shock household needs assess- As systems for monitoring and tracking haz- ments. What is the capacity to undertake ards exist in all the countries examined, the postdisaster needs assessments? Are dimensions scored address the extent to roles and responsibilities clear? Are which they inform ASP design, response, there clear guidelines and protocols for and financing. assessing needs that can be quickly translated or linked to ASP responses? The following are the foundational and adap- tive indicators under Building Block 3. 1 :   I ntroduction 9 ● Data-sharing protocols with internal/ countries or at a similar stage of devel- external partners. Are systems and pro- opment? What portion of the resources tocols in place to share data across are for core social protection benefi- departments to facilitate response? Are ciaries, and what portion are for ASP there processes to share data inter- beneficiaries? nally and externally with other service providers? ADAPTIVE AS S E S S M E NT I N D I C AT O R S BUILDING BLOCK 4: ● Disaster risk financing policy. Do any laws FINANCE or policies exist on disaster risk financing For ASP programs to respond to shocks in that set out how the government will pro- a timely manner, risk financing strategies actively manage the financing of disaster will need to be developed that can be uti- response? lized to mobilize resources rapidly through ● Quantification of post-shock ASP costs. safety net programs when needed. Build- Has the government undertaken any ing on international experience, optimal modeling or calculation of the economic risk financing strategies consist of multi- and fiscal impacts of shocks and disas- ple financial instruments established before ters? Did any such work consider the a shock occurs (ex ante) and combined in long-term/average costs of financing a risk layering approach. In the event of a any ASP response? shock, the appropriate financial instrument ● Financing instruments for ASP. Does is drawn on to finance an ASP response, the government have ex ante financ- depending on the severity (and cost) of the ing instruments in place for disaster shock. More generally, ASP will benefit from response, such as emergency funds, increased budgetary allocations that sup- contingency budgets/loans, or risk port the extension of social protection to transfer instruments? Do these financing poor households and sustain investment in instruments have clear conditions or trig- their resilience, along with investments in gers under which they can be utilized? long-term system strengthening. Have they been or can they be used for ASP response? Is the ability to use them The following are the foundational and adap- explicitly mandated or codified? tive indicators under Building Block 4. F O U N D AT I O N A L SUBREGIONAL A S S E S S M E N T I N D I C AT O R OVERVIEW ● Government spending on social protec- tion. What percentage of gross domestic The chapters that comprise the bulk of product (GDP) is allocated to social pro- this report present the individual country tection spending? Is the percentage assessments—including findings and and increasing or decreasing? What share of recommendations—for Botswana, Eswatini, these resources goes toward programs Lesotho, and South Africa. The remainder of that target the poorest? How does this this introductory chapter summarizes these share compare with those in neighboring findings and reflects on recommendations 10 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA that are pertinent at a subregional level. This With a few notable exceptions, adaptive includes findings and recommendations that indicators pertaining to additional specific emerged for more than one of the countries investments required for ASP—clear insti- and are relevant elsewhere. Some of these tutional mandates for social protection in common, country-specific recommenda- disaster response, program contingency tions could be applied subregionally—which plans, pre-positioned risk financing linked might be especially beneficial to small coun- to social protection—tend to be at a nascent tries like Eswatini and Lesotho, for which stage of development. The score for each there would be large economies of scale in building block is, however, only a rough sum- doing so. While the findings and recommen- mation of its constituent indicators. A closer dations are broadly grouped around each of examination of the indicators offers deeper the four ASP building blocks, many issues insights into specific findings, along with are relevant to more than one. recommendations for their strengthening. Overall, the findings from the assessment BUILDING BLOCK 1: suggest ASP is at a formative stage of devel- opment in southern Africa. Table 1.1 presents INSTITUTIONAL a high-level overview of the traffic light ARRANGEMENTS AND scores across the four countries for each PA R T N E R S H I P S building block, the vast majority of which The assessment of institutional arrange- fall between nascent and emerging stages ments and partnerships for ASP in the of development. This overview clearly indi- subregion placed them, for the most part, at cates that, despite the considerable climatic an emerging stage of development. As pre- risk to which southern Africa is exposed, sented in table 1.2, this building block is the subregion is not well prepared to mobi- rated as emerging in Botswana, Lesotho, lize postdisaster support to households in and South Africa; development is largely a timely and effective manner. Many of the driven in Lesotho and South Africa by the building blocks that are scored as emerg- presence of relatively strong foundational ing exhibit strong foundations within either social protection and DRM policies and leg- the social protection system (in terms, for islation, and in Botswana by relatively strong example, of social protection policy frame- coordination for shock response. Eswatini works, program coverage, and adequacy has made less progress in the development of benefits) or the DRM sector (in terms of of this building block, with foundational and early warning systems and DRM legislation). TA B L E 1 . 1   Scorecard for the four building blocks across countries Building block Botswana Eswatini Lesotho South Africa Institutional arrangements and partnerships ● ● ● Program design and delivery ● Data and information systems ● Finance ● ● ● Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. 1 :   I ntroduction 11 TA B L E 1 . 2   Scorecard for institutional arrangements and partnerships across countries Indicator Botswana Eswatini Lesotho South Africa Composite country rating ● ● ● Foundational indicators Social protection policy and legislative frameworks ● ● ● Disaster risk management policies and legislative frameworks ● ● ● ● Social protection institutional capacity and coordination ● ● Disaster risk management institutional capacity and coordination ● ● ● ● Adaptive indicators Clear mandate for social protection response ● ● ● Multisectoral coordination for social protection responses ● ● ● ● Partnerships with nongovernmental actors ● ● ● — Note: ● = nascent; ● = emerging; ● = established; — = not available. Mixed colors imply the rating is between two stages of development. adaptive indicators all assessed as being in disaster response. This is the case in between the nascent and emerging stages Botswana, where social protection is not of development. explicitly mentioned in the constitution or any legislation, nor is there any specific law KEY FINDING 1 or legislation relating to DRM. While a rel- atively recent National Social Protection The role of social protection in respond- Framework (GOB 2018a) sets out clear pri- ing to shocks in the case study countries orities for the social protection sector as a is typically not clearly mandated in either whole, it omits any explicit reference to the social protection or DRM policy or leg- use of social protection in shock response. By islative frameworks, leading to limited contrast, South Africa enshrines both social coordination and little to no investment protection and DRM provisions in its consti- in preparedness measures within the tution and wider legislative frameworks. It social protection sectors. Even when a recently amended its Social Assistance Act clear mandate exists in social protection to establish a mandate for one of its grant and DRM frameworks—as in the case of programs, the Social Relief of Distress grant, Lesotho—it has not translated into ade- to respond to disasters. quate investments in preparedness. In Eswatini and Lesotho, foundational pol- The status of foundational social protec- icies have been developed more recently, tion and DRM policies and legislation varies with substantial support from partners. In across countries. Where policies do exist, Lesotho, the National Social Protection they are often outdated and make limited or Strategy 2014/15–2018/19 (GOL 2014) no mention of the role of social protection emphasizes the role of social protection 12 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA in shock response and resilience building; managed by separate disaster response ongoing revisions for 2021–25 are expected systems with weak to nonexistent links to to articulate these roles further. In 2020, government social protection programs and a comprehensive set of legislative frame- delivery systems. The South Africa Social works and policies for DRM was developed, Security Agency (SASSA), for example, has with explicit provisions for the development no formal engagement with the National of ASP. Eswatini has shown less progress Disaster Management Centre or local equiv- in this regard, although the recent launch alents in disaster response. For its part, the of the National Social Assistance Policy is National Disaster Management Centre has promising for the whole social protection developed significant guidance on under- sector—and for ASP, too. taking postdisaster assessments, including mapping human impact, but the response The Lesotho experience also highlights that focuses mainly on immediate in-kind assis- legislative frameworks and policies need to tance (such as food parcels or temporary specify roles and responsibilities regarding shelter) provided by local disaster man- social protection and be backed with suffi- agement committee actors and does not cient investment to enable actors to fulfill include any SASSA-led cash transfer grants. their mandates. Social protection is still not In short, no coordination or linkage with fully embedded in Lesotho’s DRM plans and SASSA is apparent relative to the postdisas- strategies, lacking specificity in terms of ter assessment process or committees. concrete roles and responsibilities over the DRM cycle. Translating the policies and com- In Eswatini and Lesotho, cross-sectoral mitments articulated in these documents coordination is also weak, mostly as a result into investments in preparedness and effec- of constraints on the capacity of the disaster tive shock responses has been challenging. management authorities that limit their ability The occasional promotion and facilitation to fulfill their coordination mandates. In Leso- of such policies by partners has resulted in tho, capacity constraints on the National documents that, while comprehensive and Disaster Management Agency include not in line with global evidence, are not entirely only inadequate resourcing but also limited effective because the underlying institu- authority and autonomy to lead the sector tional bottlenecks and constraints—such and the constituent ministries, depart- as limited administrative capacity, scarce ments, and agencies involved over the DRM resources, and political economy issues— cycle. In such a context, legislative frame- are not resolved. works that promote ASP become ineffective. Indeed, in both countries, these coordination The gaps in policy and the largely absent issues have led to the development of par- mandate for social protection contribute to allel coordination structures between social weak coordination mechanisms among the protection and DRM actors, which creates relevant agencies in most of the countries. inefficiencies and further deepens the coor- Social protection departments often have dination problems. very little interaction with those responsi- ble for DRM and humanitarian response. Botswana provides a good example within Although DRM agencies may lead on disas- the subregion of effective cross-sectoral ter responses that include both national and coordination around ASP, especially at nonstate actors, these efforts are typically the local level. In Botswana, most social 1 :   I ntroduction 13 protection programs are administered by mobilize social protection responses. Con- the Ministry of Local Government and Rural sequently, the onus is on external agencies Development, whose representatives also to work with government to plan clear exit serve as key members of national and dis- strategies that systematically hand over and trict disaster management committees. build the capacity of social protection and Indeed, at the village level, ministry social DRM agencies. workers who manage social protection pro- grams are the first responders to disasters, International humanitarian actors do not play often leading postdisaster damage and any significant role in shock response in needs assessments. South Africa, where responsibility for social protection and DRM has been internalized KEY FINDING 2 by the national government. Botswana also does not currently depend on humanitar- Humanitarian and development partners ian partners to respond to shocks, although continue to play an important role in fill- the country was a recipient of humanitar- ing government technical and financial ian assistance (primarily through the WFP) capacity gaps related to ASP in Eswatini until the mid-1990s. In 1994, the govern- and Lesotho. ment established the national Pula Fund to support the development of a national Partnerships with nongovernmental actors social assistance system. The fund set up have been and will continue to be critical a range of social protection programs, out- enablers for the further development of ASP lined in the Botswana case study, among in Eswatini and Lesotho. In both countries, them several food assistance programs pre- partners also play key roles in implementa- viously delivered by the WFP, including in tion of annual vulnerability assessments as response to shocks. In the many years since, well as by conducting postdisaster needs these programs have served as the basis assessments. Partners have been instru- for government-led shock-responsive social mental in promoting policy dialogue on protection in Botswana. Hence, reliance on ASP (mainly in Lesotho) and in the devel- nongovernmental partners in Botswana has opment of policies and operational capacity been low, although national civil society (in both countries, though more success- organizations and nongovernmental orga- fully in Lesotho). In Lesotho, the United nizations do work in partnership with the Nations Children’s Fund (UNICEF), the WFP, government in the delivery of postdisaster and donors are working to strengthen the assistance. Development assistance from government’s social registry, the National partners could be catalytic in helping finance Information System for Social Assistance; crucial investments highlighted in the case as well as build the capacity of the National study, including early warning systems and Disaster Management Agency to coordinate the implementation of a social registry (dis- and respond to annual vulnerability analysis cussed further below). and needs assessments. While this relation- ship is born of necessity, it can result in a free-rider problem where governments have limited incentive to invest in the national insti- tutions and coordination systems required to develop their own national capacity to 14 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA S U B R E G I O N A L R E C O M M E N D AT I O N S O N programs and delivery systems. The assess- INSTITUTIONAL ARRANGEMENTS AND ment found, however, that the significant PA R T N E R S H I P S social assistance response in South Africa to COVID-19 was the exception to the rule of 1. Establish policy environments that encourage the develop- very limited large-scale vertical or horizontal ment of ASP, providing a clear mandate for social protection expansions in response to climatic shocks. in shock response and specific roles and responsibilities for implementing agencies. KEY FINDING 1 2. In most contexts, significantly increase resource allocations to enable DRM agencies to fulfill their coordination man- Generally, the countries with broad dates, including as those mandates relate to the coordination social assistance coverage and stronger of social protection in response to shocks. Social protection foundational social protection delivery departments should be integrated fully into DRM coordination systems have been better able to mobi- mechanisms at all levels—national, regional, and local. lize responses to shocks, including to COVID-19. 3. Ensure the support provided in Eswatini and Lesotho by non- governmental partners is part of a strategic vision that aims to strengthen government capacity to build household resilience The countries assessed in this study that have and to respond to shocks with social protection over the longer mounted the more effective ASP responses term. Key nongovernmental partners should be engaged by to shocks are those whose delivery systems the respective country governments in the development of are more established and comprehensive. All shock-responsive social protection contingency plans (see the case study countries have implemented Building Block 2) to enhance coordination across the govern- vertical and/or horizontal expansions of ment–humanitarian divide when responses are initiated. their social protection programs, particu- larly in response to the COVID-19 crisis and recent droughts. Those that have mounted the most effective ASP responses tend to BUILDING BLOCK 2: have more comprehensive and relatively PROGRAM DESIGN AND well-established core social protection sys- tems, with relatively high coverage. South D E L IVE RY Africa is a clear illustration; its social protec- The assessment of ASP programs and deliv- tion system is one of the most comprehensive ery systems found them to be at an emerging on the continent, with almost all foundational stage of development in the subregion. As dimensions under Building Block 2 scoring shown in table 1.3, Botswana, Eswatini, and as established in this assessment. In the face Lesotho are in transition from nascent to of the pandemic, the government of South emerging with respect to this building block, Africa immediately expanded payments to with these ratings largely driven by limited over 17 million social protection beneficia- investments in both adaptive and founda- ries, and a new COVID-19 Social Relief of tional social protection indicators (although Distress grant was established and opera- Botswana stands out as established for tional in just over one month. This new grant several foundational indicators—for exam- represented a horizontal expansion of cov- ple, in the coverage of its social assistance erage to working-age adults (aged 18–59) programs). South Africa is classified as affected by the pandemic. By November between emerging and established, sup- 2020, the government had registered and ported by strong foundations in terms of its million people under the new paid over 6  1 :   I ntroduction 15 TA B L E 1 . 3   Scorecard for programs and delivery systems across countries Indicator Botswana Eswatini Lesotho South Africa Composite country rating ● Foundational indicators Benefit adequacy ● ● ● ● Eligibility criteria and targeting ● ● Coverage of the main social assistance program ● ● ● Financial inclusion and national ID ● ● ● Social protection payment systems Social protection delivery systems ● ● ● Adaptive indicators Experience of vertical expansion ● ● ● Experience of horizontal expansion ● ● ● Protocols and contingency plans ● ● ● ● Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. grant—an achievement reflecting SASSA’s post—that is, when a disaster has already strong institutional capacity; its highly digi- occurred—resulting in an ad hoc approach tized application, registration, and payment that often translates into assistance being systems; and the wide national coverage of late and of limited adequacy in terms of who its programs, staff, and offices. is reached and the amount of the assistance provided. Attempts to consider the overlap KEY FINDING 2 between the beneficiaries of existing social protection programs and those populations Despite their all being subject to regular more regularly or severely affected by each climatic disasters, no case study coun- country’s most frequent climatic shocks tries have developed any protocols or have also been limited. guidelines for using social protection programs to respond to shocks. No country’s social protection program has guidelines or operating procedures for when This finding reflects a general lack of pre- and how to scale up even in response to rela- paredness across the subregion, leading to tively regular shocks. This deficiency reflects largely ad hoc mobilizations of social protec- limited planning in the development of the tion responses. As countries have expanded social protection systems and their potential and developed their social protection sys- role in offering direct and speedy assistance tems, few have considered the role of social in the wake of acute shocks and building protection in responding to shocks as part resilience to shocks in the long term—a defi- of the design or delivery of their programs. ciency that persists despite the very regular As highlighted in the case studies, most ASP and increasingly frequent climatic hazards interventions are designed and mobilized ex affecting all the countries. This situation 16 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA may be changing, though, as a result of the countries of relatively minor adjustments COVID-19 crisis. This major covariate and to targeting and benefit package design to unexpected shock has brought far more account for climate risk in regular program- attention to the role of social protection as ming. Geographical targeting, for example, a response mechanism and has provided did not prioritize the subnational areas that a significant stress test of existing systems are most exposed and vulnerable to cli- and the extent to which they are prepared mate risk as a basis for providing long-term to scale up and meet widespread household support through regular social protection needs. How much learning from this expe- programs. Similarly, the review did not find rience is being documented and used to evidence that the amounts, timing, and dura- develop more standardized guidelines or pro- tion of benefit packages were being tailored tocols for ASP responses remains unclear. to climate shocks and stresses such as the lean season. Only limited evidence indicated KEY FINDING 3 that public works programs were identifying physical works that would more effectively The design of key elements of regular reduce the impact of disasters—for instance, social protection programs (targeting, the clearing of drainage systems in urban benefit timing, duration, and amount) areas and/or the creation of water catch- could be improved to build household ments in rural areas. Lastly, no significant resilience to climate shocks. experience was found of efforts to connect beneficiaries located in high-risk areas with Given the chronic nature of climate shocks complementary resilience-building or cli- in southern Africa, the design of some ele- mate change adaptation activities, such as ments of core social protection programs livelihood grants and training for off-farm could be improved to build household resil- income-generating activities and/or the cre- ience to shocks in the longer term. The ation of savings groups. assessment found no evidence in the four S U B R E G I O N A L R E C O M M E N D AT I O N S O N P R O G R A M D E S I G N A N D D E L I V E R Y 1. The most important step in developing each country’s capacity to respond to shocks is continued strengthening of the foundational elements of national social protection programs, including their coverage and their delivery systems, which remain relatively nascent in Botswana, Lesotho, and—especially—Eswatini. South Africa provides a good exam- ple in the subregion of how strong social protection foundations result in a generally improved preparedness posture. 2. Greater ex ante analysis of disaster risk should be applied in the design and delivery of regular, day-to-day social protection programs to ensure targeting is risk informed (prioritizing poor and vulnerable households in high-risk areas) and that benefit packages maximize resilience outcomes—for example, by timing delivery to coincide with the lean season and offering complementary services to support more resilient livelihoods. 3. The use of social protection in response to COVID-19 has provided an opportunity for systematic review of the potential of existing social protection programs to respond to the most frequent and regular shocks affecting each country. Reviews of COVID-19 responses—each in and of itself a severe stress test of the capacity for social protection to respond to shocks—should inform the creation of contingency plans and protocols. These forward-looking plans and protocols should capture the lessons learned, informing the program design parame- ters and delivery processes needed for future responses. 1 :   I ntroduction 17 BUILDING BLOCK 3: systems for sharing and responding to disas- ter risk information. D ATA A N D I N F O R M AT I O N S Y S T E M S KEY FINDING 1 The assessment of social protection and Varying degrees of digitization and lim- DRM data and information systems placed ited interoperability, integration, and most countries between the nascent and coverage of social protection informa- emerging categories. As shown in table 1.4, tion systems constrain the ability of South Africa has the most digitally advanced the case study countries to respond to and comprehensive beneficiary registration shocks, with the notable exception of systems. In Eswatini and Lesotho, many indi- Lesotho. vidual programs still lack digital beneficiary registries and MISs (an exception is Leso- tho’s Child Grants Program); beneficiary Most of the countries have social protection data are thus not well integrated. That said, information systems that can be classified as Lesotho stands out as the only country in either nascent or emerging. (See the typol- the subregion with a social registry, although ogy in figure 1.3.) Beneficiary registries plans for developing social registries are enable the government to reach existing under way in Botswana and South Africa. social protection beneficiaries quickly with Lesotho’s registry covers virtually the entire additional cash transfers during an emer- population and has been used to inform gency. This process is hastened significantly recent responses to shocks. In the collection in countries where beneficiary registries are and utilization of disaster risk or early warn- digitized. Countries do not register all ben- ing information to inform social protection eficiaries digitally, however, and they use responses, all the case study countries were multiple MISs whose sophistication ranges assessed as nascent or emerging. Botswana from paper-based lists to national databases. had the most developed cross-sectoral Eswatini’s main social assistance programs, TA B L E 1 . 4   Scorecard for data and information systems across countries Indicator Botswana Eswatini Lesotho South Africa Composite country rating ● Foundational indicators Data and information on disaster risk and household vulnerability ● ● ● Beneficiary registry ● ● ● ● Social registry ● Adaptive indicators Use of preexisting data for shock response ● ● ● ● Post-shock household needs assessments ● ● ● ● Data-sharing protocols with internal/external partners ● Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. 18 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA FIGURE 1.3  Development trajectory of social protection information systems Source: Based on World Bank 2020b. Note: The development of information systems is not as linear as the figure suggests. Countries can advance development on one spectrum yet demonstrate less maturity on another; they can also leapfrog to more integrated and interoperable systems. Having social registries in place is not a precondition for an “established” social protection information system. Countries may decide to strengthen other registries and their interoperability rather than develop a social registry. The best approach will depend on the context and policy priorities; since social registries, for example, are typically associated with poverty-targeted programs, countries with universal or categorical programs may prefer or need other types of registries. for example, lack digital beneficiary regis- The interoperability of social protection infor- tries with associated MISs. In Lesotho, the mation systems within the subregion (their Child Grants Program is the exception to the ability to exchange and make use of data) same trend, with all other social assistance and their levels of integration (consolidation programs in the country lacking digital bene- of processes like data collection and analy- ficiary registries. South Africa has developed sis across programs and institutions) are very a comprehensive digitized beneficiary reg- limited. As a result, the ability to aggregate istration and payment system—the Social or cross-reference beneficiaries in different Security Pension System (SOCPEN)—which programs is also limited, which reduces their is used to process applications for all social utility in informing shock responses, even for protection grants and generate payments for relatively straightforward vertical expansions. approximately 17 million grants each month. South Africa’s SOCPEN is an exception, as its systems provide interoperability among 1 :   I ntroduction 19 several administrative databases. Integrated registries hold the valuable potential of beneficiary registries like SOCPEN, which enabling the government to expand assis- the other countries do not have in place, tance to households that are not already can be used to inform vertical expansions. social protection beneficiaries (table 1.5). Despite the merits of such registries, they Relying on social registry information has tend to include only beneficiaries (and some- a core limitation, however, that can lead to times former beneficiaries and people on missed or mistaken identifications: the data the waiting list), which limits the possibility collected before the shock cannot capture of reaching nonbeneficiaries (for horizontal the changes caused by the event. While expansions) using existing data. social registries can play a key role in inform- ing timely shock responses, the feasibility of Social registries can be a key source of using such data given the potential for inac- information for rapid responses to shocks, curacy is very context-specific and depends including horizontal expansions. When on social registry coverage, the relevance of responding to a shock, policy makers face the data the registry contains, the accuracy a trade-off between timeliness and accu- of the data, their currency, and their accessi- racy. Taking advantage of existing registry bility (Barca and Beazley 2019). Of the four data (as opposed to collecting new data countries studied, only Lesotho has a social through postdisaster needs assessments, registry in place, and although the country’s for instance) may be the fastest way to iden- experience in using it to inform responses to tify individuals, depending on the scale of recent shocks has been noteworthy, it has the shock and the capacity to collect data also presented challenges, as outlined in ex post. Unlike beneficiary registries, social box 1.1. TA B L E 1 . 5   Utility of and trade-offs between beneficiary and social registries for shock response Response type Beneficiary registry Social registry ■ Contains information on existing ■ Contains information on individuals who may or may not be ben- social protection beneficiaries eficiaries of existing social protection programs Vertical expansion used to inform vertical expan- (giving additional ■ While it can also be used to inform vertical expansions to sions during times of crisis transfers to existing existing beneficiaries, the social registry needs to retrieve infor- beneficiaries) ■ Integrated beneficiary registries mation from beneficiary registries to match those individuals to allow vertical expansions of vari- the beneficiaries in its database ous programs Cannot be used for horizontal ■ Suitable for informing horizontal expansions, as it contains expansions to new beneficiaries, information on the wider population (although the share of the since it only contains data on exist- population in a social registry can vary significantly) Horizontal expansion ing social protection beneficiaries ■ Since it is designed to inform the assessment of needs and (increasing coverage conditions of households for entry into any number of social to reach nonbenefi- protection and related programs, it contains socioeconomic ciaries affected by data useful for prioritizing those who should receive support via the shock) horizontal expansions after a shock ■ Does not, however, typically include operationally relevant data, such as bank account details, to facilitate post-shock transfers 20 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA B OX 1 .1   Lesotho’s social registry and shock response Lesotho’s National Information System for Social Assis- Despite the impressive achievement of developing a tance (NISSA), unique among the countries studied, social registry that covers virtually all households, the provides an essential foundation for shock response. system still faces several challenges and limitations. NISSA is a social registry containing socioeconomic First, NISSA lacks a mechanism for updating data fre- data for virtually all households in the country. These quently and at scale, which is a constraint not only data have been gathered from surveys going as far for regular social protection programming but also back as 2009, with the latest survey rounds for rural for responding to shocks. The “census sweep” strat- areas conducted from late 2016 to mid-2019 (GOL egy for data collection (in which data are collected on 2020d). The registry currently covers all 331,248 rural every household in the country, as is done in the roll- households in Lesotho. An additional 78,000 urban out of a national census survey) is expensive, and the households (representing 36 percent) were enrolled by ministry and key partners have indicated that a more the end of 2020, with enrollment anticipated to reach cost-effective method should be developed. Serious 100 percent by the end of 2021. Currently, NISSA data challenges to the registry’s sustainability are posed by are used to determine eligibility for the Child Grants limitations on capacity at the local level, the scale of Program and the Orphans and Vulnerable Children the registry, and the amount of data that needs to be Bursary program. The Ministry of Social Development collected. plans to utilize the registry for other social support pro- grams but has not yet done so. A further challenge lies in NISSA’s very limited interop- erability and integration, which prevents its ASP NISSA has played a central role in recent shock potential from being fully exploited. In terms of interop- responses. It informed vertical and horizontal expan- erability, NISSA’s data exchanges are rare and mostly sions of the Child Grants Program in response to ad hoc, rather than on the basis of protocols and recent droughts and the COVID-19 pandemic. Key agreements. Regarding integration, the only process actors agree that NISSA should be the main source of incorporated so far has been the targeting of the Child household-level information for planning and execut- Grants Program and the Orphans and Vulnerable Chil- ing response actions. dren Bursary program. NISSA does not yet use the national ID as a unique identifier, which is an obstacle to further interoperability and integration. KEY FINDING 2 information to inform and trigger social pro- tection responses. The information base The use of data on risk and vulnerabil- associated with disaster risk, vulnerability, ity from disaster early warning systems and early warning is very mixed across the to inform social protection responses is four countries: very limited. ● Botswana has the weakest early warn- Despite the existence of ever-more sophis- ing system of the countries assessed, ticated satellite-based tools to provide early with limited agrometeorological mon- warning data on climatic disasters, none itoring systems and almost no recent of the case study countries has used such disaster risk profiles for most of its 1 :   I ntroduction 21 population. Consequently, social pro- and the Department of Health. Together tection responses are triggered by they provide accurate climate informa- post-shock assessments undertaken by tion, such as historical trend data and national or local DRM committee teams seasonal predictions, projections, and in the field. Although these comprise early warnings of extreme weather and robust multisectoral assessments of other climate-related events to inform household needs, they do not constitute adaptation planning and disaster risk early warning. reduction. What is not being monitored, ● Eswatini’s and Lesotho’s early warning however, is the disaster risk and expo- systems are based on annual assess- sure levels of individual households, ments that combine agrometeorological particularly poor ones. Fundamen- data with baseline data on household tally, while offering granular information vulnerability. These annual assessments on weather and disasters, these sys- produce estimates of food-insecure pop- tems reveal very little about the impact ulations throughout the two countries, of monitored hydro- or agrometeoro- drawing on the WFP’s vulnerability anal- logical conditions on households. The ysis and mapping approach (WFP 2018). indications they provide of these condi- Although very useful and credible, tions need to be complemented with very these assessments are time-consuming different information to assess and quan- and expensive to carry out and are still tify their effects at the household level. often conducted ex post (for example, This is a common problem that reflects after rains or harvests), providing only the incongruity between the availability static annual assessment. Although con- of very advanced technology to monitor ducted by the governments themselves, meteorological and agricultural indica- they are heavily supported (in terms of tors and the much greater difficulty of both resources and capacity) by exter- monitoring household poverty and vul- nal agencies, primarily the WFP. It is not nerability, which requires much more clear this approach would be sustain- time and many more resources to collect able if external support were withdrawn. appropriate data. Additionally, because these assessments always identify household needs based KEY FINDING 3 on conditions that have already materi- Few countries have systematic protocols alized, they tend to be used to appeal for or guidance for conducting postdisaster assistance rather than inform or trigger household needs assessments, and the any predetermined ASP response. linkages between such assessments and ● South Africa has several sophisticated social protection responses are weak. government-led early warning systems and multiple organizations involved in All countries undertake some form of post- forecasting and monitoring climatic disaster assessment of household needs events. These include the National following both fast-onset shocks like floods Disaster Management Centre, the South and slower-onset shocks like drought, but African Weather Service, the South the linkages between the findings of these African National Space Agency, the assessments and social protection are often Department of Environmental Affairs, weak. South Africa has a network of disaster 22 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA management agencies responsible for post- impact and response can thus be ad hoc and disaster assessments, with a particular focused mainly on the provision of imme- focus on floods, fires, and other short-onset diate in-kind assistance (such as food or shocks; protocols and guidelines do exist for temporary shelter). Even the provision of these activities. Elsewhere in the subregion food assistance following assessment is not greater emphasis is placed on assessments often viewed as social protection, though. in response to drought, which tend to follow This further reflects the lack of any sys- harvests rather than disasters. Where these tematic consideration or inclusion of social assessments happen annually, such as those protection agencies and programs as part of conducted by the Lesotho Vulnerability disaster response. Assessment Committee, they tend to follow standard approaches based on global proto- There are no examples of the retention of cols used by agencies such as the WFP. In household needs assessments in central- Eswatini, postdisaster assessments of food ized databases to inform planning for future security (which are very similar to posthar- responses. In Eswatini, for instance, the vest assessments) are led by the National information from the postdisaster needs Disaster Management Agency (NDMA) assessments conducted by different actors annually and carried out by nongovern- is neither consolidated in any single data- mental organizations and agencies based base nor by the NDMA, so each partner on selection criteria and procedures jointly relies exclusively on the information it has developed by the NDMA and a consortium of collected. In the responses to COVID-19, the partners. In Botswana, similar assessments NDMA and some partners used the WFP’s only take place in years of bad drought. In all beneficiary information and transfer manage- cases, guidance linking the findings on the ment platform (SCOPE) to register affected impact of the shock to any specific social households. In sum, recognition of the role protection responses is limited due to lack of social protection information systems can financing (see Building Block 4) and to weak play in generating beneficiary lists, identify- connections among disaster management ing poor households, and understanding the agencies, their assessments, and social pro- demographics of a shock-affected popula- tection ministries. The mapping of human tion is limited. 1 :   I ntroduction 23 S U B R E G I O N A L R E C O M M E N D AT I O N S O N D ATA A N D I N F O R M AT I O N S Y S T E M S 1. Countries should continue to prioritize investments in the development of mature social protection information systems. Maturity can be achieved in different ways, with different types of registries and integration strate- gies based on country needs, and is a long-term endeavor. Having both widespread coverage of national IDs and digital registries in place is key for interoperability and integration. The development of protocols and mecha- nisms for data sharing and data protection is also crucial, as are mechanisms to ensure data are up to date and reliable. In the development of social registries, each country should follow a tailor-made, context-specific path- way. Although remarkable, Lesotho’s approach to developing a social registry—through household-level census sweeps covering the whole country—is not the only way. Alternative or complementary pathways may entail developing on-demand mechanisms at the local level (for example, allowing those who are affected or in need to enroll on an ongoing basis) and making use of other administrative registries containing information on nonben- eficiaries (further highlighting the importance of interoperability). 2. Social registries should be risk informed. With regard to ASP, social registries are risk informed when they col- lect household information useful for conducting risk and vulnerability analysis and informing shock responses and when they exchange data with DRM and humanitarian actors. When increasing the coverage of existing or planned social registries, the preregistration of households living in high-risk areas should be prioritized as a means of enabling horizontal expansions in response to predictable climatic shocks. 3. Countries should develop clear protocols and guidance for postdisaster assessments. Such guidelines should take into account how existing social protection programs, including their MISs, can serve as a basis for calcu- lating needs and delivering responses. Similarly, social protection registries should be expanded through the registration and enrollment of beneficiaries for emergency support in the wake of a shock. Consideration should be given to harmonizing data collection tools across departments to enhance interoperability and the develop- ment of social registries. 4. A strong case can be made for a more systematic approach to early warning systems for ASP across the sub- region. Because of their geographical proximity to South Africa, small nations like Eswatini and Lesotho could benefit from South Africa’s comprehensive remotely sensed early warning information. Indeed, the homoge- neity of climatic shocks makes a subregional approach to developing early warning systems that inform ASP cost-effective. Subregional actors such as the Southern African Development Community (SADC) and the WFP should consider how to help countries across southern Africa develop standardized systems to track the disaster risk and vulnerability of their populations to make it easier to predict which populations need the most assis- tance in the face of different shocks. Cross-regional support could also be helpful to governments in developing national indicators and thresholds for the most regular climatic shocks. These indicators and thresholds could be used to trigger standardized but country-specific social protection for predetermined households (see the above recommendation on social registries) in response to the more recurrent and predictable climatic shocks. Similar support should be provided for the development of robust but simple methodologies to monitor the disaster risk of different populations across the subregion, the results of which could inform the provision of long-term social protection measures that build the resilience and support the adaptation of those households. 24 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA BUILDING BLOCK 4: Three of the case study countries allocate FINANCE between 2.5 and 5.0 percent of their national budgets to social protection, which compares All case study countries were assessed as favorably against international standards. nascent with regard to financing for ASP. Eswatini is the exception, with its spend- Although all except Eswatini scored well ing the lowest of any country in the SACU in terms of national expenditures on core subregion. The higher spending levels show social protection, resources for financ- that, for the most part, political support for ing ASP were absent in almost all cases, social protection programming is relatively as table 1.6 shows. While no country has strong. The role of social protection in build- developed disaster risk financing policies, ing the resilience of the poorest to shocks the governments of Eswatini, Lesotho, and is less acknowledged, however, which may South Africa have completed disaster risk mean large amounts of social protection finance diagnostics and are in the process financing are directed to projects with limited of developing such policies, with a focus on impact in reducing poverty or vulnerability expanding the suite of risk financing instru- to shocks. Botswana’s tertiary grants and ments available to finance disaster response, sponsorships program, for instance, is the including through ASP programs. Some such country’s most expensive social protection instruments—for example, the contingency program, despite having a very low caseload reserve in South Africa—are in place to fund (less than 29,000). The program is also not unexpected expenditures, including disas- means tested, being available to any youth ters; none, however, has the explicit purpose achieving the requisite score in the second- of financing an ASP response. ary school leaving exam. By not specifically targeting the poorest children, the program KEY FINDING 1 tends to benefit wealthier households. The reallocation of such budgets to provide Governments are mostly maintaining and more generous payments to the poorest even expanding social protection budgets, and/or for complementary livelihood activ- but the effectiveness of social protection ities (highlighted under Building Block 2) expenditure in reducing poverty varies. TA B L E 1 . 6   Scorecard for finance across countries Indicator Botswana Eswatini Lesotho South Africa Composite country rating ● ● ● Foundational indicators Social protection spending ● ● ● Disaster risk financing policies and strategies ● Adaptive indicators Quantification of the expected costs for social protection responses ● ● ● ● Disaster risk financing instruments linked to social protection ● ● ● ● Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. 1 :   I ntroduction 25 would do more to build the resilience of poor Risk Management Fund in Eswatini (not yet households to climatic shocks. Ultimately, operational), the Contingency Fund in the this should reduce the numbers of those Ministry of Finance of Lesotho, and the con- requiring ASP in the event of shocks—partic- tingency reserve in South Africa. Although in ularly regular climatic shocks—and, in turn, theory the instruments should be available the implied cost of response. Robust impact to finance disaster response, they are fre- monitoring of social protection programs is quently used to support other emergencies essential to track these implicit benefits. and are depleted early in the budget cycle before shocks occur. This was the case in KEY FINDING 2 both Lesotho and South Africa for the 2016– 19 droughts. While all countries lack disaster risk financing policies and ex ante financing The diagnostic studies showed the financ- instruments for ASP, they are working to ing of shock responses relies largely on strengthen their financial resilience to ex  post budget reallocations, adjustment shocks. budgets, ex post borrowing, and humanitar- ian aid. In Botswana, a United Nations Office Governments in southern Africa are prior- for Disaster Risk Reduction review of DRM itizing the strengthening of disaster risk expenditure found that, in the five years financing, albeit from a low starting point. leading up to 2019, the government made no Having recognized their acute exposure allocations for disaster risk response, relief, to disaster shocks and the lack of ex ante or reconstruction and recovery (UNDRR financing instruments available to mobi- 2020). In response to COVID-19, Eswatini, lize a response, Eswatini, Lesotho, and Lesotho, and South Africa drew upon the South Africa have now all completed disas- ex post borrowing facilities available from ter risk finance diagnostic studies, which the International Monetary Fund’s Rapid include a range of recommendations to Credit Facility and Rapid Financing Instru- strengthen financial resilience to disasters in ment. In 2020, South Africa drew its full the subregion. Central among these are the billion from the allocated amount of $4.3  development of national risk layering strate- Rapid Financing Instrument, which may have gies to finance disaster responses, including helped enable its generous and expansive through ASP programs. Although the eco- social protection response to the pandemic. nomic and wider costs of disasters were million Eswatini gained access to $110  modeled as part of these diagnostic reports, through the Rapid Financing Instrument and the analysis was not at the sectoral level Lesotho to $49.1 million through the Rapid and thus did not quantify the cost of an ASP Credit Facility and the Rapid Financing response. While no country has any disaster Instrument, both in July 2020 (IMF 2020a, risk financing policies or strategies in place, 2020b, 2020c). the recently completed disaster risk financ- ing diagnostics will provide a useful basis for KEY FINDING 3 their development (World Bank 2019b). The quantification of the costs and Some ex ante risk financing instruments impacts of regular and frequent disas- are in place in the subregion, but they tend ters is almost exclusively undertaken to be unreliable. These include the Disaster post shock. 26 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA In all the case study countries, social protec- supplies. The gaps in income and consump- tion response costs are quantified ex post, tion of poor households that result from the and no estimation is actively undertaken most regular shocks are not analyzed. ex ante as a basis for securing financing in advance of a shock. Ex post assessment is No examples were identified of ex ante led by the relevant disaster management aggregated analysis or modeling of the indi- agency in South Africa, and by multiagency vidual- or household-level impacts or costs teams of government and nongovernmen- of disasters to affected populations. Such tal actors (such as the Lesotho Vulnerability exercises are important preliminaries in the Assessment Committee) elsewhere. Despite development of disaster risk financing strat- the chronic nature of most climate-related egies or actions to gain, for example, an disasters in the subregion, few countries understanding of the annual budget required have methodologies in place to estimate for a drought contingency fund. Such anal- their costs before they occur. In South Africa, yses could also inform the nature of social efforts to estimate the costs of disaster protection responses that would most effec- have concentrated on damage to property tively build the disaster resilience of poor and infrastructure. In the case of drought, households. Modeling should include a range the focus of financial assessments has of the most likely disasters and consider been on commercial impacts to agricultural how the costs associated with response and production—in, for instance, the wine indus- impact vary relative to the magnitude and try—or on reductions in hydroelectric power frequency of each. 1 :   I ntroduction 27 S U B R E G I O N A L R E C O M M E N D AT I O N S O N F I N A N C E 1. Social protection expenditures should be prioritized toward those programs that most effectively reduce poverty and, by extension, build resilience among the households that are likely among the most vulnerable to climatic shocks. Most countries have the scope to reorient spending away from regressive social programs, such as tertiary bur- sary schemes, toward programs that prioritize poor households. 2. Implement the recommendations of the disaster risk finance diagnostic studies carried out in Eswatini, Lesotho, and South Africa (World Bank 2019b, 2021e, 2022b). Build- ing on the momentum created by the completion of the diagnostics, the respective governments are well placed to move forward with the drafting of national disaster risk financing policies, the adoption of which is important to clarify their strategic objec- tives related to disaster response. Financing of ASP should have a central role among these objectives. 3. Expand the suite of ex ante risk financing instruments available for disaster response, including for ASP. The diagnostic studies in all three countries recommend doing so through a risk layering approach, with the inclusion of more robust contingency funds, contingent credit (for example, through the World Bank Development Policy Opera- tion with Catastrophe Deferred Drawdown), and sovereign risk insurance. Fiscal gap analyses conducted in the diagnostic studies that sought to quantify the cost savings if Eswatini, Lesotho, and South Africa expanded their risk financing instruments found that the governments in all three would save money on average if they implemented the modeled risk layering approach, with significantly higher savings ($26 million in Eswatini, $33 million in Lesotho, and $500 million in South Africa) for extreme disas- ter events. 4. In parallel to the above, all the case study countries should examine the long-term costs of adapting social protection in response to disasters. This will require ex ante analysis and modeling of historical disaster data alongside data on the actual or esti- mated impact of disasters on household consumption. Such analyses would provide a data-driven understanding of the likely financial costs each country can expect in developing various approaches to and scales of ASP programming. Significant scope exists for such analyses to be undertaken at the subregional level, given the similarity in type and frequency of the climatic shocks experienced across the subregion and the often transboundary nature of their impacts. 5. Alongside investments in linking disaster risk financing for shock response, there is significant, untapped potential for the case study countries to access climate finance in support of adaptation. As outlined by Longhurst et al. (2021), climate finance could be channeled to invest in ASP delivery systems, while social protection programs could offer the southern African countries new ways to deliver support for household adapta- tion, mitigation, and risk management. B OT S WA N A A D A P T I V E S O C I A L P R O T E C T I O N A S S E S S M E N T C A S E S T U DY 2 estimated to have contracted by 8.5 percent OVERVIEW in 2020 (World Bank 2021c). While the econ- OF RISK AND omy was expected to rebound to 8.5 percent HOUSEHOLD in 2021, GDP was expected to remain below its 2019 level until 2022 (World Bank 2021c). VULNERABILITY The government of Botswana has conse- quently been keenly aware of the need to Botswana has long been considered a top diversify the economy to maintain growth performer on the African continent in terms and address its persistently high levels of of economic development. The country suc- unemployment and inequality. Additionally, cessfully maintained an average growth rate Botswana’s emergent tourism industry— of over 5  percent from the time of its indepen- which is based on the country’s wildlife and dence in 1966 until the 2009 global financial national parks—is being affected by travel crisis, as its gross domestic product (GDP) restrictions related to the pandemic; tour- per capita grew almost five times faster than ism is also threatened by climate change. the global average during the same period. The economic impact of COVID-19 is likely This success was catalyzed by the discov- to be deep and long-lasting, and, in common ery of the world’s largest diamond deposits, with most shocks and disasters, it will dis- which have made Botswana a leading pro- proportionately affect the poor. The World ducer and exporter of the gems, accounting Bank estimates that poverty in Botswana for over 80 percent of both overall exports under the upper-middle-income poverty line and government revenues. High-end tour- ($5.50 per day) increased in 2020 by 3 per- ism has also grown in recent years, although centage points (World Bank 2021c). agriculture has remained largely dominated by subsistence farming, with some commer- cial livestock production. POVERTY Despite an impressive record of growth, Botswana’s economic growth has stalled Botswana struggles with persistent pov- somewhat since the financial crisis, as erty and inequality. In 2015/16, 16.1 percent demand for diamonds fell and has remained of the population (approximately 390,000 depressed. The COVID-19 crisis has exac- Botswana) were poor. This total masks erbated these trends, with the economy 29 30 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA significant geographical variations; while the SHOCKS AND poverty rate in cities and towns was 3.3 per- DISASTERS percent cent, in rural areas it was 26.8  (World Bank 2022a). Unemployment, dis- Botswana is exposed to numerous hazards, ability, and disease prevalence are all greater including droughts, floods, earthquakes, in rural areas. The overall unemployment strong winds, land fires, and pest infes- rate in Botswana was around 23.2 percent tations. Table 2.1 shows that floods occur before the pandemic, rising to 24.5 percent most frequently, while epidemics have in 2020 as a result of it (World Bank 2021c). caused the most deaths, largely attributed Unemployment is higher among youth and (pre-COVID-19) to an episode of acute diar- women, with 64.8  percent of the unem- rheal syndrome in 2006 (22,264 affected, ployed in 2020 under the age of 35. 1 470 deaths). Droughts are the most wide- spread natural hazard in terms of the Botswana experiences stubbornly high rates population and areas affected. of malnutrition, despite its middle-income status. Overall, 28 percent of children are F LO O D S AN D STO R M S percent in the rural stunted, rising to 32  northeast. Wasting rates are estimated at Botswana is drained by a complex system 7.2 percent, much higher than the average of transboundary rivers (map 2.1), most for poorer neighbors in the subregion, such of which dry up in the summer but fre- as Eswatini and Lesotho. Although the World quently flood during the monsoon season, Bank estimates that inequality, as measured which is also when most localized storms by the Gini index, fell from 60.5  percent occur. Flooding affects an estimated aver- to 53.3 percent between 2010 and 2015, 2 age of 8,000 people per year (CIMA 2018). it remains unacceptably high. Improving Affected populations are concentrated in income growth and economic opportunities the central district, which is the most popu- for the rural population will be crucial in cre- lous, and in Ngamiland, where the Okovango ating a more equal society. delta is located. Although floods cause rel- atively few deaths, their economic impact is significant. In 2001, they caused damage 1 Source: World Bank “Where We Work” Botswana to private and public property estimated web page. at $711,885 (UNDRR 2020). According 2 Source: World Bank “Where We Work” Botswana web page. to the Botswana Central Statistics Office, storms and floods affect hundreds of house- Incidence and impact of disaster events in TA B L E 2 . 1   holds and thousands of people countrywide Botswana, 1965–2018 annually. Localized storms are becoming a regular occurrence, particularly in areas Hazard Occurrences Deaths People affected around Mahalapye, Palapye, Selibe Phikwe, Drought 6 — 1,344,900 and Serowe, where they are deleterious to Epidemica 3 655 36,897 people’s livelihoods. The United Nations Flood 11 43 144,577 Office for Disaster Risk Reduction has esti- Insect infestation 1 — — mated that an average of $90 million of GDP Storm 1 — 400 (approximately 0.56 percent) is affected by Source: World Bank 2021a. floods each year (CIMA 2018). a. Does not include COVID-19 figures. — = not available in database. 2 :   B otswana 31 M A P 2 .1   Botswana and remains heavily dependent on sub- 20 ∞E 25 ∞E IBRD 33376 sistence crop and livestock farming. Reductions in output due to drought directly reduce farm incomes and To ZAMBIA Lusaka BOTSW ANA ANGOLA To Kasane affect dietary diversity and overall food Menongue i nt consumption. ay Ok Lin N ga ma s CHOBE av eri an go Xa ud um NAMIBIA Nokeneng NGAMILAND Okavango Swamps ZIMBABWE ● Reduced livestock production: The To livestock industry is an important agri- Maun Gweru 20 ∞S Nata 20 ∞S Tsau eb E is Sehithwa Lake cultural subsector for Botswana. Drought Ngami Makgadikgadi Salt Pans NOR TH- Rakops Bote ti EAST To Orapa Francistown Messina affects pasture and water sources, with Lake Xau Letlhakane S ha she Mo Ghanzi C E N T R A L tlo utse Seruli Selebi-Phikwe GHANZI Mamuno O kwa Serowe Palapye Sefophe Limpo p consequences for livestock conditions o Lotsa ne K a l a h a r i Mahalapye and production levels. D e s e r t Kang ● Human water supply: Groundwater, KWENENG Tshane KGA TLENG Mochudi SOUTH AFRICA which is the main source of water supply Molepolole KGALAGADI Jwaneng GABORONE Khakhea SOUTHERN 25 ∞S Kanye SOUTH- to the people of Botswana, is threatened 25 ∞S No W erda Lobatse EAST sso b be M osele To Pretoria po M olo Tshabong po BOTSW ANA by low levels of replenishment and high lo Mo salinity. Overabstraction is becoming a To DISTRICT CAPIT ALS* Kimberley Bokspits NA TIONAL CAPIT AL RIVERS problem in some areas (GOB 2019a). MAIN ROADS 0 50 100 150 Kilometers RAILROADS 0 25 50 75 100 Miles DISTRICT BOUNDARIES This map was produced by the Map Design Unit of The World Bank. INTERNA TIONAL BOUNDARIES ● Wildlife and national parks: Droughts The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank * The town councils of Francistown, Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Gaborone, Lobatse, and Selebi-Pikwe have status equal to Districts. 20 ∞E 25 ∞E affect rangeland and forage for wild ani- SEPTEMBER 2004 Source: World Bank 2015. mals, which can increase incidents of crop destruction by animals. The wild animal populations can be affected in DROUGHTS turn as farmers retaliate to protect their Drought is considered Botswana’s most livestock, farms, and families. serious natural hazard. It is a frequent occurrence locally, with particularly harsh C L I M AT E C H A N G E droughts affecting most of country in The incidence and severity of climate-related 1981–87, 1991–92, and 2018–19. An average hazards, particularly drought, are expected of 687,000 people (nearly one-third of the to worsen as a result of climate change. In population) are affected by drought annually, common with other southern African coun- and it affects areas that produce approx- tries, records indicate Botswana has gotten percent of the country’s GDP imately 31  considerably hotter in recent years. An anal- (CIMA 2018). Drought has multiple impacts ysis of climate data from 1970 to 2015 shows on individual lives and livelihoods and the an average temperature rise of around 1.5°C wider economy, including the following. (CIMA 2018). Projected trends based on current climate modeling predict a decline ● Reduced crop production: Agricultural in rainfall in much of the country and an areas are concentrated in the south- increased likelihood of drought. Increasing ern and eastern parts of Botswana, and drought will exacerbate overgrazing and soil droughts particularly affect the produc- degradation, which can lead to flooding from tion of root crops and vegetables. Over heavy rainfall as runoff rates increase. half the population lives in rural areas 32 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Records indicate climate variability is already which also lack clear and consistent guide- negatively affecting Botswana. Ten natural lines and associated training. A proposed disaster events occurred between 2000 and social protection harmonization process 2016 (CIMA 2018). Changing rainfall pat- should address this issue. terns are likely to have significant impacts on agricultural and livestock production, with Botswana lacks any current disaster risk man- associated consequences for food insecu- agement legislation and policy guidelines. As rity. Reduced rainfall is likely to affect stream a result, disaster risk management (DRM) flows in the main rivers and groundwater planning and response rely on outdated doc- reserves; combined with growing popula- uments, which, although adequate, need to tions, this is turn is likely to increase water be more current. The lack of policy guidance scarcity, further worsening food insecurity to cover wider shocks, such as COVID-19, is a and health vulnerabilities. major gap. That expanding social protection is an appropriate response to disasters, par- ticularly in terms of scaling up food assistance ASSESSMENT OF and public works, is generally accepted. Such ADAPTIVE S O C IAL approaches are not explicitly mandated with clear guidelines, but rather based on historic P R OTECTI O N trends—for example, the regular expansion of food assistance. On the positive side, the BUILDING BLOCK 1: structure and capacity for implementing INSTITUTIONAL DRM are good, with strong cross-sectoral involvement—including the Ministry of ARRANGEMENTS AND Local Government and Rural Development PA R T N E R S H I P S (MLGRD)—and a clear hierarchy of responsi- Although the constitutional and legislative bilities from the national down to the village framework for social protection in Botswana is levels. Also encouraging is the involvement of weak, commitment to a comprehensive social nongovernmental actors, including communi- protection system is strong. The National ties, in DRM decision making. Development Plan (NDP) and the National Social Protection Framework (NSPF) clearly Following are detailed findings and ratings demonstrate the government’s commitment justifications for this building block (see to improving the system. The institutional table 2.2). structures in place to deliver social protec- tion programs, however, are complicated by P O L I C Y A N D L E G I S L AT I V E the large number of such programs in opera- FRAMEWORKS: SOCIAL tion. Overall responsibility and accountability P R OTECTI O N for programs down to the beneficiary level are not always clear, as some straddle gov- Social protection is not explicitly enshrined ernment departments and local and national in Botswana’s constitution or legislation. agency control. While field-level delivery for Nonetheless, there is widespread consen- many programs is provided through a cadre sus that social protection is an investment in of well-qualified social workers, they struggle Botswana’s population and a key contributor with the administrative burden of implement- to the country’s sustainable development. ing the wide range of overlapping programs, As a result, social protection is explicitly 2 :   B otswana 33 mentioned in both National Development TA B L E 2 . 2   Scorecard for institutional arrangements and Plan  11 (GOB 2016) and Botswana’s Vision partnerships in Botswana 2036 (GOB 2019b). NDP 11 (2017–23) rec- Indicator Score ognizes social protection as a critical driver of social development and proposes mea- Composite country rating ● sures to harmonize programs, including Foundational indicators the development of a single social registry Social protection policy and legislative frameworks (SSR). The most relevant policy document relating to social protection is the NSPF, Disaster risk management policies and legislative frameworks ● approved by the Cabinet in 2020 (GOB Social protection institutional capacity and coordination 2018a). All social protection programs will Disaster risk management institutional capacity and coordination ● fall under the umbrella of the NSPF, which Adaptive indicators is a 10-year plan to establish a “comprehen- sive and well-coordinated social protection Clear mandate for social protection response ● system that is efficient, effective and sus- Multisectoral coordination for social protection responses ● tainable” (GOB 2018a). An interministerial Partnerships with nongovernmental actors ● Social Protection Technical Steering Com- Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is mittee has been established to provide between two stages of development. guidance and technical inputs for achieving the priorities set out in the NSPF, although it is not clear how regularly it meets in practice. P O L I C Y A N D L E G I S L AT I V E FRAMEWORKS: DISASTER The NSPF puts forth a comprehensive set of R I S K MANAG E M E NT goals and objectives for the social protection Botswana does not have specific legisla- system in Botswana. Key priorities include tion relating to DRM. Instead, the authority creating a more harmonized and coordi- to implement DRM actions is drawn from nated system that better addresses poverty, various pieces of legislation that address inequality, and vulnerability. Through the disaster risk reduction in one form or creation of a social registry, better target- another. Key among these are the Constitu- ing, and more comprehensive monitoring tion of Botswana and the Emergency Powers and evaluation, it aims to improve nutrition, Act. The constitution provides an overall health, and education outcomes and enable framework for the provision of safety and more people to graduate from dependence security, including from natural and manu- on social assistance. While the framework factured disasters, and thereby offers the makes limited mention of adapting current legal authority for ensuring the social and programs in response to disasters, propos- economic development that are essential to als for a social registry (discussed under reducing vulnerability. While the Emergency Building Block 3) are seen as important for Powers Act contains no guidelines specific more effectively targeting social protec- to disasters, it empowers the president to tion assistance to households in response make emergency regulations, which in prac- to disasters. The NSPF explicitly mentions tice is the mechanism often relied upon to disasters only briefly in the report—as one address them. Another legal instrument on of many risks facing all age groups—and which authorities have relied is the National “disaster relief” as a potential social protec- Disaster Relief Fund Order. This contains tion program to respond. 34 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA provisions for ensuring lives are sustained expired, also made limited mention of ASP during disasters and is often invoked to and is still awaiting revision or updating. address disaster risks affecting such sectors as agriculture, forestry, and health. C L E A R M A N D AT E F O R ADAPTIVE S O C IAL Three key policy documents guide the P R OTECTI O N management of disaster risk reduction in Botswana. The National Policy on Disaster Although no explicit constitutional man- Management, produced in 1996, advocates date exists for ASP as an ex post disaster for the integration of DRM measures into the response, acceptance is widespread that development process by planning projects expanding social protection provision is a to reduce the impact of future disasters, as default response. This widespread accep- well as by reducing disaster vulnerability at tance is largely due to the high reliance on all levels of governance—national, district, feeding programs within the social protection and local. It also establishes a framework to system (see the discussion of social protec- facilitate the preparation of local DRM plans. tion programs under Building Block  2). The improvement to the social protection systems The second policy document is the National proposed by the NSPF includes the provision Disaster Risk Management Plan, devel- of disaster relief as a social protection pro- oped in 2009. The plan provides guidance gram, specifying the use of cash transfers on planning and implementing responses with targeting mechanisms in the social reg- to disasters and on DRM, defining manage- istry that can be adjusted to reflect the nature ment structures for implementing DRM and of the disaster (meteorological data for nat- outlining the key responsibilities of different ural disasters, adjusted household targeting government and nongovernmental actors at score for other disasters) (GOB 2018a). the national, district, and local levels (GOB 2009). The plan endorses a decentralized This reference to social protection systems approach, with the active participation of should enhance their role in responding to district- and community-level initiatives. disasters, although the expansion of cash It is rather outdated and makes little or no transfers in response to shock has been lim- mention of adaptive social protection (ASP) ited, with the possible exception of public programs to build resilience or respond to works programs that pay participants in cash shocks. (see more under Building Block 2). DRM policy documents also support the provision The third document is the National Disaster of ex post social assistance but only reference Risk Reduction Strategy for 2013–18. Devel- food and other in-kind supports. The national oped in 2013 by the government and the DRM plan (GOB 2009) makes no mention of United Nations Development Programme, cash or any social protection programs. the strategy has contributed to reducing the risks of disasters by providing a framework I N S T I T U T I O N A L C A PA C I T Y for developing measurable indicators for A N D C O O R D I N AT I O N F O R the achievement of disaster risk reduction S O C IAL P R OTECTI O N components and objectives set out in it and the other two policy documents. The strat- The MLGRD is primarily responsible for the egy (GOB and UNDP 2013), which has now delivery of most social protection programs. 2 :   B otswana 35 The MLGRD works through local government information systems are inadequate, and councils to support policy, planning, and clear guidelines, training, and support are coordination in delivering social protection; lacking on such issues as beneficiary selec- social services; community development; and tion into and their exit out of programs, primary infrastructure. There are 16 districts referral of clients to other services or agen- in Botswana, which are coordinated by the cies, and implementation of entrepreneurship MLGRD at the national level. Local councils and microprojects (World Bank 2022a). are elected bodies and have devolved respon- In remote rural areas, a very wide range of sibilities for setting development priorities; responsibilities can lie with a single commu- overseeing local infrastructure development; nity development worker. Given this lack of and delivering a variety of services, includ- guidance for mainstream social protection ing primary education, primary health, water, programs, guidance in scaling assistance for and sanitation, and tertiary roads. District shocks clearly has not been considered. commissioners appointed by the central gov- ernment have executive authority, however. I N S T I T U T I O N A L C A PA C I T Y percent of The MLGRD provides up to 90  A N D C O O R D I N AT I O N : recurrent budgets (World Bank 2022a). DISASTER RISK MANAG E M E NT Nationally, the MLGRD has overall respon- sibility for most social protection programs Overall responsibility for DRM in Botswana and grants, but at the district level pro- lies with the Office of the President. The grams are delivered by multiple departments National Disaster Management Office rather than a single coordinating agency. The (NDMO) was established within the Office Orphans Care Program, the Destitute Per- of the President in 1998 and is man- sons Program, the Needy Students/Children dated to coordinate multiple actors in the Program, and disaster relief, for instance, development and implementation of multi- are managed by social workers and com- disciplinary DRM plans at national, district, munity development officers in the MLGRD and community levels. The NDMO supports Social Protection Department. On the other the development of tools, techniques, and hand, the Primary School Feeding Program systems to ensure implementation of such and the Vulnerable Group Feeding Program DRM activities as hazard identification, are managed by administrative staff in the emergency operations, and vulnerability and MLGRD Department of Finance and Pro- risk assessment. It monitors overall coor- curement. District staff report to the chief dination of the National Committee on executive officers of their districts rather Disaster Management (NCDM), the National than to a national agency for social protec- Disaster Management Technical Committee, tion. Expenditures are not monitored locally and district disaster management commit- at the program level, which undermines tees (DMCs) and facilitates the integration accountability and cost effectiveness. of disaster management into sectoral pol- icies and programs. The NDMO also takes Community-based staff working directly a role in funding arrangements for disaster with social protection beneficiaries have preparedness and relief (discussed under highlighted a lack of defined policies and Building Block 4). The current organiza- procedures for the many different social pro- tional framework for DRM in Botswana is tection programs that exist. In particular, shown in figure 2.1. 36 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA F I G U R E 2 . 1   Botswana’s current disaster risk management which are subordinate to the wider Dis- organizational framework trict Development Committee, provide guidance and technical support to village Office of the President National Disaster committees and are responsible for identify- Management Technical Committee ing disaster-prone geographical locations in National Disaster Management Office their respective districts and communicating National Committee on possible risk reduction measures to the Dis- Disaster Management District Disaster Management trict Development Committee. District DMCs Committee review and approve implementation reports of DRM activities before their submission to Village Development Committees the NDMO—and, in the case of drought, to Source: Adapted from GOB 2009. the MLGRD, which works closely with dis- trict DMCs to develop community-based DRM plans and incorporate risk reduc- Disaster management in Botswana benefits tion measures into development planning. from highly inclusive decision mechanisms Village development committees coordi- at all levels. The NDMO reports to the NCDM, nate DRM activities at the village level and which is the interministerial committee endorse community-based DRM plans and responsible for national DRM policy formu- risk reduction activities. The village commit- lation and coordination. The NCDM includes tees also raise awareness among community the MLGRD, which operates many of the members and support district DMCs and largest social protection programs. Coordi- the NDMO in organizing trainings for village nating closely with the NDMO, the NCDM volunteers for disaster response and pre- makes recommendations to the Office of paredness (World Bank 2019a). the President for national policies, plans, and strategies related to disaster risk reduc- M U LT I S E C T O R A L tion and for funding disaster preparedness C O O R D I N AT I O N F O R S H O C K and relief activities. The committee con- RESPONSE sists of deputy permanent secretaries from all relevant ministries and representatives The inclusion of all ministries on the national- from police and defense forces, the United and district-level DRM committees provides Nations (UN), and nongovernmental orga- a good basis for multisectoral coordina- nizations (NGOs). NCDM members endorse tion. The MLGRD is particularly important, disaster preparedness and relief activities as its social workers are the first respond- and can be deployed to the National Emer- ers at the village and district levels and often gency Operations Centre when it is activated undertake postdisaster damage and needs in an emergency. The NCDM is supported by assessment in emergencies. This struc- the National Disaster Management Technical ture puts the MLGRD in a key position as a Committee, a multisectoral technical advi- bridge between the district DMCs and the sory body composed of ministry specialists. NDMO in the management of ex post disas- ter response activities. The ministry is also in This multisector approach has been rep- a good position to support ex ante risk mit- licated at the local level, with a total of igation actions, including ASP, as discussed 30 DMCs formed in the country’s 10 rural further under Building Block 2. and 6 urban districts. The district DMCs, 2 :   B otswana 37 PA R T N E R S H I P W I T H R E C O M M E N D AT I O N S F O R I N S T I T U T I O N A L N O N G O V E R N M E N TA L A R R A N G E M E N T S A N D PA R T N E R S H I P S I N ACTO R S B OT S WA N A The Botswana Red Cross Society, the United 1. Ensure that new or revised legislation and policy (social pro- Nations Development Programme, and the tection and DRM) outline the objectives and responsibilities of Botswana Council of Non-Governmental social protection and the MGLRD in disaster response. Organisations are all members of the NCDM. 2. Develop updated and relevant DRM legislation and policies, The Botswana Red Cross Society and NGOs including a strategy for disaster risk reduction, placing greater are also represented on many district- and emphasis on investments that reduce disaster risk and build village-level DMCs. All government policy resilience, including for households, through the provision of documents recognize the role of these agen- social protection. cies in all DRM activities, including disaster 3. Consider a high-level commitment in DRM and social protec- response. Nonetheless, the government of tion policy updates to the expanded use of emergency cash Botswana may want to consider developing transfers, diversifying from sole reliance on food assistance, ex ante agreements or contingency plans depending on what is most appropriate as determined by an outlining the roles and support that could assessment of needs. Such a transition would be in line with be played by nongovernmental actors— international trends in both the social protection and human- for example, the World Food Programme itarian sectors and would confer the following associated or other UN agencies in the event of major benefits, among others: increased efficiency gains relative to covariate shocks that would stretch govern- in-kind distribution by eliminating the need for storage, trans- ment capacity. portation, and so on; increased speed with which assistance can reach affected households, especially in the case of elec- tronic payments; increased beneficiary choice and agency BUILDING BLOCK 2: over the use of the assistance by enabling recipients to pur- PROGRAM DESIGN AND chase the items they need most; and indirect, positive impacts D E L IVE RY for local economic recovery through the stimulation of local Botswana operates a wide range of social markets. protection interventions, incorporating social 4. Consider ex ante agreements with nongovernmental actors to insurance, labor market and employment support ASP and other responses in the event of large covari- support, and social assistance programs. ate shocks that may stretch government capacity. The social assistance programs in particular have wide coverage, and they have substan- tially reduced poverty and inequality. Most criteria. They also typically use a mixture of are well targeted to the poor, with the sig- paper-based and computerized systems for nificant exception of the tertiary grants and selection, registration, and transfer deliv- sponsorships program, over half of whose ery. Several studies have recommended the beneficiaries come from the wealthiest two rationalization and harmonization of social quintiles. This program also has the high- assistance programs and the prioritization est budget allocation of all social protection of resources to those that target the poorest programs. (GOB and UNDP 2020a; World Bank 2013), but progress has been slow. A standardized Delivery of social protection is fragmented, proxy means test (PMT) to assess poverty and programs have overlapping eligibility across programs was developed in 2009 38 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA (and subsequently updated) but is still not Persons Program grant into a form of tempo- widely applied. Social assistance remains rary social support grant that can be scaled heavily dominated by feeding programs and up quickly to all registered households in programs offering in-kind benefits rather response to shocks. than those that support graduation out of poverty. Shifting to a single delivery system Following are detailed findings and ratings using Visa cards that can hold cash and justifications for this building block (see in-kind benefits has only just started. table 2.3). Social protection responses to disaster in S O C IAL P R OTECTI O N Botswana are also dominated by the ex post PROGRAMS OVERVIEW expansion of feeding programs, because of which the adaptive dimensions of programs Botswana has a mature social protec- and their delivery systems have been scored tion system with a long history of providing as nascent or emerging, at best. The social centrally funded, locally delivered social protection response to COVID-19 entailed protection programs. A recent World Bank the delivery of food packages to nearly assessment of Botswana’s social protec- two-thirds of households and expanded tion system identified 29 different programs school feeding programs. A post-COVID implemented across nine ministries (World social protection recovery plan has recom- Bank 2022a). These include a range of cash mended modifying the temporary Destitute transfers, in-kind transfers, feeding pro- grams, fee waivers, public works, programs to enhance employability and livelihoods of youth and other adults, and pension pro- TA B L E 2 . 3   Scorecard for programs and delivery systems in grams. They can be placed under three broad Botswana categories: social insurance, labor market and employment programs, and social assis- Indicator Score tance. This assessment focuses primarily on Composite country rating ● social assistance programs, which form the Foundational indicators foundation for ASP. They are discussed in Benefit adequacy ● detail below, following brief descriptions of the other two categories. Eligibility criteria and targeting Coverage of the main social assistance program ● The main government-operated social Financial inclusion and national ID insurance program is the Botswana Public Officers Pension Fund. This fund covers Social protection payment systems public servants at the national and local Social protection delivery systems levels on their retirement at age 60. It is now Adaptive indicators a defined contribution pension scheme with Experience of vertical expansion 162,056 members. Such a program has lim- ited capacity for adaptation in response to Experience of horizontal expansion shocks and disasters, and, as ex-civil ser- Protocols and contingency plans ● vants, its beneficiaries are least likely to be Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the vulnerable to them. rating is between two stages of development. 2 :   B otswana 39 Botswana’s labor market and employment protection transfers were eliminated (World programs aim to address the country’s seri- Bank 2022a). Among the social assistance ous unemployment problem. According to program incomes examined in the Botswana the 2015/16 Botswana Multi-Topic House- Multi-Topic Household Survey 2015/16, the hold Survey (GOB 2018b), the situation is Old Age Pension makes the highest contri- particularly severe for youth, among whom bution to the fight against poverty. Without the unemployment rate at the time of the the pension, the national poverty headcount survey was an estimated 34 percent. Con- would increase from 16.0 percent to 18.1 per- sequently, the government has established cent, while the poverty gap would rise from nine youth job creation programs and at percent to 5.5  4.6  percent (GOB 2018b).3 least three employability programs. The Primary School Feeding Program also reduces poverty significantly, with the sim- Thirteen main social assistance programs ulated poverty headcount and poverty gap are currently operating in Botswana. They percent and 5.2  going up to 17.6  percent, represent a mixture of targeted and univer- respectively, in the program’s absence. The sal benefits, offering a range of cash and poverty reduction impact of these two pro- in-kind transfers. The programs are summa- grams is achieved through a combination of rized in table 2.4, which ranks them in order program size, poverty targeting, and trans- of total expenditure and enumerates the pop- fer amounts. While the adequacy ratio of the ulation assisted where figures exist. As the benefits they provide is not high, the pro- table shows, the tertiary grants and sponsor- grams are sufficiently large to be effective. ships program is the most expensive, despite having a relatively low caseload. The schol- Generally, implementation of so many social arships and sponsorships it offers pay for protection programs raises operational chal- tuition fees, books, and school supplies, as lenges. A recent World Bank (2022a) review well as provide a living allowance and cover highlights several problems, including vari- other expenses, such as clothing, medi- ations in selection criteria and benefits cal fees, and insurance. The program is not depending on staff interpretation in differ- means tested but is available to any youths ent locations and programs. The multiple achieving the requisite score on their sec- methods of benefits delivery (discussed ondary school leaving exams; thus, it does below) make it difficult to keep track of indi- not explicitly target the poorest children. By vidual transfers and overall program costs. contrast, the Old Age Pension and school Updated figures for beneficiaries of all the feeding programs have far more beneficia- programs are not readily available, which ries but represent much smaller portions of reflects the lack of any centralized admin- GDP. istrative management information system (MIS). The predominance of feeding pro- Social protection programs have a significant grams has been questioned, particularly the impact on poverty and inequality reduction in Botswana. On average, social protection programs contribute to the reduction of the 3 The poverty gap is the mean shortfall in income poverty rate by nearly 8 percentage points. or consumption from the poverty line (counting This means the poverty rate of 16 percent, the nonpoor as having zero shortfall), expressed as a percentage of the poverty line. This mea- calculated using the national poverty line sure reflects the depth of poverty as well as its ,would go up to almost 24 percent if social incidence. 40 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 2 . 4   Social assistance programs in Botswana Cost Implementing Benefit No. of Program agency provided beneficiaries P millions $ millions % of GDP Destitute Persons (Food Basket)a MLGRD Food, electronic 38,973 351.9 32.7 0.18 food card, in-kind Destitute Cash Allowance MLGRD Cash 37,556 90.1 8.4 0.05 Old Age Pension a MLGRD Cash 126,424 590.0 55.0 0.30 World War II Veterans Allowance a MLGRD Cash 1,270 9.7 0.1 0.005 Orphans Carea MLGRD Electronic food 24,351 222.1 20.6 0.11 card, social care Community Home-Based Carea MLGRD Electronic food 1,252 15.0 1.4 0.01 card, social care Disability Allowancea MLGRD Cash 7,805 29.9 2.8 0.02 Ipelegeng (public works program) a MLGRD Cash 73,045 646.7 60.1 0.33 Vulnerable Group Feeding Programb MLGRD Food — 714.0 70.0 0.38 Primary School Feeding Programb MLGRD Food 268,761 Secondary School Feeding Programb MOBE Food — 308.5 30.2 0.16 Tertiary grants and sponsorships b MOTE Sponsorships — 2,151.0 210.9 1.13 and scholarships Early Childhood Developmentb MLGRD Day care/nursery — 0.4 0.039 0.0002 Source: World Bank 2022a. Notes: — = not available. MOBE = Ministry of Basic Education; MOTE = Ministry of Tertiary Education, Research, Science and Technology. Beneficiary data for all programs are from fiscal year (FY) 2019/20. Data for the Vulnerable Group Feeding Program and the Primary School Feeding Program could not be obtained independently. The number of beneficiaries for the latter is from FY 2012/13. Ipelegeng beneficiary numbers are estimated from the Quarterly Multi-Topic Survey Labour Force Module Report 2020, Q1 (GOB and UNDP 2020b). a. Expenditure data are from FY 2018/19. b. Expenditure data are from FY 2017/18. Vulnerable Group Feeding Program, which S O C IAL P R OTECTI O N provides take-home rations to vulnerable C O V E R A G E , TA R G E T I N G , children aged six months to five years, preg- A N D B E N E F I T A D E Q UACY nant and lactating women, and some others. percent of all house- An estimated 56  The program is targeted based on nutri- holds in Botswana are covered by some tional status, with no means testing. Despite form of social protection program, rising this program’s being in place since 1988, percent for households in the poor- to 79  malnutrition rates in Botswana remain stub- est quintile. This compares favorably with bornly high. A recent study (UNICEF 2019) other upper-middle-income countries, recommended improvements to reduce inef- including those in the subregion. Social pro- ficiencies and enable the program to support tection coverage of the poor is dominated local economic development. by social assistance programs that account for 76 percent of the coverage for the poor- est quintile. This is because social insurance (the Botswana Public Officers Pension 2 :   B otswana 41 Fund) only covers ex-civil servants, who are F I G U R E 2 . 2   Total and poor populations covered by Botswana unlikely to be in the poorest groups; and social assistance programs, 2015/16 (%) labor market programs cover only very small Poor portions of the population. All social assistance Total School feeding (primary) Social assistance coverage in Botswana is generally progressive and well targeted to Old Age Pension poorer households. As shown in figure 2.2, the Old Age Pension and Primary School Ipelegeng Feeding Program together provide the great- School feeding (secondary) est coverage, reaching nearly 48 percent of the poor and 32 percent of the total pop- Destitute Persons (in-kind) ulation. There is, however, a good deal of overlap between these—and several other— Needy student social assistance programs. This reflects the Orphan care (in-kind) existence of many multigenerational house- holds and the many elderly grandparents who Student allowance raise grandchildren and sometimes orphans. As most other social assistance programs Tertiary scholarship are poverty targeted, the same groups or 0 20 40 60 80 100 households are likely to be eligible for mul- Percent tiple programs—for instance, Ipelegeng, the Source: World Bank 2022a. Destitute Persons Program, and the Orphans Care Program. The tertiary grants and spon- sorships (also known as student allowance same poor populations represents a duplica- and scholarships) are the exceptions to the tion of effort and an inefficiency of the current progressive nature of social assistance in social protection system. In 2009, the gov- Botswana. More than half of the beneficia- ernment developed an initial standardized ries of these programs are from the upper PMT that could be used to identify house- two quintiles because students from poor holds’ poverty status objectively and created households are less likely to go on to tertiary a standard assessment questionnaire for all education, given their higher school dropout programs. The digital questionnaire, admin- rates and lower education attainment levels. istered using smartphones, was successfully field tested in 2016 and has since been Most of the largest social assistance pro- updated using the latest household poverty grams are categorical and thus available to data (GOB 2018b). For reasons that are not all who meet the criteria. These programs clear, however, rollout and adoption by the include, for example, the Old Age Pension, programs has been slow. tertiary grants and sponsorships, and the school feeding programs. Several programs Social protection transfers represent an are means tested, and each has different important proportion of total household con- eligibility and selection criteria, although sumption for the poorest quintile. Again, they often target the same households. The social assistance grants are particularly administration of multiple targeting pro- important in expanding household consump- cesses by the same social workers for the tion for the poor, particularly those who are 42 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA eligible for several programs. Cash transfers mature and well-functioning civil registra- range from approximately P 300 ($28) per tion system, with over 96 percent of adults month for the Destitute Persons Program to having a national ID. Each individual is up to P 600 ($60) for the Ipelegang public assigned a unique identification number works program. Programs providing in-kind at birth that appears on a national ID card commodities and food allowances provide issued at age 16. Such a comprehensive a similar level of support. The World Bank system provides a useful basis for improv- (2022a) estimates that social protection pro- ing social protection beneficiary information grams bring the poverty headcount down in systems and developing an SSR (discussed Botswana by 8 percentage points from what under Building Block 3). Financial inclu- it would be without the transfers—16 percent sion is still quite low, with only 41 percent of rather than 24 percent. Again, most of this adults having active bank accounts (World drop (6 percentage points) is attributable to Bank 2020c). And although over 83 percent social assistance programs. This reinforces of adults have mobile phones, the use of the value of transfers targeted to the poor- mobile money is also relatively low, with only est households. 42 percent of adults reporting having used it in 2017 (World Bank 2020c). S O C IAL P R OTECTI O N The Social Benefits Registration System P AY M E N T A N D D E L I V E R Y (SOBERS) currently serves as the main pay- SYST E M S roll system for cash benefits. These benefits Botswana’s social protection programs come from four programs: the Old Age Pen- deliver in-kind and cash transfers using a sion, the Destitute Persons Program, the plethora of systems operated by a range of World War II Veterans Allowance, and the different ministries. Individual payments are Disability Allowance. Recent years have made as direct deposits to bank accounts, seen efforts to expand and consolidate pay- through post offices, or at pay points of ment systems to improve monitoring and payment service providers. In-kind benefits accountability. From 2021, payments should (including, for example, the in-kind com- be carried out by BotswanaPost and the ponents of the Destitute Persons Program, Botswana Savings Bank, which are para- the Orphans Allowance Program, and Com- statals. Beneficiaries who have personal munity Home-Based Care) and other adult bank accounts can be paid through these. In food grant benefits are provided through the future, all beneficiaries will be issued a a SmartSwitch card. SmartSwitch cards, Botswana Savings Bank Visa card that allows which look and function just like automated withdrawals from any ATM that accepts Visa. teller machine (ATM) cards, can be loaded The objective, accelerated by the COVID-19 with monthly food or other allocations in the crisis, is to avoid crowding at pay points. It form of electronic vouchers. SmartSwitch is also proposed that Visa cards replace the electronic food coupons can be used for SmartSwitch cards for beneficiaries receiv- point-of-sale purchases in designated pri- ing in-kind and food benefits. Although Visa vate shops. cards can also be loaded with e-vouchers for noncash benefits, the ultimate aim is to It is unclear why the shift to cash-based move to a fully cash-based system (World transfer delivery in Botswana has been Bank 2022a). so limited, given that the country has a 2 :   B otswana 43 EXP E R I E N C E I N ADAPTIVE DRM responses in collaboration with district S O C IAL P R OTECTI O N and village development communities and P R O G RAM D E L I V E RY social workers from the MLGRD. Local social workers are required to complete a rapid The linkage between disaster response postdisaster needs assessment of affected and social protection is relatively strong in households within 24 hours. The most usual Botswana, as community-based develop- immediate response is the registration of ment and social workers usually play a key affected households as temporary benefi- role in assessing and delivering disaster ciaries of the Destitute Persons Program. response. ASP responses are almost entirely This provides affected households with food ex post, however, and heavily dominated by relief, temporary shelter, clothing, medica- the expansion of feeding programs. As high- tion, and other necessities rather than cash. lighted above, drought is the most common Social workers are not involved in the distri- and widespread hazard, affecting large parts bution of food and goods; they do, however, of Botswana. Consequently, most ASP expe- help make the necessary arrangements for rience is related to drought response. In delivery or pickup from district storehouses. years when drought risk is high, the MLGRD In addition to assessment and registration, leads a drought and household food security social workers follow up with temporary vulnerability assessment. This involves data Destitute Persons Program beneficiaries to collection and field assessments of affected assess whether they should exit the program districts by multidisciplinary teams that meet (GOB and UNDP 2020a). with district DMCs to discuss and receive reports on drought severity and impacts. The The COVID-19 pandemic generated a cen- assessments generally recommend the fol- trally led response by the government of lowing ASP responses: Botswana following the president’s declara- tion of a state of public emergency on March ● Expansion of vulnerable groups feeding 31, 2020. In the weeks that followed, the programs (usually via health centers). government established a range of social These programs could be expanded protection initiatives to support the most vul- in terms of numbers fed or rations pro- nerable households, primarily the COVID-19 vided (that is, horizontally or vertically, Food Relief Program. With lockdown restric- respectively). tions leaving few or no markets functioning, ● Expansion of school and other feed- food aid needed to be expanded through ing programs. Increased rations or more the relief program, which made provision food relief packages are recommended for an estimated 325,000 food baskets for for affected areas. needy households for at least three months. ● Expansion of the Ipelegeng public works As this represented nearly two-thirds of program. The program would expand to all households, the program constituted a include more beneficiaries. Wage rates major scale-up of assistance achieved in a would remain unchanged. relatively short time. Efforts were made to ensure existing social protection programs In the event of faster-onset or more local- continued and that the Ipelegeng remained ized shocks such as floods, district- and operational. Emergency operations centers village-level authorities have the first respon- were established in all districts to facilitate sibility to react. District commissioners lead 44 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA and coordinate the provision of essential targeting, payment, grievances, and so forth. supplies to needy households. With the expansion of the SSR (discussed under Building Block 3), more households The economic impact of the strict lock- could take advantage of the proposed Tem- down on workers led the government to porary Social Support Grant. This modified put in place an economic response pack- version of the current “temporary” compo- 4  age estimated at P  million). billion ($378  nent of the Destitute Persons Program would This provided a range of economic support use the existing PMT tool to register house- to businesses and the economy, including holds on the SSR. The grant could supply the provision of a wage subsidy (approxi- immediate cash (and food) transfers to 1  mate budget of P  million) billion, or $95  households affected by a serious shock, and for employees of businesses most directly it would provide the backbone of a scalable affected by COVID-19, for a period of three (or shock-responsive) safety net that could months. The government also provided loan cope with any substantial covariate shock— guarantees to the most affected businesses; like COVID-19—that might strike the country facilitated access to credit; and introduced in the future. tax concessions, including a deferral of pay- ment for six months (UNICEF 2020). P R OTO C O LS AN D C O N T I N G E N CY P L A N S The COVID-19 response highlighted the weakness of Botswana’s fragmented social Nationally, no explicit protocols or guidelines protection system, as maintaining so many explain the role of ASP in disaster response existing programs through the crisis left little in Botswana. In theory, the MLGRD supports space to expand beyond them. The gov- district DMCs and village development com- ernment of Botswana and the UN recently mittees in undertaking community-based produced a National Social Protection DRM planning and risk reduction activities. Recovery Plan (GOB and UNDP 2020a), reit- These should also include disaster prepared- erating a World Bank 2013 social protection ness measures, including the development review’s recommendations for consolidat- of contingency plans and protocols for ing programs to provide a social protection disaster response. In reality, it is unclear how system that better addresses the vulnera- common such plans are and how consis- bilities of individuals throughout their lives. tently they are developed in practice. Given The plan proposes reducing social assis- the widespread and default reliance on food tance to five primary programs supported aid as a response, however, any plans that by a residual household-based safety net. do exist are likely to make some reference to The programs would be cash based and sup- this form of ASP. ported by common systems for registration, 2 :   B otswana 45 R E C O M M E N D AT I O N S F O R P R O G R A M D E S I G N A N D D E L I V E R Y I N B OT S WA N A 1. Implement the proposals in the NSPF and the “life-course framework” proposed in the National Social Protec- tion Recovery Plan (GOB and UNDP 2020a). These proposals would simplify and strengthen the foundations of the social protection system. Both documents propose measures to harmonize social protection systems for increased efficiency and effectiveness, creating a stronger platform from which to initiate social protec- tion responses to disasters. In particular, approaches to targeting (including continued rollout of the SSR using the revised PMT), registration, and payment delivery should become more standardized across programs. That said, the restructuring of social protection programs proposed in the National Social Protection Recovery Plan requires a full examination of the costs and benefits of the proposed reforms, along with an analysis of the avail- ability of the required fiscal space. 2. Create the proposed Temporary Social Support Grant (GOB and UNDP 2020a) to alleviate the current heavy emphasis on food relief, which is cumbersome to administer and deliver. The grant would be a modification of the current temporary component of the Destitute Persons Program, with a focus on poverty targeting and cash-based delivery. Accordingly, the recommendation should be implemented as a priority alongside the rollout of the SSR and the standardized PMT. The introduction of the Temporary Social Support Grant should entail the development of a shock response manual and operational guidelines for its use in response to a range of local- ized and substantial covariate shocks. 3. Create stronger linkages between the Ipelegeng public works program and Botswana’s other labor market inter- ventions to provide clearer pathways for graduating the working-age poor out of poverty, thus making them more resilient to shocks. 4. Consider reviewing or producing new guidance or criteria for the selection and design of the public works pro- gram’s projects to ensure a subset of the assets created by beneficiaries contributes to disaster risk reduction and climate change adaptation in high-risk areas. BUILDING BLOCK 3: proceeded since. When fully operational, D ATA A N D the SSR is envisioned to provide automated information on existing and future beneficia- I N F O R M AT I O N S Y S T E M S ries. This would improve the government’s Botswana has multiple social protection ability to respond to shocks with social pro- MISs, reflecting the fragmented nature of tection by using the applicants list in the programs operating across different govern- registry to increase benefits to existing ben- ment departments. The systems vary in their eficiaries, as well as expand programs to sophistication from paper-based lists to dig- additional individuals or households. itized databases, which limits the ability to aggregate or cross-reference beneficia- Information systems on disaster risk and ries across programs. The development of household vulnerability in Botswana are an SSR for Botswana is a key commitment weak. The government produces regular in the NDP and a priority in the NSPF. Cre- agrometeorological information products, ation of the SSR started in 2016, and the but the capacity to utilize and apply these registry was rolled out to four pilot districts; to assess disaster risk is limited. Even it is not clear how much farther rollout has where they exist, disaster risk assessment 46 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA and integrated DRM plans are very out- paper-based lists to digital databases linked dated. Consequently, early warning systems to payment systems such as SOBERS. As are led by nongovernmental agencies, pri- highlighted under Building Block 1, Botswana marily the Southern African Development has multiple programs operated by a range of Community (SADC) and the UN. The assess- different departments, each program with its ment of disaster needs tends to be ex own registration system. The programs are post, with multiagency needs assessments so varied that having a single system would commissioned during shock-affected peri- be neither possible nor desirable. The ability ods, particularly drought. ASP responses to cross-reference and compare beneficia- in the form of scaled food assistance and ries across systems, however, is regarded public works are a primary DRM response. as a valuable requirement. The government ASP needs are calculated by assessment fully recognizes the need to integrate MISs teams and district and village DRM commit- across programs as part of its framework for tees during field visits that follow no clearly harmonization and has established a generic defined methodology. MIS for most of its cash-based systems. This is a first step toward fulfilling the govern- Following are detailed findings and ratings ment’s commitment, articulated in the NDP justifications for this building block (see and the NSPF, to establish an SSR. table 2.5). Development of an SSR for Botswana com- menced in 2016, and the first phase of the pilot B E N E F I C I A R Y R E G I S T R AT I O N during 2019/20 showed promising results. SYST E M S A N D T H E S I N G L E The MLGRD is leading the development of S O C IAL R E G I ST RY the SSR in collaboration with the Department The quality of MISs in use by social pro- of Information Technology in the Ministry of tection programs in Botswana varies from Transport and Communications. The SSR will enable inclusion of the following: TA B L E 2 . 5   Scorecard for data and information systems in ● Dynamic updating of the registry from Botswana multiple data sources Indicator Score ● Beneficiary profiles at individual and Composite country rating household levels Foundational indicators ● Data on participation in social pro- Data and information on disaster risk and household grams, including details of benefits (cash vulnerability ● or in-kind) received over time at both household and individual levels Beneficiary registry ● ● Integration with the national civil registry Social registry for validation of identity and elimination Adaptive indicators of duplication Use of preexisting data for shock response ● ● Customized reports (for example, by pro- Post-shock household needs assessments ● gram and geographical area) Data-sharing protocols with internal/external partners ● Integration with the PMT (targeting) Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is assessment tool. between two stages of development. 2 :   B otswana 47 When fully operational, the SSR is envisioned exposure, and vulnerability of communi- as providing automated information on exist- ties to be used for social protection program ing and future beneficiaries. This would design (World Bank 2019a). To date, this has improve the government’s ability to reach not happened. households in disaster-affected areas and to allow increases in benefits for existing social Nongovernmental agencies play a greater protection beneficiaries—as well as the role than the government in producing and expansion of programs to additional individ- monitoring early warning information for uals and households using the applicants list Botswana. The SADC operates a subregional in the social registry. The registry thus would vulnerability assessment and analysis pro- improve the government’s ability to respond gram that provides regular synthesis reports to natural or economic shocks through its on the state of food security and nutrition social protection system. That said, rollout in southern Africa (SADC 2020). The World appears to be slow, and the existing system Food Programme and other UN agencies was not utilized in any scaled-up social pro- support production of reports to identify and tection support in response to the COVID-19 assess the impact of a full range of crises pandemic. and shocks on countries in the subregion. Given the large geographical area covered, recommendations for Botswana are high D ATA A N D I N F O R M AT I O N level and generic. ON DISASTER RISK AND HOUSEHOLD VULNERABILITY USE OF DISASTER RISK The Department of Meteorological Ser- M A N A G E M E N T I N F O R M AT I O N vices in the Ministry of Environment, SYST E M S TO I N F O R M Natural Resources Conservation, and Tour- ADAPTIVE S O C IAL ism produces agrometeorological data for P R OTECTI O N R E S P O N S E the country. The data are not, however, easily accessible or openly available on the The dearth of early warning information and department’s website, nor is it clear that the disaster risk assessments makes it difficult government of Botswana is currently system- to identify where or when such informa- atically collecting any significant data on ex tion has been used to inform ASP response. ante disaster risk and vulnerability or produc- SADC reports provide a summary number of ing any regular early warning information. the food-insecure population for each coun- try in the subregion every 6–12 months. More The latest disaster risk assessments for detailed national assessments appear to have each district in Botswana were completed in been undertaken in Botswana in response 2010. In 2019, the University of Botswana, to slower-onset disasters (droughts). In in collaboration with Oxfam and the Uni- years when these disasters occur, national versity of Cape Town, organized trainings drought and household food security of risk assessors for conducting new dis- vulnerability assessments are led by a Vul- trict risk assessments. These have since nerability Assessment Committee chaired been completed for the districts of Bobirwa, by the NDMO. The assessments are under- Chobe, and Mahalapye. The plan was for taken by multiagency teams comprising the pilot to be expanded to all districts to representatives of all key ministries as well acquire updated information about hazard, as the UN and other NGOs. As highlighted 48 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA under Building Block 2, the social protection complete more formal postdisaster damage recommendations from these assessments and needs assessment reports. While it is not focus almost exclusively on the expansion clear if any standard or formalized guidelines of food assistance and the Ipelegeng public actually exist for undertaking such needs works programs (GOB 2019a). Even these assessments, the current system may pro- recommendations are highly generic and do vide a natural link between DRM and social not specify numbers of beneficiaries. protection agencies at the field level, with social workers playing a key role in the iden- In the case of quicker-onset or more local- tification and coordination of assistance to ized disasters, social workers in the MLGRD affected households. are often the first responders at the dis- trict and village levels. Members of district DMCs and village development committees BUILDING BLOCK 4: FINANCE R E C O M M E N D AT I O N S F O R D ATA A N D At 3.6 percent of GDP, Botswana’s spending I N F O R M AT I O N S Y S T E M S I N B O T S W A N A on social protection compares favorably with that of other upper-middle-income countries. 1. Accelerate the rollout of the SSR to all parts of the country, par- This is lower than the 4.4 percent spent in ticularly those that are regularly disaster-affected. This will 2018 when the NSPF was completed, how- require all relevant social protection programs to have digi- ever. A key challenge is to ensure a greater tized and interoperable MISs so they can be linked with the proportion of social protection resources is SSR. Consider whether the SSR could link to the MISs of other focused on the social assistance programs government services (for example, health or unemployment pro- that do most to reduce poverty. grams) to expand the information on potential shock response beneficiaries for the Temporary Social Support Grant. The government of Botswana has no explicit 2. Undertake systematic disaster risk and vulnerability assess- budget lines allocated to DRM, although up ments for all parts of the country. The assessments should to $191 million can be classified as implicitly examine current and projected risks arising from climate addressing DRM objectives. Between 2013 change and the COVID-19 experience and can feed into and 2019, the government made no alloca- national DRM strategies and policies, as well as into integrated tions to disaster risk response and recovery. district DRM plans. All disaster or shock response expendi- 3. Use disaster risk and vulnerability assessments to identify the ture during this time was met directly by key early warning data required to monitor disaster risk and the responding government department in assess the likely impacts of shocks. Assessment of the likely the form of budget reallocations. Thus, the geographical and demographical impacts of various shocks resources available for shock response will upon household and individual welfare and consumption is a depend on the financial resources of the challenge that would benefit from a subregional initiative. The responding agency or authority. government of Botswana could work with the SADC or other subregional players to adopt or develop simple but robust Disaster costs are quantified ex post by methodologies by which information on recurrent climatic multiagency teams, including local DRM shocks is combined with poverty and vulnerability data to gen- committees and local social protection staff. erate more specific information on the location and scale (in It is not clear what guidance or methodol- terms of numbers of affected individuals or households) of the ogy is used to assess the number of affected ASP response required. individuals or households or the costs of 2 :   B otswana 49 the response required. Where significant TA B L E 2 . 6   Scorecard for finance in Botswana costs are incurred, a humanitarian appeal appears to be the default option. There are Indicator Score no pre-agreed disaster risk financing instru- Composite country rating ● ments in place for Botswana and none Foundational indicators specifically designed to finance a social Social protection spending ● protection response. It is not clear how the Disaster risk financing policies and strategies COVID-19 Relief Fund has been financed. Adaptive indicators Following are detailed findings and ratings Quantification of the expected costs for social protection responses ● justifications for this building block (see table 2.6). Disaster risk financing instruments linked to social protection ● Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. SPENDING ON SOCIAL P R OTECTI O N Botswana’s spending on social protection F I G U R E 2 . 3   Social assistance program expenditures as compares favorably against the average for percentage of Botswana’s total budget upper-middle-income countries. Botswana is also one of the few countries in Africa 2% 1% whose social protection programs are fully 4% funded by the government. The country has, 7% however, seen a gradual decline in social Tertiary sponsorship protection spending in recent years, drop- School feeding ping from 5.2 percent of GDP in 2009/10 11% Ipelegeng to 3.7 percent by 2017/18. Most social pro- 43% Old Age Pension tection expenditure is on social assistance, Destitutes food comprising 2.6 percent of GDP in 2017/18. 12% Orphans As discussed under Building Block 2, social Destitutes cash assistance has been shown to have the great- 20% Disability est impact of all social protection spending in reducing poverty. It is worth consider- ing, therefore, how to ensure this element Source: World Bank 2022b. of social protection spending is ringfenced Note: Budget data do not include administrative costs. or expanded effectively—taking particular note of the regressive nature of the tertiary grants and sponsorships program. Indeed, DISASTER RISK FINANCING as shown in figure 2.3, the tertiary grants P O L I C I E S A N D S T R AT E G I E S and sponsorships program currently has the A N D B U D G E T A L L O C AT I O N S highest expenditure of these social assis- tance programs (1.13 percent of GDP), while Technically, the NDMO is entitled to endorse the remaining programs are considerably and recommend ex ante DRM funding for smaller (covering less than 2.5 percent of costs related to disaster preparedness the population) with expenditures of less and capacity building from the Ministry of than 0.5 percent of GDP. Finance and Development Planning. Other 50 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA ministries should also have contingency bud- methodology used to calculate the number gets for DRM, but it is not clear that they do. of individuals or households affected and A recent review by the United Nations Office the costs of any ASP response is not clear. for Disaster Risk Reduction found that while Reports are submitted to the NDMO, which no element of the national budget is explic- proposes a recovery and rehabilitation plan itly earmarked for such activities, an annual to the NCDM. Most response costs are million in government average of $191.7  covered primarily by the relevant local or expenditures could be categorized as DRM sectoral department, using budget reallo- expenditure (UNDRR 2020). 4 Investments cation rather than any form of additional that could be defined as directly related to or contingency budget. In the case of more disaster risk reduction projects only repre- significant disasters, the NCDM may make percent of the total, however. sented 9.2  recommendations to the Cabinet for addi- This equated to $17.7 million annually and tional funding or ask to make a wider appeal was accounted for by a single program: the for more funding from th SADC, international social protection and disaster preparedness humanitarian and relief organizations, and program located in the Ministry of Pres- others. idential Affairs, Governance, and Public Administration. Most disaster risk reduction In response to the COVID-19 pandemic, the expenditure was related to ongoing devel- government established the COVID-19 Relief opment expenditures in ministries such as Fund. The private sector, individuals, and land management, water and sanitation, or donors were encouraged to add to the initial agriculture—all of which can be categorized investment of P 2 billion ($189 million). This as disaster risk mitigation or preparedness fund was used to pay for the additional food spending. The review also found that, in the assistance provided to 350,000 households five years up to 2019, the government made as COVID-19 relief. It is not clear how this no allocations for disaster risk response and level of assistance was established (World relief and reconstruction and recovery. This Bank 2020a). would seem to reflect a relatively low level of disaster incidence and impact. DISASTER RISK FINANCING MECHANISMS FOR SOCIAL When disasters do occur, requests for fund- P R OTECTI O N ing are calculated at the field level. Such calculations are done either by the district There are no pre-agreed disaster risk financ- development committees or in line with the ing instruments in place for Botswana and vulnerability assessment process described none specifically designed to finance a in Building Block 3. Such reports quan- social protection response. tify the assistance needed, including social protection elements, particularly food assis- tance but also expansions in other social protection cash projects. The guidance or 4 By applying guidance from the Sendai Frame- work for Disaster Risk Reduction 2015–2030 (UNDRR 2015). 2 :   B otswana 51 R E C O M M E N D AT I O N S F O R F I N A N C E I N B O T S W A N A Botswana’s expenditure on social protection reflects the high political and fiscal priority placed on expenditure in this sector. By comparison, the government has not prioritized disaster risk financing efforts to the same degree. The following are recommended to strengthen finance for ASP in Botswana: 1. Considering the ongoing fiscal situation, ensure that social protection spending is allocated efficiently, reorient- ing spending toward programs that reach the poorest households. Review current program expenditures and consider reallocations away from regressive programs, such as the tertiary grants and sponsorships program, so the government’s scarce resources are prioritized for those most in need. In that regard, the rollout of the PMT and the SSR should be prioritized as vital tools for improving allocative efficiency. 2. Undertake a disaster risk diagnostic review as a precursor to the development of a national disaster risk financ- ing strategy. This would examine in more detail the disaster risk faced by Botswana and its potential associated costs. By developing an ability to calculate costs of disaster risk ex ante, the government can better devise strategies and plans that more efficiently allocate resources to disaster preparedness, mitigation, and response interventions. 3. Using information collected as part of disaster risk and vulnerability assessments, undertake further analytical work to model and quantify ex ante the costs of disasters in Botswana. Analysis should focus specifically on the consumption, asset loss, and financial resilience of poor households and be used to inform the nature of ASP responses that would most effectively build their disaster resilience. PRIORITY of climate change in the form of extreme weather events. INTERVENTIONS F O R B OT S WA N A While the social protection system is rel- atively well funded and provides good This analysis showed that ASP in Botswana coverage of the population, the system is is at a nascent to emerging phase. Overall, complicated by the large number of social the existing social protection system is rel- protection programs in operation. These pro- atively well established and comprehensive grams straddle government departments but with some gaps in elements that are and local and national agency control. They essential for an effective ASP capacity. are also skewed toward educational support, which benefits wealthier groups, and food The constitutional and legislative frame- assistance programs. At the field level, deliv- work for social protection in Botswana is ery for many programs is provided through weak; however, the government is strongly a well-trained cadre of social workers. committed to developing a comprehen- While expanding social protection is gener- sive social protection system. Similarly, ally accepted as an appropriate response to Botswana lacks current DRM legislation disasters, that response has so far focused and policy guidelines. Updating these docu- primarily on scaling up food assistance and ments is particularly important in light of the public works. COVID-19 pandemic and the growing impact 52 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA The NSPF makes clear commitments to ratio- National Social Protection Recovery nalizing and harmonizing social protection Plan. Consolidation should examine the systems. This includes the proposed use potential for each program to adapt in of a standardized PMT to assess poverty response to shock, and at least one pro- across programs; the establishment of an gram should be explicitly designed to do SSR; and the shift to a digitized, cash-based this. This should include examination of delivery system. While all of these initiatives the Ipelegeng public works program and would significantly support the development other employment and labor market pro- of ASP systems, implementation progress grams to assess their relative strengths, has been slow. National early warning and weaknesses, and success in provid- DRM systems are not well developed. Disas- ing temporary relief and graduating the ter response, including the assessment working-age poor out of poverty. of costs, is undertaken ex post—although ● Accelerate efforts to harmonize oper- a good system of DRM committees that ational systems across programs, include social protection players is in place particularly those for targeting, registra- at the local level. There are no pre-agreed tion, payment delivery, and monitoring. disaster risk financing instruments in place This should include rollout of the SSR for Botswana and none specifically designed and the standardized PMT tool. to finance a social protection response. ● Consider subregional collaboration on early warning data collection and analysis Table 2.7 summarizes the priority actions of disaster risk and household vulner- for both foundational social protection and ability. The government’s systems for ASP. The key areas for strategic investment collecting and analyzing data on disaster include the following: risk and vulnerability are currently lim- ited. Botswana is not alone in this regard ● Strengthen and update legislative frame- and therefore could consider working works for social protection and DRM. with other southern African states or the National social protection and DRM SADC to address this issue. policy should be updated or created to make explicit reference to the role of ● Undertake a disaster risk financing diag- ASP in mitigating and responding to nostic to examine opportunities to expand disaster risk. disaster risk financing. This should be a general examination that includes the ● Continue to consolidate and rational- costs of ASP. ize the many social protection programs in line with the recommendations in the 2 :   B otswana 53 TA B L E 2 . 7   Summary of all recommendations by building block for Botswana Building block Foundational social protection Adaptive social protection ■ Strengthen the constitutional and legislative ■ Ensure any legislative or policy documents for basis for social protection in Botswana social protection include explicit mention of the role and objectives of ASP in addressing poverty and ■ Update national DRM policy and strategies, vulnerability Institutional and ensure these make explicit mention of arrangements ASP as a response ■ Ensure national DRM strategies and policies are and partnerships reflected in integrated district DRM plans ■ Consider ex ante agreements with nongovernmental actors to support ASP responses in the event of large covariate shocks ■ Implement the proposals in the National Social ■ Modify the temporary component of the existing Desti- Protection Recovery Plan and elsewhere to tute Persons Program to create a dedicated Temporary consolidate social protection programs Social Support Grant, with an emphasis on increasing the provision of cash-based relief alongside, or in place ■ Redirect social protection funding from expen- of, food relief sive programs that do not effectively target the poor, such as those for tertiary education ■ Develop accompanying ASP shock response guidelines Program design that set out how ASP would operate in response to a and delivery ■ Review the potential for harmonizing the range of localized and substantial covariate shocks Ipelegeng program with other labor market programs so they play a stronger role in gradu- ating the working-age poor out of poverty ■ Move forward with efforts to harmonize target- ing, registration, payment, and other systems across social protection programs Accelerate efforts to roll out the SSR and wide- ■ Consider whether the SSR could link to MISs of other spread use of PMT government services (e.g., health or unemployment programs) to expand the number of potential shock response clients for the Temporary Social Support Grant ■ Conduct the systematic disaster risk and vulnerability Data and infor- assessments required for all parts of the country, exam- mation systems ining current and projected risks arising from climate change and the COVID-19 experience ■ Consider cross-national and regional efforts to map disaster risk and vulnerability, and identify the key early warning data required to trigger appropriate ASP responses Ensure that social protection spending is allo- ■ Undertake a disaster risk diagnostic review as a pre- cated efficiently, reorienting spending toward cursor to the development of a national disaster risk programs that reach the poorest households financing strategy ■ Use the diagnostic review to enhance the government’s Finance ability to calculate costs of disaster risk ex ante ■ Use such analysis to inform the nature of ASP responses that would most effectively build the disaster resilience of poor households E S W AT I N I A D A P T I V E S O C I A L P R O T E C T I O N A S S E S S M E N T C A S E S T U DY 3 OVERVIEW F I G U R E 3.1   Poverty headcount ratios in SACU countries OF RISK AND 30 29.2% 27.2% Poverty headcount ratio (%) HOUSEHOLD 25 VULNERABILITY 20 18.7% 14.5% Eswatini is a small landlocked country bor- 15 13.4% dering Mozambique and South Africa. It 10 has a population of 1.1 million people and is 5 classed as a lower-middle-income country, based on average per capita gross domestic 0 Botswana Eswatini Lesotho Namibia South Africa product (GDP). (2015) (2016) (2017) (2015) (2014) Source: DataBank: World Development Indicators. Note: Poverty headcount ratio refers to the ratio of the population that is earning or POVERTY consuming an amount below the poverty line. Poverty is measured according to the global poverty line of $1.90 per day in 2011 PPP U.S. dollars. Eswatini’s poverty level is higher than those of other Southern African Customs Union (SACU) countries. Measured by the interna- percent in rural areas (Raju and to 70.1  tional poverty line of $1.90 per person per Younger 2021). Poverty rates also varied day in 2011 purchasing power parity (PPP) markedly across administrative regions in U.S. dollars, the poverty rate in Eswatini the country, ranging from 51.5 percent in percent in 2016 (figure 3.1), the was 29.2  Manzini to 71.5 percent in Lubombo. With highest in the SACU subregion and 3.8 per- respect to age groups, 2016/17 poverty centage points higher than the average for rates were higher among young children lower-middle-income countries (Raju and (0–5 years of age), older children (6–14 Younger 2021). years), and older people (60+ years) than among youth (15–35 years) and adults of Poverty in Eswatini is geographically and prime working age (36–59 years). demographically differentiated. For 2016/17, for example, the overall poverty rate in percent, compared urban areas was 19.6  55 56 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA SHOCKS AND Eswatini’s population has been affected by DISASTERS droughts over the past two decades, rang- ing from 20 percent during the most recent Eswatini is vulnerable to several types of percent during the drought in 2019 to 96  shocks. Hazards affecting the country 2001 drought, which affected the whole include droughts, floods, disease outbreaks, country (figure 3.2). The El Niño phenome- windstorms, forest fires, and invasive spe- non of 2015–16 induced the worst drought cies (World Bank 2022b). Moreover, climate Eswatini had experienced in more than three change is expected to increase the fre- decades. It produced large crop and livestock quency and severity of weather-related losses and acute shortages of water; an esti- disasters through more intense tempera- mated 260,000 individuals in rural areas tures, more variable rainfall patterns, (30 percent of the rural population) required prolonged heat waves, and water scarcity urgent food assistance between April and (World Bank 2022b). June 2016 (Raju and Younger 2021). Drought remains the most recurrent, frequent, The COVID-19 pandemic has been projected and pervasive shock affecting the country. to increase poverty and food insecurity even Historical data show that among the different further. According to an Integrated Food hazards, it has caused the highest economic Security Phase Classification (IPC) report, losses and death tolls (table 3.1). The gov- ernment of Eswatini declared a national Nearly 262,000 people (22% of the pop- state of disaster due to drought in 2001, ulation) in Eswatini are experiencing high 2007, 2015/16, and 2019. Droughts occur levels of acute food insecurity (IPC Phase 3 nearly every second year, although recently or above) between June and Septem- ber 2021 and require urgent humanitarian droughts of varying severity have occurred assistance. Of these, 240,000 people are almost annually (World Bank 2022b). experiencing Crisis food acute insecurity (IPC Phase 3) and 22,000 Emergency acute Droughts increase the risk of nonpoor house- food insecurity (IPC Phase 4). An addi- holds falling into poverty and food insecurity tional 342,000 people (29%) are Stressed and decrease the chances for poor house- (IPC Phase 2). During the projected period (October 2021–March 2022), which corre- holds to exit poverty. A significant share of sponds with the lean season, the number of people expected to experience Crisis or worse acute food insecurity is expected to TA B L E 3 . 1   Incidence and impact of disaster events in increase by an estimated 5% from the cur- Eswatini, 1981–2019 rent levels of 262,000 to around 317,000. The likely impact of the COVID-19 pan- People affected Financial demic, high commodity prices and poor Hazard Occurrences Deaths (cumulative) impact ($) performance of the agricultural season will Drought 8 500 2,389,000 309,739,000 greatly influence this increase. (IPC 2021, 1) Storm 4 74 742,639 54,152,000 Flood 3 11 544,400 50,000 Economic recovery is expected to support a Wildfire 1 2 1,500 n.a. modest reduction in poverty levels, with the Epidemic 3 142 3,677 n.a. $1.90/person/day international poverty rate Total 19 729 3,681,216 363,941,000 projected to reach 28.4 percent in 2022 and Source: World Bank 2022b, based on EM-DAT database (1981–2019). 28.3 percent in 2023 (World Bank 2021c). Note: n.a. = not available. 3 :   E swatini 57 FIGURE 3.2  Disaster risk profile of Eswatini a. Risk frequency and severity b. Population affected by drought 60 100 50 Economic loss (million $) Drought 80 40 60 Percent 30 40 20 Storm 10 20 0 0 0 5 10 15 20 2001 2007 2015 2019 Frequency (years) Source: World Bank 2022b, based on EM-DAT database (1981–2019). Note: Bubble size in figure 3.2a is relative to the percentage of the population affected by the shock. ASSESSMENT OF Important efforts have been undertaken recently toward developing DRM frame- ADAPTIVE S O C IAL works and guidelines. They do not provide P R OTECTI O N concrete indications of how to use social protection capacity for DRM actions, how- ever, and coordination mechanisms in both BUILDING BLOCK 1: the social protection and DRM sectors face INSTITUTIONAL significant challenges and constraints on ARRANGEMENTS AND functioning effectively. These include the PA R T N E R S H I P S limited capacities of the Deputy Prime Min- ister’s Office (DPMO) and the National The institutional arrangements required for Disaster Management Agency (NDMA) to adaptive social protection in Eswatini are lead their respective sectors, coordination assessed as between nascent and emerg- structures that are complex and duplica- ing. The recent development of the National tive, and resource constraints. Partnerships Social Assistance Policy (NSAP) is promis- with humanitarian agencies and nongov- ing; with successful implementation, it can ernmental organizations (NGOs), under the improve the institutional foundations nec- NDMA’s leadership, have been fundamental, essary for adaptive social protection (ASP). particularly in relation to conducting vulner- Currently, social protection policies and ability and postdisaster needs assessments disaster risk management (DRM) legisla- and delivering relief to affected households. tive frameworks do not envision a role for They have also played a significant role in ASP, although the new NSAP does include capacity development. The NSAP, for exam- the objective of linking the social assistance ple, was developed with the support of the system to emergency preparedness and United Nations Development Programme disaster risk reduction mechanisms. and the World Food Programme (WFP), and 58 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA various NGOs have supported the NDMA in vulnerable, it does not explicitly articulate developing protocols for targeting relief to functions related to ASP. households (see discussion under Building Block 2). In contrast, the recently developed NSAP does include some ASP provisions. As the Following are detailed findings and rat- policy has not yet been formally launched, ings justifications for this building block it is still too early to know how and when it (table 3.2). will be operationalized. However, its devel- opment is promising for the sector (box 3.1). P O L I C Y A N D L E G I S L AT I V E DRM legislative frameworks do not give FRAMEWORKS social protection a key role in disaster pre- The role of social protection in responding paredness and response activities. The to shocks and promoting household resil- DRM Act of 2006 is the legislative back- ience is only mildly recognized in Eswatini’s bone of the sector. It establishes the main existing relevant policies. The 2010 National emergency management structures: the Social Development Policy recognizes the Ministerial Disaster Management Team, impact of disasters on human well-being: “It the National Disaster Management Coun- is national policy to mitigate the effects of cil, and the NDMA. Reportedly, the National natural disasters by ensuring preparedness Disaster Management Policy from 2010 is and providing appropriate support to vic- not fully functional yet (GOE 2015). Recent tims” (GOE 2010). Although the plan clearly efforts have gone to developing DRM poli- indicates that social protection should cies and frameworks. The WFP, for example, target vulnerable populations and that provided technical support to the NDMA people affected by disasters are considered for the development of the National Disas- ter Management Operational Framework (GOE 2019a)—which outlines a mechanism Scorecard for institutional arrangements and TA B L E 3 . 2   for coordination, accountability, and orga- partnerships in Eswatini nizational arrangements—and the National Operational Guideline for Emergency Man- Indicator Score agement (GOE 2019b)—which outlines the Composite country rating multiagency framework for emergency, Foundational indicators disaster response, and recovery at national Social protection policy and legislative frameworks ● and local levels. Despite the importance of these frameworks and guidelines, they Disaster risk management policies and legislative frameworks ● do not provide concrete indications of how Social protection institutional capacity and coordination ● to use social protection capacity for DRM Disaster risk management institutional capacity and coordination ● actions. Adaptive indicators Clear mandate for social protection response ● C O O R D I N AT I O N A N D C A PA C I T Y Multisectoral coordination for social protection responses ● The DPMO’s capacity to coordinate the social Partnerships with nongovernmental actors ● protection sector through the Social Welfare Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. Department is limited. Although the DPMO 3 :   E swatini 59 B OX 3.1   The National Social Assistance Policy The NSAP is grounded in several preceding docu- there is great potential for closer collaboration between the National Disaster Management Agency (NDMA) ments that set the policy context for social assistance and Deputy Prime Minister’s Office (DPMO) on the in Eswatini. Prominent among these are the Constitu- harmonization of humanitarian relief and social assis- tion (2005), the National Social Development Policy tance, in line with current thinking and best practice (2010), the National Social Security Policy (2018), and on making social protection more “shock-responsive.” the Strategic Road Map (2019). This includes sharing of data from the single registry to allow for social assistance programmes to be scaled up either horizontally (by registering new beneficiaries on The policy draws on three approaches to social assis- a temporary basis) or vertically (by paying higher trans- tance: the life-cycle approach, to ensure all individuals fers for a limited period to existing beneficiaries). The and vulnerable groups that are in need are covered delivery of humanitarian assistance would then be pro- and no one is left behind; the social protection floor, to vided through established social assistance platforms ensure social assistance benefits are adequate to meet for the registration, payment, and management of ben- basic income security; and transformative social pro- eficiaries. (GOE 2021) tection, to ensure social assistance is linked to efforts to promote livelihoods and achieve self-reliance where The launch of the NSAP is promising for the entire possible, along with the enhancement of social justice. social protection sector and for the development of ASP specifically. Although the policy is very compre- Among the policy’s several objectives is one spe- hensive and ASP is only briefly referenced, it can serve cifically related to ASP: to provide more effective as the framework from which to develop and invest in protection against shocks by linking the social assis- the ASP capacities outlined in this report. tance system to emergency preparedness and disaster risk reduction mechanisms. The policy states that Source: GOE 2021. is in charge of multisectoral coordination for and services are delivered regularly, pre- implementation of the National Social Devel- dictably, on time and in full, to every citizen and resident of Eswatini who needs and is opment Policy and the forthcoming NSAP, in entitled to receive such support from the practice it faces challenges in coordinating state. (GOE 2021) social protection across different ministries and government levels. Capacity constraints The NDMA’s capacity to lead the DRM sector are the result not only of inadequate resourc- and coordinate activities, including ASP pre- ing, but also the lack of clear strategies and paredness and response actions, is limited. policies and of the power to enforce any The NDMA is the central authority respon- such coordination. This issue is specifically sible for day-to-day DRM activities at the addressed in the NSAP: central level, as it was assigned by the DRM Act the role of “promoting an integrated and It is imperative that existing institutional coordinated system of disaster manage- arrangements are strengthened and new institutional mechanisms are established, ment focused on decreasing vulnerability to ensure that social assistance benefits and increasing preparedness and mitigation 60 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA capacity.” To fulfill this role, the NDMA needs While clear on paper, the respective roles of to lead the Inter-Cluster System, which was the NDMA and the Disaster Management Unit established to ensure intersectoral coordi- within the DPMO are not clear in practice. nation within the government and between Although their mandates differ (with the unit it and humanitarian agencies along sectoral being a regulatory body only), some overlaps lines (see below). According to the National in their activities occur in practice, and the Multi-Hazard Contingency Plan, however, NDMA sees its mandate challenged by the the NDMA has insufficient institutional Disaster Management Unit. This leads to a capacity to lead the Inter-Cluster because lack of collaboration between two key enti- its institutional structures and activities ties in the DRM sector. have not been effectively decentralized. One finding of the national progress report M U LT I S E C T O R A L on implementation of the Hyogo Framework C O O R D I N AT I O N A N D for Action, for example, is that most of the PA R T N E R S H I P S sectoral working groups formed through the National Multi-Hazard Contingency Plan at Partnerships and multisectoral engagements the technical level are not functional and do are coordinated through the Inter-Cluster not meet regularly (GOE 2015). System, as established by the National Multi-Hazard Contingency Plan. As noted Although the DPMO’s Social Welfare Depart- above, the NDMA is the coordinator of ment participates in the Inter-Cluster the Inter-Cluster System, which includes System, this has not translated into stronger the relevant government agencies, the coordination. The department’s participation United Nations, and national and interna- has not resulted in stronger collaboration tional NGOs. Each cluster of agencies and between this entity and the NDMA (and organizations in the system is responsible others) with respect to exchanging informa- for ensuring the implementation of emer- tion, conducting joint needs assessments, or gency preparedness and response actions supporting delivery. through all structures from central admin- istration to the regional and community Resource constraints limit the ability of key levels (GOE 2012). A strong collaboration actors to fulfill their mandates related to DRM exists between the NDMA and humanitar- and coordination. Lack of adequate resourc- ian agencies and NGOs, with a focus on ing for the NDMA and other bodies, including postdisaster responses. Key activities like subnational coordination mechanisms, is conducting vulnerability and postdisaster one of the main constraints to adequate needs assessments and delivering support DRM and ASP actions (GOE 2015). entail partnerships with humanitarian agen- cies and NGOs, under NDMA leadership. 3 :   E swatini 61 R E C O M M E N D AT I O N S F O R I N S T I T U T I O N A L A R R A N G E M E N T S A N D PA R T N E R S H I P S I N E S WAT I N I The main constraints on effective institutional arrangements in Eswatini are related to a lack of adequate policy and legislative frameworks and of capacity on the part of the leading agencies to ensure effective coordination. The fol- lowing are recommended to strengthen institutional arrangements and partnerships for ASP in Eswatini: 1. Operationalize the NSAP, and track and review its implementation progress. 2. Develop a strategy for ASP, in line with the NSAP, to establish a policy framework exclusively focused on defining the role of social protection in preparing for and responding to shocks. The strategy should be developed jointly by the DPMO and the NDMA, among other actors (including partners), and should establish the institutional arrangements needed for ASP, including coordination mechanisms, partnerships, roles, and responsibilities. 3. Conduct an institutional capacity assessment of the NDMA to determine the main bottlenecks in fulfilling its role as the leading agency in the DRM sector. The assessment should cover issues related to legislation and man- dates, the coordination structure, and staffing, as well as incentives and power relationships. It should also provide concrete recommendations and a roadmap for strengthening the NDMA’s capacity, and it should identify strategies for enhancing the collaboration between the NDMA and the Disaster Management Unit. 4. Revise and improve the Inter-Cluster System to ensure better intersectoral coordination, and identify concrete actions to improve the collaboration between the DPMO and the NDMA (and other entities involved in shock responses), such as joint contingency planning and information sharing. BUILDING BLOCK 2: wide coverage. The NDMA and its partners PROGRAM DESIGN AND also deliver support as part of the agency’s food aid program. A transition is under way D E L IVE RY to electronic delivery of cash assistance, Building Block 2 on programs and their deliv- including in the food aid program. Finally, the ery systems in Eswatini is assessed as being NDMA and its partners developed guidelines between nascent and emerging stages. On for social protection assistance during the the one hand, the limited number of social pandemic, including targeting criteria and assistance programs offered and the lack protocols, that can be the basis for future of a cash transfer program to support poor contingency protocols. households is a constraining factor for ASP. Accordingly, no contingency protocols are Following are detailed findings and rat- in place for social protection support in ings justifications for this building block times of crisis. And, although they have (table 3.3). been increased recently, benefit amounts remain low, which limits their potential Eswatini offers a limited number of social impact on household resilience building as assistance programs. Although a pilot pro- well. There have been a few experiences of gram was implemented between 2016 and vertical expansions in response to recent 2018, no cash transfer program is currently shocks, and programs such as the Old Age in place for children and/or poor households. Grant and the School Feeding Program have 62 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 3 . 3   Scorecard for programs and delivery systems in ● Neighborhood Care Points are Eswatini community-run centers providing free cooked meals and basic early childhood Indicator Score development services to needy, young Composite country rating OVC. The program is fully funded by Foundational indicators the WFP and administered by it through NGOs, with the collaboration of the Benefit adequacy ● government. Eligibility criteria and targeting ● The School Feeding Program provides Coverage of the main social assistance program cooked lunches for students in public Financial inclusion and national ID ● primary, secondary, and high schools. Social protection payment systems The program is administered by the Ministry of Education and Training and Social protection delivery systems ● implemented in 860 schools. The WFP Adaptive indicators provides meals to 24,392 children in Experience of vertical expansion ● 28 primary and 22 secondary schools. During some months in 2020, school Experience of horizontal expansion ● meals were disrupted by school closures Protocols and contingency plans ● in response to the COVID-19 pandemic Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is (WFP 2020a). between two stages of development. ● Emergency Food Aid is distributed by the NDMA during emergencies, typically in The main social assistance programs are response to droughts. In recent shocks, described below and in table 3.4. such as the pandemic, a greater share of the benefits was distributed in cash. ● The Old Age Grant is provided to all citi- zens of Eswatini once they reach age 60. C OVE RAG E, B E N E F IT With 75,232 beneficiaries and monthly A D E Q U A C Y, A N D TA R G E T I N G transfers of E 500 per beneficiary, this grant is by far the largest social cash Despite Eswatini’s relatively high coverage transfer program in the country. for social protection, ASP is constrained by the lack of a cash transfer program tar- ● The Disability Grant has only 7,715 ben- geted to the poor and/or children that eficiaries, accounting for approximately could be scaled up after a shock. The over- 60 percent of people living with disabil- all coverage of social protection and social ities in Eswatini (GOE 2021). It pays a assistance is relatively high in Eswatini com- monthly transfer of E 280 and stopped pared to neighboring countries, and is driven registering new beneficiaries some years by the School Feeding Program and the Old ago. Age Grant (figure 3.3). The School Feeding ● Orphans and Vulnerable Children (OVC) Program has the most beneficiaries, reach- Education Grants subsidize tuition and ing 79.3 percent of primary school students exam fees for OVC attending public sec- percent of secondary and high and 60.8  ondary and high schools. The subsidy is school students in 2016/17. The Old Age paid directly to the schools. Grant is the main cash transfer program 3 :   E swatini 63 TA B L E 3 . 4   Main social assistance programs in Eswatini Implementing Number of Program Description agency Targeting Benefit value beneficiaries Social pension Social Welfare Universal, age ≥60 E 500 ($33) 75,232 (2021) Old Age Grant Department monthly, paid quarterly Poor disabled Ministry of Social ■ Means-tested disabled E 180 ($19) 4,744 (2021) Disability Grant people Development monthly ■ Registration currently closed Grants to poor Ministry of Social Poor OVC in 260 second- E 1,950 ($130) per 58,000 (2019/20) OVC Education OVC Development ary and high schools year, paid directly Grants to schools Free meals and Ministry of Edu- OVC, pre-primary age Not applicable 54,481 (2020) Neighborhood basic early child- cation and Care Points hood development Training; WFP services to OVC School meals Ministry of Edu- Children in public pri- Not applicable 360,000–380,000 School Feeding cation and mary, secondary, and (2014–19) Program Training high schools (860 schools) Emergency NDMA Households selected Not applicable 72,745 households Emergency in-kind or cash based on food insecurity (Apr. 2016–Dec. 2017) Food Aid assistance and other measures of 49,150 (2021) vulnerability Sources: GOE 2021; Raju and Younger 2021; Schubert 2018. percent of in the country, covering 81.1  percent of tons of food commodities (87  those age 60 or older (Raju and Younger which was corn) to 72,745 households across 2021). Although categorical programs like 1 the country in response to drought. Food aid the Old Age Grant can be leveraged when coverage has been estimated, based on the responding to shocks, the lack of wider cash Swaziland Household Income and Expendi- transfer programs constrains mobilization of ture Survey 2016/17, at roughly 23 percent cash-based assistance through horizontal of severely food-insecure individuals. and vertical expansions after a shock. Benefit amounts are generally low, limiting Food aid reaches only a small percentage their potential for building household resil- of people identified as being severely food ience to shocks, although recent efforts have insecure. Between April 2016 and Decem- aimed to improve the adequacy of some ben- ber 2017, the NDMA distributed 4,834 metric efits. In early 2020, the Old Age Grant was increased from E 400 to E 500 monthly ($27 to $33) and the Disability Grant from E 180 280 ($12 to $19). Amounts are not to E  1 Older people are among the population groups with the highest poverty levels in the country; indexed to price inflation, and the increases based on 2016/17 data, 66.5 percent are poor have occurred at irregular intervals (Raju (Raju and Younger 2021). 64 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA FIGURE 3.3  Social protection coverage in SACU countries: direct and indirect beneficiaries 90 Botswana Eswatini Lesotho Namibia South Africa (2015) (2016) (2017) (2015) (2014) 80 70 % of population receiving benefits 60 50 40 30 20 10 0 All social protection All social assistance Cash transfers Social pensions School feeding and labor Source: ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database. and Younger 2021). In addition, although the Food distributed by the NDMA as part of the real value of the Old Age Grant increased by Emergency Food Aid program is targeted 250 percent between 2005 and 2021, cate- to those in need. The distribution of aid is gorical programs such as the Old Age Grant based on targeting criteria and a mecha- and the Disability Grant do not take house- nism developed jointly by the government, hold size into account. This leads to many humanitarian agencies, NGOs, and commu- of the extremely poor households reached nity leaders (box 3.3). The selection criteria by the Old Age Grant (which average seven comprise several indicators of household members) receiving very low benefits per socioeconomic disadvantage and distress member (Schubert 2018). related to household composition, member health and disability status, coping strat- Social assistance programs in Eswatini rely egies for food insecurity, farming output, mostly on categorical targeting approaches. assets, livelihoods, and income sources; Both the School Feeding Program and the these are accompanied by guidance on their Old Age Grant target specific segments of use to categorize households into groups the population—children and youth in the considered predictive of the severity of their former and older people in the latter—with- food insecurity. Food is expected to be dis- out any income requirement. The OVC Cash tributed to the households categorized as Transfer Pilot tested the use of a proxy means experiencing the severest food insecurity test (PMT), but the pilot ended in 2018, and (Raju and Younger 2021). it was not scaled up (box 3.2). 3 :   E swatini 65 B OX 3.2   OVC Cash Transfer Pilot (2016–18) Between 2016 and 2018, the DPMO implemented a and pay for education (by covering the costs of uni- pilot, fully funded by the World Bank, of a cash transfer forms, stationery, and school fees, allowing children program to support poor OVC between birth and age to attend school) and transportation to health centers. 18. The pilot, which reached 15,920 OVC by its end, Households were able to save, which made them more tested procedures and mechanisms that were in line resilient to shocks. Livelihoods improved significantly with international experiences, although innovative for for children and the families in general, while the com- Eswatinit; these included mobile payments and a digi- munities also benefited from the cash inflow and the tal management information system. general welfare benefits. The OVC Cash Transfer Pilot underwent a rigorous The pilot was discontinued after donor funding ended. impact evaluation, which assessed its main impact as The lessons learned may be useful in guiding the restoring the dignity of the most vulnerable members design and implementation of existing or future pro- of society by decreasing the gap between the poor- grams. Indeed, the experience with mobile payments est and other community groups. The cash transfer under the program has informed the adoption of this had a statistically significant impact on the treatment approach for other interventions in Eswatini. group, especially regarding the ability to buy food Sources: GOE 2018; Raju and Younger 2021; UNICEF 2017. S O C IAL P R OTECTI O N commits to continuing to shift the delivery D E L I V E RY SYST E M S of social assistance benefits from manual to electronic mechanisms where this is feasible The transition from manual to electronic and appropriate. It also proposes a review delivery of cash benefits has already started of the payment mechanisms of all social in Eswatini, which can speed up response assistance programs to identify opportu- mobilization. Before the pandemic, most Old nities to improve efficiency and reliability. Age Grants were transferred manually in two Financial inclusion and national ID coverage ways: on a quarterly basis in cash, through are essential for setting up and expanding local government offices or post offices; electronic payment mechanisms. Box 3.4 and on a monthly basis, through electronic discusses these issues in greater detail. transfers to the beneficiary’s personal bank account. Only about 20 percent of benefi- Food aid provided by the NDMA and its part- ciaries were paid via the electronic transfer ners is delivered both in cash and in kind. method (Raju and Younger 2021). In April Although the mechanisms differ depending 2020, manual disbursement of the Old Age on the implementing agency or NGO, the Grant was transferred to electronic pay- NDMA and the WFP have delivered cash via ments to minimize COVID-19 infection risks mobile money in response to the pandemic. (GOE 2021), although information about the In-kind benefits are collected by registered proportion of payments delivered electron- households at distribution points admin- ically during the pandemic is not available. istered by local government officials and The Disability Grant is presently disbursed NGOs (Raju and Younger 2021). entirely through mobile money. The NSAP 66 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA B OX 3.3   Food Response Plan for COVID-19 In response to the socioeconomic effects of the The food security sector members agreed on the tar- geting criteria that beneficiaries of the program shall COVID-19 pandemic, the Eswatini government devel- include but not limited to the following: oped a Food Response Plan, establishing criteria and processes for identifying people to be supported. ■ Household with no formal constant/sustainable The plan’s main provisions (lightly copyedited) are income excerpted below: ■ Household headed by an elderly person ■ Household hosting more than two orphaned and The NDMA, working with NGOs, will be reaching all vulnerable children Tinkhundla [constituencies] in the country with an ■ Household head is a chronically ill person with no emergency cash-based transfer program. Each imple- external support menting partner has been allocated an operational Inkhundla where it is going to do the targeting, bene- ■ Household head or breadwinner living with disability ficiary registration, local stakeholder engagement and ■ Household with no or limited assets that can be monitoring of the intervention. These partners include used to generate income the Africa Co-operative Action Trust, the Adventist ■ Household has a pregnant or lactating mother Development and Relief Agency, the Baphalali Eswa- ■ Temporary loss of income by household head (e.g., tini Red Cross Society, Caritas Swaziland, Nazarene street vendors) Compassionate Ministries, Save the Children, Woman Farmer Foundation, and World Vision Eswatini. ■ Temporary loss of remittances ■ Household income severely affected due to partial Beneficiary identification will be done by the imple- lockdown. menting partner in collaboration with community-level structures like the relief committees and community A household to be targeted for the COVID-19 assis- leaders at the chiefdom level. To ensure better tar- tance should at least have a combination of the above geting, the participation of community leaders and characteristics that will determine the degree of vulner- affected people is key for this response. Households will ability. A household with one or two characteristics is be ranked based on the severity of their food-insecurity not as vulnerable as a household with three or more status within the particular community. criteria. This can be discussed with the involvement of relevant stakeholders including the relief committee and community leaders in order to rank the households accordingly. Source: GOE 2020. P R OTO C O LS AN D the roles of government and nongovernmen- C O N T I N G E N CY P L A N N I N G tal entities. The plan was originally developed for the fiscal year 2012/13, but it is a living Eswatini has no contingency plan in place for document that has been updated a few times cash responses to shocks. Moreover, wider subsequently (World Bank 2022b). NDMA-led contingency plans lack specificity on social protection interventions. This is the In practice, responses are planned in the case with the National Multi-Hazard Contin- food security cluster before the beginning gency Plan, which provides a framework for of the lean season or when a shock such as implementation of preparedness, response, the recent pandemic hits. Under NDMA lead- mitigation, and recovery activities, including ership, humanitarian agencies and NGOs 3 :   E swatini 67 B OX 3.4   Access to financial services and national ID cards in Eswatini Compared to neighboring countries, Eswatini has the three dimensions of the economy—people, govern- lowest percentages of people with bank accounts and of ment, and business—and GDP per capita, shows that people who have used digital payments (figure B3.4.1). Eswatini, globally ranked at 298 out of 359 countries, In addition, the World Bank’s Digital Adoption Index, places lowest in the SACU subregion. which measures countries’ digital adoption across F I G U R E B 3.4.1   Financial inclusion and national ID coverage in Eswatini Botswana Eswatini Lesotho Namibia South Africa 100 80 60 Percent 40 20 0 Has account Has national ID Made/received digital payments Digital Adoption Index (% age 15+) card (% age 15+) in past year (% age 15+) Source: Global Findex Database, World Bank. Note: Eswatini data for account access are from 2011 and probably underestimate current coverage; other countries’ data are from 2017. Eswatini is the only SACU country for which the Global Findex Database does not have information on national ID. share information and coordinate responses, country’s social assistance programs, com- although no consolidated response plan is bined with its limited administrative capacity, produced. Such planning is informed by the reduce the scope to respond to shocks with food security assessments conducted by the vertical or horizontal expansions. This has Vulnerability Assessment Committee (see been demonstrated in the responses to two Building Block 3). recent shocks: the El Niño–induced drought of 2015–16 and the COVID-19 pandemic. S O C IAL P R OTECTI O N AN D In response to the 2015–16 drought, the H U M A N I TA R I A N R E S P O N S E S School Feeding Program expanded verti- TO S H O C KS cally. The program scaled up from offering Beyond the Emergency Food Aid program, lunch to also providing breakfast for one which is responsive by design, the govern- term (September–December 2016). The esti- ment of Eswatini’s few experiences in the mated cost of adding breakfast for this term use of social protection programs to respond was E 7 million (about $467,000) and was to shocks have tended to rely on exter- financed through the Ministry of Education nal support. The limited offerings of the within its annual and supplementary budget 68 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA allocations. The response encountered sub- cash transfer, with total direct cash amount- stantial delays, however; the first top-up was ing to E 25.27 million ($1.7 million) from the given seven months after emergency dec- Baphalali Eswatini Red Cross Society and laration. The main reason reported for the the WFP (UNICEF 2017). These responses delay was the lack of available resources were implemented in coordination with the to fund the additional meal (Drechsler et al. NDMA. 2019). The response to the COVID-19 pandemic In addition to using the OVC Cash Transfer included vertical expansion of the Old Age Pilot to cushion the effects of the 2015–16 Grant. The monthly allowance was increased drought, other cash responses were imple- from E 400 to E 500 ($27 to $33) in early mented by humanitarian agencies and NGOs. 2020 to cushion older people from the A total of 58,800 beneficiaries received effects of the crisis. A shift was made from cash support from the Baphalali Eswatini manual to electronic payment of the grants Red Cross Society and 27,955 beneficiaries to contain the spread of COVID-19 (Dhemba from the WFP. In total, 86,755 beneficiaries 2021). (15,391 households) were reached through The NDMA responded to the COVID-19 crisis with cash and in-kind transfers. The NDMA provided a cash payment of E 700 R E C O M M E N D AT I O N S F O R P R O G R A M ($47) to 14,750 households in 14 constit- D E S I G N A N D D E L I V E R Y I N E S WAT I N I uencies. Cash was transferred via mobile 1. Continue rolling out digital payment mechanisms for most money. Additionally, food was transferred beneficiaries of programs that transfer cash benefits, includ- at different distribution points to 34,400 ing shock response programs. Ensure digital mechanisms are households. Beneficiaries were selected sufficiently flexible for vertical/horizontal scale-up when the according to the criteria and mechanisms need arises. Consider designing and pre-positioning alterna- described in box 3.3. tive payment mechanisms to cope with disruptions to the main mechanism. Various partners delivered support comple- ment these efforts, reaching nearly 32,375 2. Develop contingency protocols for the Emergency Food Aid additional households. Implementing agen- program and a methodology for annual response planning at cies such as the WFP, the Baphalali Eswatini the beginning of the lean season, consolidating all responses Red Cross Society, and Save the Children, by the government and its partners. Continue the transition with funding from various donors, pro- from food to cash assistance. vided cash and in-kind assistance. A total 3. Develop contingency protocols for scaling up social assistance of 77 percent of the overall benefit values programs, linked to the proposed strategy for ASP (under the comprised transfers in cash, made through 1 recommendations). Both the School Feed- Building Block  electronic mechanisms. Transfer values for ing Program and the Old Age Grant have already been used to the cash benefits were agreed on by the respond to shocks, although without previous preparedness NDMA and the partners, based on informa- measures. Lessons from these experiences can be drawn on and tion from vulnerability assessments. guidelines formalized ahead of future shocks and scale-ups. 4. Consider designing and implementing a mechanism for regu- lar, seasonal cash support to households exposed to recurrent droughts. 3 :   E swatini 69 BUILDING BLOCK 3: TA B L E 3 . 5   Scorecard for data and information systems in D ATA A N D Eswatini I N F O R M AT I O N S Y S T E M S Indicator Score The information systems required for effec- Composite country rating ● tive ASP in Eswatini are assessed as nascent, Foundational indicators with the main weakness in this area related Data and information on disaster risk and household vulnerability ● to the lack of digital registries and informa- tion systems for the major social protection Beneficiary registry ● programs. Vulnerability assessments are Social registry ● focused on food insecurity and follow inter- Adaptive indicators national methodologies. However, the information they generate has some limita- Use of preexisting data for shock response ● tions for ASP: it is static, available only once Post-shock household needs assessments ● a year, and based on forecasts of future Data-sharing protocols with internal/external partners food needs. Further, while it leads to esti- Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is mates of people in food insecurity at various between two stages of development. administrative levels, it does not allow the identification of specific households. A pro- established information systems have cess is in place for conducting postdisaster digitized and electronic registries, high confi- food needs assessments at the household dence among users and the public regarding level, led by the government and in col- data quality, strong security mechanisms, laboration with various partners. The data and a high level of integration and interop- collected through this process inform individ- erability. From an ASP perspective, the ual responses by NDMA and partners, but are interoperability between social protection not consolidated or retained to inform future and DRM information systems can inform the planning and responses. Data are sometimes preparedness and response actions of both shared across actors but without protocols. sectors (Barca and Beazley 2019). Following are detailed findings and rat- Most social assistance programs in Eswatini ings justifications for this building block lack digital beneficiary registries with man- (table 3.5). agement information systems, which limits the possibility of sharing data or making S O C IAL P R OTECTI O N use of them for social protection responses. I N F O R M AT I O N S Y S T E M S For example, an evaluation of the national School Feeding Program commissioned by The capacity to collect, manage, and share the WFP and the Ministry of Education and data across social assistance programs Training in 2019 identified the lack of digital and beyond the social protection sector can data collection tools and of a database as a be fundamental for responding to shocks. key constraint for program monitoring (WFP Nascent systems are typically characterized 2020a). As a result of this finding, the Min- as being largely paper based and with few istry of Education and Training is planning security protocols, few controls on data qual- to develop and pilot a management infor- ity, no interoperability, and no social registry mation system (MIS). Moreover, the DPMO’s in place. At the other end of the spectrum, action plan up to 2022 emphasizes the 70 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA development and implementation of an inte- needs among the population. Data for the grated MIS for program monitoring. The lack assessments are collected through annual of such an MIS is seen as a major weakness field-based surveys conducted across Eswa- (Raju and Younger 2021); for this reason, tini’s four regions every year between May the NSAP proposes concrete actions in this and June during the post-harvest season. regard (box 3.5). “Hotspots” for data collection are identi- fied based on meteorological data received from the Eswatini Meteorological Service. RISK AND VULNERABILITY Preliminary results of the assessments are M A P P I N G A N D E A R LY disseminated in June, and the final report is WA R N I N G SYS T E M S available in July, before the start of the lean The vulnerability assessments led by Eswati- season in October. The main assessment ni’s Vulnerability Assessment Committee are is often complemented by additional data currently the main source of information for collection exercises conducted toward the ASP and DRM actions in the country. Under end of the year to adjust the original fore- NDMA auspices, the Vulnerability Assess- casts. The information from the vulnerability ment Committee collects household-level assessments is used for funding appeals and information to assess chronic food insecurity, also by the NDMA and partners in develop- malnutrition, livelihoods, and vulnerability in ing multiagency and intersectoral responses. rural households in all regions of the coun- Actors generally consider this information try. Vulnerability assessments are based robust and reliable. In addition to the vul- on the Household Economy Approach, nerability assessments, an annual National which also can be used to determine food Agricultural Survey is conducted in collab- oration with the Central Statistical Office, employing the Household Economy Analysis B OX 3.5  Actions proposed to improve Eswatini’s and Household Survey to provide informa- information systems tion on cropped areas. This survey is not The NSAP establishes that the government will commission a consistently conducted every year, however, review of options for developing a unified single registry or MIS due to resource and technology constraints that compiles information on all social assistance programs in (World Bank 2022b). Eswatini within a single database. The following are among the activities indicated in the policy: The capacity to produce and share early warning data to inform ASP and DRM inter- ■ The DPMO will work with the Ministry of Home Affairs to ventions is very limited. Although both the ensure the availability of accurate and comprehensive identi- Meteorological Service and the Ministry of fying information from civil registration records for all citizens Agriculture’s Early Warning Unit generate and residents. and provide early warning data, the early warning system faces many constraints. ■ Manual data sets and registries of beneficiaries of existing These constraints range from deficiencies social assistance programs and complementary initiatives will in data collection, management, and storage be captured electronically. to limited provision of timely and actionable ■ All relevant data sets will either be linked electronically for the al. early warning information (Drechsler et  MIS or merged into a unified database for the single registry. 2019). A lack of qualified personnel to run Source: GOE 2021. the weather stations and analyze the data also limits system capacity (GOE 2015). 3 :   E swatini 71 U S E O F P R E E X I S T I N G D ATA NGOs. A number of sectoral postdisaster TO I N FO R M ADAPTIVE needs assessments are conducted in the S O C IAL P R OTECTI O N country, with varying degrees of quality and RESPONSES reliability. The strongest are the food secu- rity assessments because they are linked The use of preexisting data can allow quick to the vulnerability assessments. Postdis- response, especially when shocks are rela- aster assessments of food security are led tively predictable and recurrent. They can by the NDMA and carried out by NGOs and include social protection data (for example, agencies, based on selection criteria and from social registries) or DRM data (from procedures jointly developed by the NDMA vulnerability assessments), among others and a consortium of partners (see box 3.3). (Barca and Beazley 2019). Eswatini has no single database that con- Overall, preexisting household-level data solidates information from the postdisaster to support and enable faster responses are needs assessments conducted by different lacking. The only recent experiences in this actors. The information is not consolidated regard are the vertical expansions of the Old by the NDMA, so each partner relies exclu- Age Grant in response to COVID-19 and of sively on the information it has collected. the School Feeding Program in response to In the responses to COVID-19, the NDMA the 2015–16 drought. No social registry is in and some partners used the WFP’s benefi- place in Eswatini, nor is a database with his- ciary information and transfer management torical information on households affected platform (SCOPE) to register affected by previous disasters. households. The WFP also offered to use the platform to keep a registry of all support POSTDISASTER NEEDS delivered, consolidating the work of different A S S E S S M E N T S A N D D ATA organizations; this ultimately was not done. C O L L ECTI O N R E C O M M E N D AT I O N S F O R D ATA A N D Depending on the type and scale of the shock I N F O R M AT I O N S Y S T E M S I N E S WAT I N I and the response strategy, the use of preexist- ing data may need to be complemented by ex 1. Develop digital registries with associated MISs for the main post data collection exercises. Such exercises social protection programs in the country. Develop digital data would aim to improve initial targeting, which collection mechanisms and ensure registries are designed to relied on ex ante data that do not describe allow data sharing and exchanges. Use the data from the reg- ex post conditions, and/or to reach those left istries to inform program operations and policy planning. behind because they are not included in the 2. Develop an integrated beneficiary registry that consolidates registries used (Barca and Beazley 2019). information from different programs. The registry should be linked to (1) a mechanism for consolidating postdisaster needs Eswatini has a process in place for conduct- assessments conducted by various partners, with this informa- ing assessments of postdisaster household tion made available for response planning; and (2) a database food needs that relies on partners for data with historical information on households affected by shocks, collection. The postdisaster needs assess- which can be used for planning and preparedness. ment process involves local government 3. Strengthen capacity to produce early warning data that inform officials, community leaders, appropriate social protection responses and/or make use of international community committees and structures, and remotely sensed early warning information to trigger earlier action. 72 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA BUILDING BLOCK 4: S O C IAL P R OTECTI O N SPENDING FINANCE Social protection and social assistance Eswatini does not have financing mecha- spending in Eswatini is very low. Among nisms in place to support shock-responsive the SACU countries, Eswatini spends the social protection; of the four building blocks, smallest share of GDP on social assistance this is the least developed in Eswatini. (figure 3.4). As figure 3.5 shows, its largest Social assistance spending in the country share of spending on social assistance is for is low, as reflected by low coverage, limited the Old Age Grant; in 2017/18, this spend- types of programs available to households ing was equivalent to approximately half the (no cash transfer program for poor house- annual total spending on social assistance holds or child grants), low benefit amounts, program benefits (Raju and Younger 2021). and significant institutional capacity con- Spending has been second highest for OVC straints. Furthermore, the government does Education Grants, followed closely by the not have a disaster risk financing strategy in School Feeding Program. As noted earlier, place and no financial commitments to fund 37.9  spending on food aid reached E  mil- scale-up of social protection when the need lion in 2015/16 (9 percent of total spending arises. At a basic level, the probable costs of on social assistance program benefits that social protection responses to shocks—even year), due to the El Niño–induced drought those that are relatively recurrent and pre- emergency; in each of the other years, how- dictable—have not been quantified. ever, spending on the program accounted for less than 5 percent of total spending on Following are detailed findings and rat- social assistance program benefits (Raju ings justifications for this building block and Younger 2021). (table 3.6). DISASTER RISK FINANCING P O L I C Y A N D S T R AT E G I E S TA B L E 3 . 6   Scorecard for finance in Eswatini Eswatini does not have a contingency Indicator Score disaster risk financing instrument. It lacks Composite country rating ● contingency planning, risk pooling, and risk Foundational indicators transfer facilities (World Bank 2022b). While the country’s Public Financial Management Social protection spending ● Bill creates a legal framework for a disaster Disaster risk financing policies and strategies risk management fund, it has not yet been Adaptive indicators operationalized or capitalized. Quantification of the expected costs for social protection responses ● Shock responses rely largely on ex post Disaster risk financing instruments linked to social protection ● reallocations of funds and borrowing. Real- Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is locations involve significant bureaucratic between two stages of development. procedures, which delays responses. Fur- ther, reallocations can only be made without parliamentary approval if the amount is less than 5 percent of the budget for the given 3 :   E swatini 73 activity (World Bank 2022b). The govern- F I G U R E 3 . 4   SACU social assistance spending as a ment also made use of funding available percentage of GDP to it through concessional borrowing from 6 5.4% multilateral agencies during the COVID-19 pandemic. In July 2020, the International 5 Monetary Fund approved assistance up to Percent of GDP 4 3.4% a total of $110.4 million; in November 2020, 2.7% 2.9% the World Bank approved a COVID-19 recov- 3 ery loan of $40 million. The downgrading of 2 1.5% Eswatini’s credit rating in July 2020, how- 1 ever, will increase its cost of borrowing and is likely to reduce its ability to issue further 0 Botswana Eswatini Lesotho Namibia South Africa debt. (2019) (2015) (2017) (2018) (2016) Source: ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database. Q U A N T I F I C AT I O N O F P O S T - S H O C K ADAPTIVE S O C IAL FIGURE 3.5  Distribution of social assistance spending in P R OTECTI O N C OSTS Eswatini The government of Eswatini has not con- 100 ducted any quantification of future ASP needs based on risk assessments or pos- 80 Old Age Grants sible social protection responses. ASP is in OVE education grants its infancy in the country, and developing 60 School feeding Percent such quantification requires a clear under- Food aid standing of the response strategies, as well 40 Other grants and benefits as relatively rich information about previ- ous shocks, household vulnerabilities, and 20 other factors. The World Bank estimates the average annual cost of disaster responses 0 2014/15 2015/16 2016/17 2017/18 at $10.8 million (World Bank 2022b). The 2 Source: Raju and Younger 2021, based on Eswatini Ministry of Education and Training same study highlights that donor funding official budget books. percent of the estimated covered only 20  Note: Other grants and benefits include disability grants, military pensions, benefits for foster children, and transfers to Caritas Swaziland and the Baphalali Eswatini Red Cross Society. annual cost of disaster response in Eswatini from 1984 to 2019. Donor funding spiked 2 The simulated average annual cost of disaster response is $10.8 million. The estimated cost of after the 2015–16 drought; even then, a disasters between 2000 and 2020 was scaled funding gap of 80 percent remained. for population growth and used to fit a statis- tical distribution and conduct a Monte Carlo simulation to estimate the frequency and sever- DISASTER RISK FINANCING ity of shocks in Eswatini. Given that the analysis was partly based on the vulnerability assess- M EC HAN I S M S FO R ADAPTIVE ment data, the estimates were not specific to S O C IAL P R OTECTI O N a type of disaster but applicable to any disas- ter that left people in need of assistance in the Eswatini currently has no financial commit- country. The analysis was limited by scant data on disaster losses in Eswatini. ment to fund the scale-up of social protection 74 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA when need arises. The use of social protec- PRIORITY tion in response to shocks has been scarce and has relied mostly on ad hoc strategies. INTERVENTIONS F O R E S W AT I N I The analysis shows that ASP in Eswatini is R E C O M M E N D AT I O N S F O R still at a nascent stage. This is the result not F I N A N C E I N E S WAT I N I only of very limited investments in increas- As the ASP building block least devel- ing the flexibility and adaptability of the oped in Eswatini, finance requires more social protection system but, mostly, of the attention, and investments in its develop- underdeveloped foundations of the system— ment should be prioritized. The following including very limited spending on social are recommended to strengthen finance assistance, no cash transfer program to for ASP in Eswatini; for further rec- support poor households, and limited admin- ommendations, see the World Bank’s istrative capacity. disaster risk finance diagnostic for the country (World Bank 2022b): Table 3.7 summarizes the priority actions for both foundational social protection and 1. Increase spending on social assis- ASP. The key areas for strategic investment tance. Spending is lower than in include the following: neighboring countries and is insuf- ficient to address not only chronic ● Develop adequate policies and legislative poverty but also food insecurity frameworks for ASP. The new NSAP is an resulting from shocks. important step in this direction, as well as the DRM guidelines and framework. It is 2. Develop a disaster risk financing also necessary, however, to develop poli- strategy, including strategic priori- cies linked to the NSAP that are focused ties for disaster preparedness and on ASP and provide concrete indications response. Provisions and financial of how to use social protection capacity commitments for ASP should be for DRM actions. covered by the strategy. ● Develop digital databases for social 3. Consider developing a risk layering protection programs and an inte- approach that combines financing grated beneficiary registry. In addition instruments to match needs aris- to ensuring all social protection pro- ing from different disasters, and grams have digital registries in place engage with development part- with good-quality data, a registry should ners to secure funding for future be developed that integrates data from ASP responses to shocks, particu- most or all social protection programs to larly for the lean season. Securing enable better monitoring and to inform such funding may depend on invest- policies and operations, including ASP ing in an overarching ASP strategy actions. and enhanced institutional capacity for shock response (discussed in the ● Develop protocols and capacity to Building Block 1 recommendations). increase the effectiveness of the Emer- gency Food Aid program, with the support 3 :   E swatini 75 of partners. Develop contingency pro- ● Undertake a disaster risk financing diag- tocols that describe how the program nostic that assesses the actual and scales up based on different scenarios potential direct and wider costs of disas- and how the data from the vulnerability ters. Examine how an effective ASP and postdisaster assessments are used; response could reduce impacts and build define alternative ways to deliver cash resilience over time. and in-kind support to monitor the ● Engage with development partners program; and agree on roles and respon- and humanitarian agencies to develop sibilities with nongovernmental partners, a disaster risk financing strategy that on a methodology for conducting the incorporates external financing to sup- postdisaster assessments, on data shar- port ASP mobilization. Given the nascent ing, and so on. status of ASP in Eswatini, and of the ● Embed ASP into existing and new social finance building block, it is important protection programs in addition to the that the government integrate with exter- NDMA-led food aid program. The main nal agencies in the strategy development shock affecting the country, drought, process, and that partner agencies work is recurrent and strikes with a severity to earmark contingent funding to ensure that can generally be predicted well in the government has the liquidity available advance. Social protection programs— to quickly ensure people are supported from the School Feeding Program to the as needed during crises. Old Age Grant—can undertake ex ante ● Consider creating a cash transfer pro- planning and design tweaks (to eligibility gram based on the lessons from the OVC criteria, benefit packages, and so on) and Cash Transfer Pilot and international preparedness measures (along the deliv- experiences. Regular and frequent cash ery chain) to support households. Higher transfer values can contribute to reduc- benefit amounts, for example, could ing poverty while enhancing resilience. be provided to vulnerable households Such a program can be designed from during the lean season, or lump sum pay- the onset to include ASP provisions. ments could be provided in advance of it. 76 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 3 . 7   Summary of all recommendations by building block for Eswatini Building block Foundational social protection Adaptive social protection Strengthen the DPMO/Social Welfare ■ Operationalize the NSAP, track its progress, and review if as Department’s capacity to lead and necessary coordinate the social protection sector ■ Develop a strategy for ASP in line with the NSAP to establish a policy framework exclusively focused on ASP needs Institutional arrangements and ■ Ensure collaboration takes place between the NDMA and the partnerships Disaster Management Unit ■ Revise and improve the Inter-Cluster System to ensure better intersectoral coordination, and identify concrete actions to improve collaboration between the Social Welfare Department and the NDMA (and other entities involved in shock responses) ■ Continue rolling out digital payment ■ Develop contingency protocols for the Emergency Food Aid mechanisms for most beneficiaries program of cash transfer programs and the ■ Develop contingency protocols for scaling up programs— social pension beyond food aid—based on a national strategy for ASP ■ Consider creating a cash transfer Program design ■ Develop a methodology for annual response planning at the program based on lessons from the and delivery beginning of the lean season, consolidating all responses by OVC Cash Transfer Pilot and inter- the government and its partners national experiences ■ Consider designing and implementing a mechanism for regu- lar, seasonal cash support to households exposed to recurrent droughts (through the Emergency Food Aid program or a new program) ■ Develop digital registries with asso- ■ Strengthen the capacity to produce early warning data that ciated MISs for the main social inform ASP actions protection program in the country ■ Develop a mechanism for consolidating postdisaster needs ■ Develop an integrated beneficiary assessments conducted by various partners, and make this Data and informa- registry that consolidates informa- information available for response planning tion systems tion from different programs ■ Develop a database with historical information on house- holds affected by shocks that can be used for planning and preparedness ■ Once databases are created, establish mechanisms and proto- cols for data sharing between social protection and DRM Increase spending on social ■ Develop a disaster risk financing strategy that includes strate- assistance gic priorities for disaster preparedness and response; include provisions and financial commitments for ASP ■ Consider developing a risk layering approach that combines Finance financing instruments to match needs arising from different disasters ■ Engage with development partners to secure funding for future ASP responses to shocks, in particular for the lean season L E S OT H O A D A P T I V E S O C I A L P R O T E C T I O N A S S E S S M E N T C A S E S T U DY 4 OVERVIEW F I G U R E 4.1   Poverty headcount ratios in SACU countries OF RISK AND 30 29.2% 27.2% Poverty headcount ratio (%) HOUSEHOLD 25 VULNERABILITY 20 18.7% 14.5% Half of Lesotho’s population lives in pov- 15 13.4% erty and about a quarter (24.1 percent) 10 in extreme poverty, with significant geo- 5 graphical heterogeneity. Among Southern African Customs Union (SACU) countries, 0 Botswana Eswatini Lesotho Namibia South Africa only Eswatini has a higher poverty rate; rates (2015) (2016) (2017) (2015) (2014) in Botswana, Namibia, and South Africa are Source: DataBank: World Development Indicators. percentage points lower by around 10–15  Note: Poverty headcount ratio refers to the ratio of the population that is earning or (figure 4.1). Poverty levels in Lesotho are also consuming an amount below the poverty line. Poverty is measured according to the global poverty line of $1.90 per day in 2011 purchasing power parity (PPP) U.S. dollars. higher than in most middle-income coun- tries (figure 4.2). Although there are pockets of poverty in every district, poverty is more predominant in rural areas, where it goes up Lesotho is exposed to different types of cli- to 60 percent; it is notably high in the rural matic shocks, particularly drought, that mountain and Senqu River Valley regions, increase the risk of its people falling into which are drought-prone due to less pro- poverty and decrease their chances of ductive land and sparse rainfall (figure 4.3; exiting it. The hydrometeorological and eco- World Bank 2019c). Maseru District, on the logical hazards to which Lesotho is prone other hand, where the capital city of Maseru include droughts, episodes of heavy rain- is located, has the lowest poverty rates fall, snowstorms, hailstorms, strong winds, in the country. Districts on the border with localized floods, and early frost, as well as South Africa are also less poor than others, pest infestations and disease outbreaks. possibly because their residents can take Among these, droughts are the most fre- advantage of economic opportunities in the quent and severe (figure 4.4). According to neighboring country. the Emergency Events Database (EM-DAT), 79 80 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA FIGURE 4.2  Poverty in middle-income countries 60 Poverty headcount rate (%) 50 40 30 20 10 0 Zambia Kenya Côte d’Ivoire Lesotho Solomon Islands Cameroon South Africa Honduras Djibouti Botswana Micronesia, Fed. Sts. Bangladesh Namibia Ghana Philippines Myanmar Mauritania Bolivia Georgia Pakistan Nicaragua Vietnam El Salvador Bhutan Kyrgyz Republic Egypt, Arab Rep. Morocco West Bank and Gaza Sri Lanka Mongolia Tunisia Moldova Ukraine Source: World Bank 2019c. Note: Poverty headcount ratio refers to the ratio of the population that is earning or consuming an amount below the poverty line. Poverty is measured according to the global poverty line of $1.90 per day in 2011 PPP U.S. dollars. FIGURE 4.3  Rural and urban welfare distribution and poverty by district a. Welfare distribution (2017) b. Poverty headcount rate by district National poverty line Butha−Buthe 0.001 Leribe Berea Mokhotlong Rural population <40% Density 0.0005 Maseru Thaba−Tseka 40−50% 50−60% Mafeteng >60% Urban population Qacha’s Nek s Hoek Mohale’ 0 Quthing 0 1,000 2,000 3,000 4,000 5,000 Monthly expenditure per adult equivalent (M) Source: World Bank 2019c. 4 :   L esotho 81 droughts affected a total of 4,914,015 people F I G U R E 4 . 4   Population affected by climatic shocks in in Lesotho between 1968 and 2020, while Lesotho, 1968–2020 floods affected 185,000 and storms 9,351. 4% The frequency and intensity of droughts and other hazardous natural events have been increasing over the last two decades, with Flood the increase expected to continue (GOL Drought 2020f). Food insecurity is widespread in Lesotho and 96% highly correlated with instances of drought. According to the vulnerability assessment and analysis (VAA) conducted annually by the Lesotho Vulnerability Assessment Source: CRE n.d. Committee (LVAC), an average of almost 400,000 people have been in food insecurity every year since 2014 (figure 4.5).1 Droughts FIGURE 4.5  Food insecurity in Lesotho, 2014–20 result in poor harvests and livestock losses because agriculture is mostly rainfed, and 800 766,000 People in food insecurity (thousands) they affect the entire country, extending into 700 all 10 of its administrative districts, includ- 600 545,467 ing Maseru (Kardan, O’Brien, and Masasa 507,700 500 2017). In the two most recent El Niño– 390,440 400 379,399 induced drought occurrences (2015–16 and 2019–20), grain production decreased 300 244,172 224,665 by percent, 60  and the food-insecure 200 population reached an estimated 500,000– 100 65,679 700,000. As with poverty, droughts and food insecurity have stronger effects in 0 2014 2015 2016 2017 2018 2019 2020 Average rural areas. According to the VAA, 85 per- cent of Lesotho’s people in food insecurity Source: VAA 2014–20. in 2019 and 76 percent in 2020 lived in rural settings. at an alarming pace. According to estimates Existing vulnerabilities and risks are being of the World Bank Macro Poverty Outlook, compounded by the COVID-19 pandemic, the $1.90 poverty rate was expected to which is pushing people further into poverty have increased from 27.7 percent in 2019 to 30.5 percent in 2020, inclusive of mitigation strategies. 1 A household is in food insecurity if it has sur- vival deficits. According to the LVAC, survival deficits occur when income is insufficient to cover the survival food requirements, mea- sured as the total amount to cover 100 percent of minimum food energy needs (2,100 kcal) per person plus costs associated with food prepara- tion and consumption. 82 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA ASSESSMENT OF of Social Development (MOSD) in develop- ing adaptive social protection (ASP). Social ADAPTIVE S O C IAL protection is still not fully embedded in DRM P R OTECTI O N plans and strategies; specificity is lacking in terms of concrete roles and responsibilities over the DRM cycle. Translating the policies BUILDING BLOCK 1: and commitments articulated in these doc- INSTITUTIONAL uments into investments in preparedness ARRANGEMENTS AND and effective shock responses has been PA R T N E R S H I P S difficult, given capacity constraints within both the Disaster Management Agency The institutional arrangements for adaptive (DMA) and the MOSD. Although coordina- social protection in Lesotho are largely ade- tion structures related to DRM and ASP do quate on paper, as reflected in policies and exist, duplications and inefficiencies occur legislation, but capacity and coordination between those of the MOSD and the DMA; in support of their implementation remain while the DMA is tasked with coordination weak. In recent years, Lesotho has mounted of DRM-related matters, severe capacity an impressive effort to update its policies constraints have undermined its ability to and legislative frameworks for social protec- manage this function effectively. Strong col- tion and disaster risk management (DRM), laboration has taken place between the with strong donor support. Recent policies government and development partners, both and legislation in both areas include ref- in terms of developing ASP and responding erences to the role of social protection in to recent shocks. These partnerships have responding to shocks and building resilience, been ad hoc, however, rather than based on providing a clear mandate for the Ministry longer-term commitments. Following are detailed findings and rat- Scorecard for institutional arrangements and TA B L E 4 . 1   ings justifications for this building block partnerships in Lesotho (table 4.1). Indicator Score P O L I C Y A N D L E G I S L AT I V E Composite country rating ● FRAMEWORKS Foundational indicators The role of social protection in responding to Social protection policy and legislative frameworks ● shocks and promoting household resilience Disaster risk management policies and legislative frameworks ● in Lesotho is explicitly recognized in social Social protection institutional capacity and coordination protection legislation and policies. Lesotho’s National Social Protection Strategy (NSPS) Disaster risk management institutional capacity and coordination ● 2014/15–2018/19, for example, is framed Adaptive indicators around four key stages of the life cycle— Clear mandate for social protection response pregnancy and early childhood, school age Multisectoral coordination for social protection responses ● and youth, working age, and old age—and two cross-cutting life-cycle events—disability Partnerships with nongovernmental actors ● and chronic illness and (importantly, where Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. this assessment is concerned) shocks (GOL 4 :   L esotho 83 2014). The rationale behind this approach approach of the NSPS to a more complex is to combine programs targeting differ- mixture of functions that include preven- ent parts of the life cycle with needs-based tive, promotional, and transformative goals. programs to address the full range of vulner- Specific priorities center on harmonized tar- abilities, inclusive of covariate shocks. The geting, integration, and coordination tools NSPS, for example, emphasizes the role of for social assistance that tie social protec- social protection in reducing the sources of tion programs to other sustainable livelihood vulnerability to shocks ex ante: and DRM programs and on stronger coor- dination and institutional capacity (World The expansion of social protection and the Bank 2021b). resultant increase in resilience, especially of the poor and vulnerable, will signifi- Revised DRM legislative frameworks give cantly reduce the future cost of emergency response. Donors are regularly called social protection a key role in DRM, where upon to provide “humanitarian” assistance the Disaster Management Act and accom- to Lesotho, to deal with a problem that is panying frameworks did not. The Disaster essentially one of chronic poverty and vul- Management Act of 1997—and the accom- nerability. By reducing that poverty and panying National Disaster Management vulnerability in an effective pre-emptive Plan and Disaster Management Manual— manner, the inflated cost of rushed emer- gency reaction will be greatly diminished. still govern DMA functioning in Lesotho. The (GOL 2014) legal frameworks are limited to the fairly narrow mandate of coordinating emergency Ongoing revisions to the NSPS for 2021– response, given that this was the initial orga- 25 are anticipated to articulate further the nizational mandate for the DMA. The policy role of social protection in responding to shift from disaster response to disaster man- shocks.2 The design of the NSPS 2021–25 agement and resilience-building is, therefore, is based on the emerging consensus among not incorporated into these frameworks, governmental and nongovernmental part- nor is a role outlined for social protection in ners in the country that social protection disaster response (GOL 2020b). In 2020, should focus on expanding protective mea- a comprehensive set of legislative frame- sures while including preventive, promotive, works and policies for DRM was developed. transformative, and shock-responsive ones Although these are only starting to become (EPRI 2021). effective, they are likely to change the sector with a stronger focus on disaster risk reduc- Lesotho’s second National Strategic Devel- tion and management, along with explicit opment Plan 2018/19–2022/23 (NSDP II) provisions for ASP development. The DRM also outlines a linkage between social pro- Act calls for “adopting a shock responsive tection and disaster risk management. social protection mechanism in collaboration NSDP II (GOL 2018) aims to improve the effi- with relevant sectors, development partners, ciency of the social protection system and nongovernmental organizations, academia, represents a broadening from the life-cycle financing institutions and humanitarian agencies in the response to disaster induced emergencies” (GOL 2020a). When it goes into effect, the act will constitute the main 2 Source: Notes from the Strategic Dialogue Workshop conducted on February 17, 2021, to legislative framework for DRM, and its call update the NSPS. for adopting adaptive social policy provides 84 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA an important foundation for future invest- ● Ensuring alignment with external (i.e. parallel) responses to shocks to create ment in ASP development. ground for future expansion or to pre- serve and/or further strengthen an The 2020 National Early Warning–Early existing social protection system. For Action Plan describes the importance of example, this could be an alignment ASP—“integrating shock-responsive social of objectives, targeting method, and protection.” The first section of the plan transfer value or delivery mechanism. Despite a possibility that social protec- (excerpted here) recommends the following tion system may not be appropriate for a priority actions for ASP: humanitarian response, there is a strong rationale, in terms of longer term sus- ● Ensuring routine social protection tainability, to avoid as far as possible the programming is based on a solid under- setting up of completely parallel human- standing of the recurrent risks, shocks, itarian programmes and systems. (GOL and stressors; 2020f) ● An explicit focus on programming that contributes to enhancing the resilience Although the emphasis on ASP in DRM pol- of individuals, households and commu- icies and frameworks is promising, for the nities to future shocks; most part it is not elaborated upon or trans- ● An improved focus on vulnerability to shocks (beyond chronic poverty or cate- lated into specific actions. Despite the gorical, life-cycle criteria) e.g. expanding inclusion of high-level ASP recommenda- routine coverage in areas frequently tions in the respective first sections of the affected by shocks; incorporating vul- National Early Warning–Early Action Plan nerability criteria into routine targeting, and Multi-Hazard Early Warning System etc.; Manual (GOL 2020e, 2020f), neither sub- ● Ensuring continuity of service deliv- sequently presents linkages between the ery for routine programmes in actions they establish regarding early warn- shock-affected contexts, at the moment when recipients need support the most; ing systems and social protection. Moreover, ● Enhancing preparedness to enable the and surprisingly, the National Early Warn- flexing and/or scaling up of existing pro- ing Strategic Action Plan describes the grammes, or activating new emergency roles of more than 15 ministries and agen- programmes that build on existing sys- cies but not of the MOSD. While the National tems, to accommodate new populations Multi-Hazard Contingency Plan does men- and needs as a result of a shock. Options tion the possibility of responding to shocks include: i) Vertical expansion of an exist- ing programme or set of programmes. by scaling up social protection programs, New components may also be added; it does not include any concrete plans for ii) Horizontal expansion of an existing doing so. The standard operating proce- programme, or set of programmes, to dures for drought responses, for example, temporarily include new beneficiaries indicate only the following: from affected communities; iii) The acti- vation of a new emergency programme Call for a stakeholder meeting to brief that rides on existing systems (e.g. data, partners on projected situation and ask capacity, delivery mechanism) where sectors to discuss areas of possible inter- possible. In Lesotho, this could be led by vention, particularly where existing MOSD working with Development Part- resilience-building and social safety net ners and humanitarian organizations; programmes could be scaled up temporarily or have the targeting mechanisms account 4 :   L esotho 85 for the expected spikes, in anticipation of (ECHO 2021; see box 4.1). The development LVAC results. Stakeholders, particularly of policies, legislative frameworks, and plans sectors should already seek ways of identi- providing for ASP across the social protec- fying areas of complementarity to enhance tion and DRM sectors shows a clear interest such responses. (GOL 2020c) in this area. A recent review, however, has identified a need for deeper understanding C O O R D I N AT I O N A N D and shared visions between the social pro- C A PA C I T Y tection and DRM sectors as well as clarity An obstacle to effective coordination is the on roles, responsibilities, and coordination existence of parallel coordination struc- mechanisms (EPRI 2020). tures. This lack of clarity in leadership of the coordination function related to ASP has The severe resource constraints on the DMA been recognized as a challenge by the gov- limit its ability to fulfill its DRM-related man- ernment. In 2020, the MOSD and the DMA dates and to coordinate effectively with signed a memorandum of understanding the many actors involved, including the to harmonize the coordination structures MOSD. The DMA’s constraints are widely B O X 4 . 1   Integrated policy framework for aligning the government of Lesotho’s coordination structures The memorandum of understanding between the will aim to improve coordination among institutions MOSD and the DMA to harmonize their coordination for ASP. Members will include the MOSD, the DMA, structures builds on a proposal for an Integrated Policy other agencies, development partners, and civil Framework (GOL 2020c), developed with support from society partners. the United Nations Children’s Fund (UNICEF) as part ■ Creation of a focal point at the MOSD for disaster of a project funded by the European Commission’s preparedness and response. Establishing an MOSD Directorate-General for European Civil Protection and focal point whose responsibility is to support the Humanitarian Aid Operations (ECHO). The framework shock-responsive social protection system within proposes the following: the ministry and strengthen coordination between the ministry and the DMA will ensure there are ■ Rationalization of horizontal coordination structures clear lines of accountability within the MOSD for at the policy level. This will facilitate the process of shock-responsive social protection. alignment between the MOSD and the DMA, par- ■ Continued investment in the DMA’s capacity. ticularly given the lack of enthusiasm conveyed by Although the DMA has the legal mandate to coor- the latter for the structures proposed by the former. dinate across sectors, integration between it and Essentially, the MOSD Subcommittee on Disas- the MOSD can be undermined by the long-standing ters, Shocks, and Crises would be absorbed by challenges to the DMA’s capacity to convene and the DMA, becoming a dedicated working group on coordinate effectively. Ultimately, effective coor- shock-responsive social protection. dination between the two depends on continued ■ Creation of a shock-responsive social protection investment in the DMA’s capacity. community of practice. The community of practice Source: GOL 2020c. 86 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA recognized—even by its own assessment, Council so DRM can be mainstreamed into which observes that the authority “cur- the national development process. The rently lacks the capacity, resource streams secretariat of the council will be the DMA. and supporting legislative power to effec- This proposed organizational structure is tively undertake the coordination mandate” expected to enable the DMA to carry out (GOL 2020g). These capacity constraints the necessary activities at all phases of the stem from two issues: the authority is inade- disaster management cycle—predisaster, quately resourced, in terms of both staffing disaster response, and postdisaster (GOL and budget; and it does not have the power 2020a, 2020g). and autonomy to lead the sector as envi- sioned (GOL 2020b). The DMA Act of 1997 As with the DRM sector, coordination of provides for six permanent sectoral work- social protection programming across imple- ing groups to monitor disasters and ensure menting agencies remains a challenge. The preparedness. As a former constituent of MOSD is charged with policy design and the Ministry of Health, the MOSD is repre- coordination in the social protection sector, sented in the health and nutrition working which has been reported as one of the main group. Despite its legally mandated author- challenges to effective social protection pro- ity, the DMA finds it challenging to convene gramming (EPRI 2020). To carry out this these meetings because its de facto power is mandate, the MOSD relies on a coordination limited (Kardan, O’Brien, and Masasa 2017; structure that includes six technical sub- Ulrichs and Mphale 2016). While the working committees, corresponding to the life stages groups are meant to meet quarterly, reports underlying the NSPS. These subcommittees indicate meetings are seldom conducted in rarely meet (GOL 2020f). practice (GOL 2020b). In addition, working group membership has, over time, devolved The MOSD is relatively well staffed, but from comprising senior representatives of has weaknesses in fundamental capacity, the various government ministries, as pre- especially in information technology, data scribed in the act, to junior delegations with management, and monitoring and evaluation. no legal mandate to make policy decisions. The MOSD information technology depart- This diminution has undermined the role of ment relies on externally funded short-term the working groups (GOL 2020g). consultants to perform some basic tasks and is greatly understaffed (EPRI 2020). This is The DRM Act of 2020, yet to be enacted by notable because this department is respon- Parliament, establishes a new coordination sible for the National Information System for structure to be led by the DMA. In the act, Social Assistance (NISSA), which is intended the DMA still serves “as the central plan- to become the backbone of social assis- ning, coordinating and monitoring institution tance registry information (see discussion in respect of all disaster risk reduction and under Building Block 3) and for implemen- management issues and postdisaster recov- tation of shock-responsive social protection ery matters in Lesotho” (GOL 2020a). The in the country. To date, the development and new structure increases the number of sec- management of the social registry has been toral working groups to 10 and specifies that substantially funded by donors. they be chaired by senior representatives. The act also provides for the formation of Subnational structures also suffer from a high-level National Disaster Management capacity constraints that limit coordination 4 :   L esotho 87 vital to ASP implementation in Lesotho. The Southern African Development Communi- DMA coordinates its district-level activities ty’s Regional Vulnerability Assessment and through district disaster management teams; Analysis Programme (SADC RVAA). The these teams are chaired by district adminis- LVAC’s role and the assessments it produces trators who report to the Ministry of Local are described under Building Block 3. Government and Chieftainship Affairs and are not accountable to the DMA. The teams As a single source of information for uni- include district-level counterparts of various fied vulnerability assessments, the LVAC national ministries, including the MOSD, as has enhanced ASP coordination. VAA data well as representatives of civil society (GOL are at the center of DRM contingency plans, 2020b). However, most sectoral ministries such as the National Early Warning Strategic do not have dedicated budgets for disaster Action Plan (GOL 2020g) and the Lesotho management at the district level. And, since Multi-Hazard Contingency Plan 2020–2023 district disaster management teams are not (GOL 2020c). They have also been used in accountable to the DMA, DMA leadership recent years by the government and partners and oversight rely on setting up an appropri- to implement social protection responses. ate incentive framework and on the DMA’s power—which is not as strong as required. Partnerships with nongovernmental actors The MOSD recently established district social for social protection preparedness and protection committees to improve vertical response initiatives are relatively strong, coordination with subnational actors. The although not based on longer-term commit- existence of parallel district-level coordina- ments. In addition to the support provided tion structures, all comprising essentially the by the LVAC as a coordination and partner- same sectoral representatives, is a burden ship platform, key initiatives to improve ASP on local structures and has created tension policies and provisions have been aided between the MOSD and the DMA. technically and financially by nongovern- mental actors through, for example, the development of policies and legislation; M U LT I S E C T O R A L workshops promoting intersectoral policy C O O R D I N AT I O N A N D dialogue, technical assistance, and assess- PA R T N E R S H I P S ments and evaluations; and implementation The LVAC is an important multisectoral coor- of pilots (for instance, a mechanism for dination platform for ASP. Established in electronic payment). Responses to recent 2002, it is a government-led multidisciplinary shocks have also benefited from partner- committee within the Office of the Prime ships with organizations—such as the World Minister whose members include representa- Bank, the United Nations Children’s Fund tives of government ministries (including the (UNICEF), and the World Food Programme DMA and the MOSD), United Nations agen- (WFP)—playing key roles in ASP responses cies, nongovernmental organizations, and to droughts and the COVID-19 pandemic; the private sector. The LVAC carries out vul- this is discussed under Building Block 2. nerability analysis and needs assessments to provide timely analysis for planning emer- gency interventions, as well as medium- to long-term programming. The LVAC is linked both technically and financially to the wider 88 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA R E C O M M E N D AT I O N S F O R I N S T I T U T I O N A L BUILDING BLOCK 2: A R R A N G E M E N T S A N D PA R T N E R S H I P S I N PROGRAM DESIGN AND LESOTHO D E L IVE RY The main constraints on effective institutional arrangements in Lesotho has recently utilized its Child Grants Lesotho are related to the capacity of major actors such as the Program in response to shocks, but sev- DMA and the MOSD. There has been no shortage of intersectoral eral design and delivery constraints have forums with the purpose of agreeing on ASP matters or of leg- limited the timeliness and effectiveness of islative frameworks, policies, and plans. These initiatives have such interventions. The offer of social assis- not been as effective as expected, however, precisely because of tance programs in Lesotho is substantial, capacity constraints, especially in the DRM sector. For this reason, and coverage is overall high, which pro- the main recommendations for this building block relate to capac- vides a good platform for their use in shock ity strengthening: response. Poverty-targeted programs reach only a relatively small share of the groups 1. The DMA’s capacity to lead and coordinate the sector needs they are meant to cover, however, and to be strengthened. In recent years, capacity constraints have benefit amounts are overall low in relation- been dealt with by creating policies and frameworks and pro- ship to program objectives of consumption moting dialogue, but this has been only partially successful. smoothing and poverty reduction. Such The DMA’s constraints are the result of limited power and, shortcomings limit the ability of the social therefore, inadequate resources, despite the establishment of protection system to reach poor households relevant legislation. that are continually vulnerable to recur- 2. The MOSD’s ability to coordinate the social protection sector rent shocks like drought—and, when it does and deliver assistance in normal times and during crises reach them, the limited benefit amounts can requires additional, foundational investments in underlying minimize the impact on household resilience capacity—in particular, in the capacity of the information tech- building. For example, the Child Grants Pro- nology department, which manages NISSA. gram (CGP) has been mobilized to respond 3. An important next step is the implementation and evaluation to drought and COVID-19, expanding cov- of the coordination framework proposed as part of the mem- erage of poor households with children and orandum of understanding between the MOSD and the DMA. increasing benefit amounts. The responses, The capacity of both entities to engage with district structures however, have been mostly ad hoc and not and to support and monitor program delivery at the local level implemented according to predetermined should also be strengthened. contingency plans or protocols. The flexi- bility of the CGP to adjust its parameters in 4. Although impressive progress has been made with regard to the response to shocks has also been limited by ASP legislative framework and policies, ASP and the specific its mostly manual mechanisms for service roles and responsibilities of the MOSD need to be integrated delivery and the payment of cash grants. more concretely into DRM policies and plans. The government of Lesotho should go beyond high-level statements and estab- Following are detailed findings and rat- lish how social protection and DRM should work together in both ings justifications for this building block preparedness and responses to shocks. (table 4.2). Lesotho offers a relatively substantial number of social assistance programs that address needs arising along the life cycle, in line with the approach outlined in the NSPS. 4 :   L esotho 89 Table 4.3 summarizes these programs. TA B L E 4 . 2   Scorecard for programs and delivery systems in Orphans and vulnerable children (OVC) Lesotho are targeted through the CGP, the Public Indicator Score Assistance program, and the OVC Bursary program; they also receive support from the Composite country rating universal School Feeding Program. Older Foundational indicators people receive the Old Age Pension, and Benefit adequacy ● working-age adults can be engaged in the public works program Fato Fato. Eligibility criteria and targeting ● Coverage of the main social assistance program ● Three social assistance programs imple- Financial inclusion and national ID ● mented by the MOSD are the primary Social protection payment systems focus of this assessment. The CGP is a government-funded cash transfer pro- Social protection delivery systems ● gram targeted to poor households with Adaptive indicators children under the age of 18. Beneficia- Experience of vertical expansion ● ries are identified through a combination of community-based targeting and proxy Experience of horizontal expansion ● means testing. The Public Assistance pro- Protocols and contingency plans ● gram supports the destitute. It provides Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is permanent and temporary assistance to between two stages of development. OVC, the severely disabled, the severely ill, and the elderly, among others. The OVC Bursary supports access to education, espe- invasive species, construction of infiltration cially secondary education, for OVC. ditches, and terracing to reduce erosion; and water harvesting, including the installa- Other social assistance programs of note that tion of roof tanks, water storage, and dams are not implemented by the MOSD include (World Bank 2019d). social pension, school feeding, and public works programs. Some large programs, C OVE RAG E, B E N E F IT such as the National School Feeding pro- A D E Q U A C Y, A N D TA R G E T I N G gram and the Old Age Pension program, are managed by the Ministry of Education and The overall coverage of social protection in Training and the Ministry of Finance, respec- Lesotho is high, offering a strong founda- tively, although a transition from the Ministry tion for building resilience and responding of Finance to the MOSD is planned. Finally, to shocks. Overall coverage is comparable Fato Fato is implemented by the Ministry to that in neighboring countries (figure 4.6). of Forestry, Range, and Soil Conservation. Almost 72 percent of the population lives in The program provides up to 20 days of sea- households with at least one member who sonal employment to low-skilled workers to benefits from a social protection program. engage in a range of land management and The two largest social assistance programs soil conservation activities, such as planting are those that provide universal cover- trees for fuel and wood; planting fruit trees; age: the National School Feeding program land rehabilitation, including rehabilitation and the Old Age Pension. The coverage of of rangeland and grass cover, removal of poverty-targeted programs—usually those 90 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 4 . 3   Main social assistance programs in Lesotho Implementing Program Description agency Targeting Benefit value Coverage Provides permanent and MOSD Self-targeting, iden- Average monthly benefit: 12,000 Public temporary assistance to tification by social M 133 ($11); quarterly ben- beneficiaries Assistance OVC, the severely disabled, workers, selection efit between M 250 and program the severely ill, and the by district managers M 500 elderly OVC under age 18 enrolled MOSD Community tar- Bursary varies by grade 23,000 in secondary school who geting and proxy and type of school but beneficiaries have lost one or both par- means testing generally includes tuition OVC ents; have a sick, disabled, based on NISSA fees, examination fees, Bursary or incarcerated parent; or and application registration cost, sta- are considered needy review tionery, books, special subject fees (e.g., science fees, boarding fees) Public works program Ministry of For- Self-targeting, not M 960 ($70) per month for 80,500 Public employing able-bodied indi- estry, Range, restricted to the a maximum of one month beneficiaries works viduals living in rural areas and Soil poor; first-come, per year and on a rota- (Fato Fato) for conservation-related Conversation first-served tional basis activities Provides free meals to all Ministry of Universal One or two free meals 400,000 National children attending primary Education and daily at school children School schools that offer free edu- Training Feeding cation and some preschools Pension for any person over Ministry of Universal Monthly cash grant of 83,751 Old Age the age of 70 not receiving a Finance M 800 ($50) persons Pension civil service pension (2017/18) Source: Based on Kardan, O’Brien, and Masasa 2017; World Bank 2019d, 2021b; administrative data. with mandates closest to ASP goals— the program can be a good platform for build- remains low. Targeted programs reach, ing resilience and responding to shocks. overall, a relatively small share of the groups Although the program’s operations in the 10 they are meant to cover. The World Bank districts of Lesotho do not include every vil- estimates that, based on 2019 figures, the lage, its coverage largely extends to areas Public Assistance program reaches only where food deficits are larger (figure 4.7). 6 percent of households in extreme poverty, and the CGP reaches 16 percent of children Benefit amounts of these social safety net in poverty (up from 11 percent in 2017).3 programs are generally low, which limits their effects on building household resil- The CGP’s geographical coverage is highly ience to shocks ex ante. Benefit amounts correlated with food insecurity, which means in Lesotho are similar to those in the other countries studied, although their adequacy is lower than in South Africa and higher than in Eswatini (figure 4.8). Interestingly, 3 Source: ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database. when calculated for the poorest quintile (the 4 :   L esotho 91 FIGURE 4.6  Social protection coverage in SACU countries: direct and indirect beneficiaries 90 Botswana Eswatini Lesotho Namibia South Africa (2015) (2016) (2017) (2015) (2014) 80 70 % of population receiving benefits 60 50 40 30 20 10 0 All social protection All social assistance Cash transfers Social pensions School feeding and labor Source: ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database. total transfer amount received by all benefi- F I G U R E 4 . 7   Correlation between CGP coverage and food ciaries in the first quintile as a share of the deficits in Lesotho, by district total welfare of beneficiaries in that quin- 60,000 Number of households with food deficits tile), Lesotho’s benefit adequacy improves compared to Botswana’s and Eswatini’s. 50,000 R² = 0.6645 Lesotho’s most generous program is the Old Age Pension; its spending per beneficiary 40,000 is slightly above the upper poverty line. The 30,000 per beneficiary spending of the other pro- grams is well below the food poverty line; in 20,000 the case of the Public Assistance program, for example, the amount equates to 71 per- 10,000 cent of the poverty line. The value of CGP benefits, which depends on household size, 0 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 is diluted when calculated on a per person Number of CGP households (2019) basis—it is the equivalent of 13 percent of Source: Based on LVAC data from 2014–18 and MOSD data from 2019. average monthly household consumption Note: CGP coverage has increased since 2019. Number of households with food deficits (World Bank 2021b). Because the method by is average for 2014–18. which benefit values are determined varies by program, the real value of benefits is in some cases (for example, the CGP) eroded, since transfers are not pegged to inflation. 92 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA F I G U R E 4 . 8   Benefit adequacy of all social assistance eligible for the CGP (GOL 2020d; World programs in SACU countries Bank 2019d). This procedure is well docu- mented and is designed to target the chronic 120 114.0 Overall population Poorest quintile poor—but not necessarily those among the 100 poor who are most vulnerable to shocks nor those who are vulnerable to poverty due Benefit adequacy (%) 80 to shocks. The Public Assistance program relies on self-targeting and identification 60 by social workers or district managers; this 42.4 process is not well documented. Both the 40 32.2 29.4 National School Feeding program and the 21.1 20.7 20 Old Age Pension are universal, aiming to 13.1 10.8 12.0 6.5 reach all children in primary schools and 0 people over the age of 70 not receiving civil Botswana Eswatini Lesotho Namibia South Africa (2015) (2016) (2017) (2015) (2014) service pensions, respectively. The public works program is self-targeted. Source: ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database. Note: Benefit adequacy = total transfer amount received by all beneficiaries in a population group as a share of the total welfare of beneficiaries in that group. S O C IAL P R OTECTI O N D E L I V E RY SYST E M S Consequently, benefit amounts that are too Cash transfer programs rely mostly on low to close the poverty gap and ensure food manual payment mechanisms (cash in tran- security fall even shorter for building resil- sit), which can limit the speed with which ience among poor households. responses can be mobilized. Because they lack digital mechanisms, most programs are Social assistance programs in Lesotho use without proper control and reconciliation a variety of targeting approaches, although systems (World Bank 2021b). Transfers are none is designed to identify people who typically delivered through district offices, are vulnerable to recurrent shocks such as security company services, and even by drought. Among the poverty-targeted pro- helicopter to remote regions (GOL 2020b; grams, the CGP has the strongest and most World Bank 2019d). This is the result of a transparent targeting mechanisms, relying on combination of factors, including the limited a combination of community-based targeting penetration of banking and mobile payments and proxy means testing. Community-based (see box 4.2); it is also partially due to Leso- targeting is conducted through an inten- tho’s topography, which includes remote and sive mobilization process involving hard-to-reach areas. The CGP has piloted community leaders, district authorities, and mobile payments in several districts, but the village chiefs. Households are classified into mechanism has not been scaled up yet. Both four income categories—ultra-poor, poor, CGP and Public Assistance transfers provide moderate, and well-off—corresponding to cash payments on a quarterly basis without NISSA categories 1–4. Proxy means testing substantial delays—at least in the case of is used to validate the information gleaned the former; delivering cash more frequently through community-based targeting, with would be a challenge for these programs. those households classified as ultra-poor or poor through both methods deemed 4 :   L esotho 93 B OX 4.2   Access to financial services and national ID cards is still limited and inequitable in Lesotho Compared to neighboring countries, Lesotho has Although national ID coverage in Lesotho is fairly high the lowest percentage of people with bank accounts (70 percent), it is substantially lower than in Botswana, and of people who have used digital payments (see Namibia, and South Africa, which are all above 90 per- figure B4.2.1; while percentages for Eswatini are nomi- cent. It is likely that the poor and most vulnerable, nally lower, they are probably underestimated because including those living in remote areas, are within the data for that country are from 2011, while those for 30 percent that do not have national IDs. Efforts to other the countries are from 2017). In addition, the address this—in collaboration with the Ministry of World Bank’s Digital Adoption Index—which provides Home Affairs—are reported as ongoing for the Old Age measures of countries’ digital adoption across three Pension, the CGP, and the OVC Bursary programs to dimensions of the economy (people, government, and identify beneficiaries who have national IDs and roll business) and gross domestic product per capita— out IDs to all those who do not yet have them. shows that Lesotho and Eswatini have the lowest levels in the subregion. Lesotho ranks 291 out of 359 coun- tries in the Digital Adoption Index global ranking. F I G U R E B 4.2 .1   Financial inclusion and national ID coverage Botswana Eswatini Lesotho Namibia South Africa 100 80 60 Percent 40 20 0 Has account Has national ID Made/received digital payments Digital Adoption Index (% age 15+) card (% age 15+) in past year (% age 15+) Source: Global Findex Database, World Bank. Note: Eswatini data for account access are from 2011 and probably underestimate current coverage. Other delivery processes related to beneficiaries, for instance, apply at the outreach, intake, registration, needs assess- district office, where they complete an appli- ment, and enrollment largely rely on manual cation form that is accompanied by a referral mechanisms, which limits their flexibility and letter from the village chief. Old Age Pension scalability during crises. This is the case for applicants submit their applications through most social assistance programs in Leso- local government offices of the Ministry of tho, except for the CGP and, more recently, Finance (World Bank 2021b). These mecha- the OVC Bursary. Most Public Assistance nisms are cumbersome and not automated, 94 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA which greatly limits their flexibility to scale and criteria for scale-up of the program in up and address sudden increases in demand. response to droughts. The framework uses VAA data to forecast the number of house- Grievance redress mechanisms, often espe- holds likely to be affected, and the scale-up cially important in a post-shock environment, plan relies largely on NISSA data. Although remain relatively weak. Grievances and case approved by the government, this framework management mechanisms could play a fun- has not yet become effective and is not inte- damental role both in social protection grated into any other plans or strategies. and ASP by ensuring households receive the support required and raising over- Wider DMA-led contingency plans lack spec- all accountability. These mechanisms are, ificity on social protection interventions however, generally weak in Lesotho. While and are usually developed in relation to an grievances are usually received by auxiliary imminent or recent shock. A number of con- social workers, community councils, and dis- tingency plans have been designed recently trict offices, records are not always kept, the for early warning and early action, includ- information does not always reach central ing plans for responding to different hazards offices, and the processes for dealing with with social protection (see Building Block 1). the grievances are unclear. Similarly, with These plans, however, do not establish how case management, support for households social protection programs and systems like in navigating the social assistance sector NISSA should be used to inform and deliver (and beyond) and assurance they will receive assistance. Contingency plans with concrete the benefits to which they are entitled are response actions are elaborated by the DMA lacking. The absence of management infor- and partners on an ad hoc basis—that is, mation systems (MISs) from some programs when a shock has already hit or is about to. and of unique ID numbers are among the In 2019, for example, the Drought Response challenges for setting up a case manage- and Resilience Plan targeted 508,125 people ment mechanism. Even in the cases of the identified as being at risk of food insecurity CGP and the OVC Bursary—both of which for the 2019/20 consumption year based on have operating MISs, rely on NISSA data, VAA data, and the scale-up of the CGP was and belong to the same ministry—it is not included as a response (GOL 2020c). possible to link program beneficiaries across the MISs (World Bank 2021b). V E R T I C A L A N D H O R I Z O N TA L E X PA N S I O N I N R E S P O N S E T O P R OTO C O LS AN D SHOCKS C O N T I N G E N CY P L A N N I N G Lesotho has responded to recent droughts A contingency plan (scalability framework) by temporarily expanding the CGP vertically for use of the CGP in response to shocks has (by increasing benefit amounts) and horizon- been developed but is not yet effective. As tally (by increasing coverage). Most such part of a program funded by the European responses have relied on the financial and Commission’s Directorate-General for Euro- technical support of development partners pean Civil Protection and Humanitarian Aid and on NISSA data. Based on these expe- Operations (ECHO), UNICEF supported the riences, the consensus among stakeholders MOSD in developing a scalability framework is that the CGP is the social assistance pro- for the CGP that establishes parameters gram best placed to provide cash support 4 :   L esotho 95 in response to shocks because, despite its Social protection responses to the COVID-19 relatively low coverage, it has the stron- pandemic have resulted in more horizontal gest delivery mechanisms. The expansion expansions of the CGP and of other pro- of the CGP can be, and has been, comple- grams. Partners such as the World Bank, the mented by the expansion of other programs WFP, and UNICEF have supported both ver- such as the Public Assistance program and tical and horizontal expansions of the CGP. the public works program. The World Bank The Public Assistance program has also (2019d) has simulated the scale-up of dif- been expanded to complement responses ferent social protection programs in Lesotho from the CGP and provide support to and has concluded that the CGP would be households without children. The precise the most cost-effective. parameters of these responses—beneficia- ries, payment amounts, and so on—were not Following are among the recent responses available at the time of this writing. to droughts in Lesotho. Overall, the temporarily increased amounts ● CGP vertical expansion (cash top-ups of cash transfers during vertical expansions to existing beneficiaries) in response to have been low relative to estimated food defi- the El Niño–induced drought (2016): Two cits of affected households. The adequacy quarterly top-ups of M 500 ($37) each of responses is always difficult to assess were given to 23,000 CGP households because it depends on the policy objectives, (or 115,000 people) included in NISSA, as well as on other complementary support supported by UNICEF in 33 community households may receive. Since food insecu- councils (Kardan, O’Brien, and Masasa rity in Lesotho combines chronic and acute 2017). A third top-up, financed by the needs, the amounts required to cover food World Bank, covered 27,000 households. gaps are substantially larger than both reg- ● CGP vertical expansion in response to ular benefit values and recent responses to the 2019–20 drought: Funding came droughts. According to LVAC calculations, from two sources. The World Bank sup- the average survival deficit in 2019/20 was ported two rounds of CGP top-ups to M 756 ($51) per month, calculated over a 26,453 households in selected dis- period of six months—that is, each house- tricts. Each household received a total hold in survival deficit should have received of M 3,000 ($204). The WFP provided a total transfer of M 4,536 ($306) to cover cash and vouchers to 16,500 households the food gap. On average, the CGP bene- in four districts, with plans to scale up to fit was M 133 ($11) per month, amounting 43,500 households (Tlakane 2020). to M 678 ($66) over six months. The value of the expansion of the CGP in response to ● CGP horizontal expansion in response to the 2019–20 drought was M 3,000 ($204) the 2019–20 drought: With support from for vertical expansion and between M 1,500 the World Bank, one-off transfers were and M 2,250 ($102 and $153) for horizon- made to 10,669 non-CGP beneficiaries tal expansion. In other words, both regular who were already in NISSA. Each house- and temporary benefits have fallen short, hold received a transfer of between on average, in meeting the food deficit, M 1,500 and M 2,250 ($102 and $153) although the policy objective may not have (Tlakane 2020). been to cover the full deficit. 96 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Reaching nonbeneficiaries through horizon- UNICEF’s horizontal expansion in response tal expansions in recent years was enabled to COVID-19, the MOSD committed to bring- through the use of VAA and NISSA data. VAA ing those households into the program after data indicated the number of people in need the temporary emergency support ended in each district and the size of the food defi- (ECHO 2021). cit; the NISSA registry and its poverty ranking were used by partners and the government Vertical and horizontal expansions have to identify and prioritize nonbeneficiaries. often been delayed. One reason for such Targeting through NISSA was validated by delays is limitations in terms of coordination, local actors. The use of NISSA data pre- leadership, and institutional capacity, as sented some difficulties, however; these are described under Building Block 1. Second, described under Building Block 3. a lack of protocols and of pre-agreed plans has slowed the decision process. A lack of Temporary horizontal expansions in ex ante financial commitments has been a response to shocks have led to increases third constraining factor. Finally, a reliance in the longer-term coverage of regular pro- on cash-in-transit mechanisms for delivering grams such as the CGP. Some horizontal support has proved ineffective and imposed expansions have entailed reaching house- significant limitations on horizontal expan- holds that were in NISSA’s registry and sions, creating a key operational barrier, were eligible for the CGP but had not yet particularly in places where the CGP is not been enrolled in the program. In the case of functioning. 4 :   L esotho 97 R E C O M M E N D AT I O N S F O R P R O G R A M D E S I G N A N D D E L IVE RY I N L E S OTH O The assessment of programs and their delivery systems for ASP places Lesotho between the nascent and emerging stages, with opportunities arising from recent experiences in scaling up social protection programs in response to droughts and the pandemic. The fol- lowing are recommended to strengthen this building block in Lesotho: 1. The coverage of poverty-targeted programs needs to be expanded to reach a larger share of the intended population, and benefit amounts need to be revised in line with program objectives related to consumption smoothing and poverty reduction. 2. A key means of strengthening service delivery is to make it more cost-effective, trans- parent, and responsive by rolling out digital payment mechanisms. This has been a policy concern in Lesotho for some time, and a few pilot programs have been imple- mented. It is important to build on the lessons from these pilots and start using these mechanisms, most likely in combination with cash-in-transit methods in locations where the use of electronic mechanisms is not feasible or cost-effective. Any new dig- ital payment mechanisms and associated contracts with payment service providers should be designed to ensure they are flexible and scalable; standby agreements with service providers can be established, for example. 3. For programs other than the CGP, establishing transparent criteria and processes for targeting and service delivery and documenting these processes in operations manu- als are crucial. 4. Social protection should be an entry point for expanding the coverage of the national ID by ensuring all beneficiaries have them. 5. Programs should develop contingency protocols for business continuity during crises to maintain service provision to regular beneficiaries throughout a shock or crisis. Addition- ally, programs that are intended to be scaled up vertically and horizontally in response to a shock should develop protocols to inform those kinds of expansions. These pro- tocols would need to be accompanied by adequate training for staff and resources for implementation. 6. The Public Assistance program should be adapted to provide support after a shock to households not included in the CGP, as proposed in the 2020 workshop for updating the NSPS (EPRI 2020). 7. Since droughts are recurrent and most of the people affected are in chronic poverty, transferring cash through the CGP and the Public Assistance program at the beginning of the lean season could be a more effective way of responding to them than waiting until the drought has become a crisis. Consideration should be given to designing and implementing a mechanism for regular cash support to households exposed to recur- rent droughts. 98 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA BUILDING BLOCK 3: Following are detailed findings and rat- ings justifications for this building block D ATA A N D (table 4.4). I N F O R M AT I O N S Y S T E M S The existing information systems in Leso- S O C IAL P R OTECTI O N tho that are required for effective ASP have I N F O R M AT I O N S Y S T E M S both strong points and shortcomings, with the national social registry (NISSA) providing The capacity to collect, manage, and share a critical foundation for shock response that data across social assistance programs can be further strengthened. While NISSA and beyond the social protection sector can contains data on virtually all the households be fundamental for responding to shocks. in the country and has been used to inform Nascent systems are typically largely paper social protection responses to shocks, the based, with few security protocols and social registry’s interoperability and inte- controls to ensure data quality, no interoper- gration are very limited. Currently, no ability, and no social registry in place. At the mechanism is in place for updating the data other end of the spectrum are established frequently and at scale. Similarly, while vul- information systems. With their digitized nerability data produced by the LVAC have and electronic registries and strong security been essential to informing MOSD- and mechanisms, they inspire confidence among DMA-led response interventions, the capac- users and the public regarding the quality of ity to conduct postdisaster household needs the data they contain. They also offer a high assessments and share the resulting data level of integration and interoperability. From with social protection programs is very lim- an ASP perspective, the interoperability ited. Early warning data have rarely been between social protection and DRM informa- used to inform direct ASP or DRM support tion systems can inform the preparedness to households affected by climatic hazards. and response actions of both sectors (Barca and Beazley 2019). Most social assistance programs in Lesotho TA B L E 4 . 4   Scorecard for data and information systems in Lesotho lack digital beneficiary registries with MISs, which constrains the possibility of shar- Indicator Score ing data and making use of them for social Composite country rating protection responses. The capacity of most programs for data sharing, monitoring and Foundational indicators evaluation, and updating information at Data and information on disaster risk and household vulnerability ● decentralized levels is limited. The CGP is a Beneficiary registry ● notable exception; the program has a digital Social registry MIS and a beneficiary registry that evolved over time to become NISSA, the national Adaptive indicators social registry. The CGP relies on NISSA Use of preexisting data for shock response ● data for poverty targeting. Post-shock household needs assessments ● NISSA contains socioeconomic data on vir- Data-sharing protocols with internal/external partners ● tually all the households in the country, Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. providing a unique platform for ASP actions. 4 :   L esotho 99 Early surveys collected data for NISSA, going programming but also on its use in respond- back as far back as 2009, while the latest ing to shocks. The census swipe strategy for survey rounds for rural areas were conducted data collection implemented so far is very from late 2016 to mid-2019 (GOL 2020d). expensive, and both the MOSD and key part- The registry currently covers all 320,254 ners have called for a more cost-effective rural households in the country. An additional method to be developed. The sustainabil- 157,309 urban households were surveyed at ity of this registry is also in question, with the end of 2020, with 100  percent enroll- serious challenges presented by its limited ment anticipated by mid-2021. NISSA data capacity at the local level, its scale (provid- are used to determine eligibility for the CGP ing full national coverage), and the amount and the OVC Bursary. The MOSD also plans of data that need to be collected. For this to utilize the registry for other social support reason, a strategy has been developed for programs, including the Public Assistance updating NISSA through local structures program, but this has yet to occur. (GOL 2020d); this is currently being piloted. Despite its comprehensiveness, NISSA’s RISK AND VULNERABILITY interoperability and integration are very lim- M A P P I N G A N D E A R LY ited, and its potential with regard to ASP WA R N I N G SYS T E M S has not been fully exploited. Interoperability refers to the ability to exchange and make Annual vulnerability assessments are at pres- use of information. NISSA’s data exchanges ent the main source of information in Lesotho are rare and mostly ad hoc, rather than based for ASP and DRM actions. Lesotho has con- on protocols and agreements (GOL 2020d). ducted annual vulnerability assessments that Integration refers to sharing or consolidat- estimate the number of people in need of ing the processes related to data collection, food assistance since 2003. Led by the LVAC, management, and use among several pro- these assessments involve a range of national grams and institutions. The only process and regional government institutions, as well integrated so far with respect to NISSA as a number of development partner agen- has been the targeting of the CGP and the cies. The resulting VAA is published each OVC Bursary. NISSA (and hence the CGP) May or June. The VAA is based on the Inte- does not yet use the national ID as a unique grated Food Security Phase Classification identifier, which is an obstacle for further (IPC) and the Household Economy Approach interoperability and integration. Under the (HEA). The IPC classification can be used to leadership of the MOSD and the Ministry of estimate the number of people in each phase Home Affairs, and with UNICEF support, a of food insecurity in each district; the HEA portal was recently developed to ensure the methodology is used to assess food and cash accessibility of NISSA data. The goal of this needs among the vulnerable population. The initiative is to enable stakeholders to apply HEA uses a livelihood-based analytical frame- data as a targeting tool during both normal work to measure food security, based on the and emergency situations. Reportedly, how- Livelihood Impact Assessment Sheet meth- ever, use of the portal is still very low. odology (Anderson, Matthew, and Murray 2017). A number of challenges are associated NISSA lacks a mechanism for updating data with using VAA data for ASP. For one thing, frequently and at scale, which is a con- as they are available only once a year, these straint not only on regular social protection data are not designed to capture sudden 100 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA changes in consumption, food security, and data have informed ASP responses to shocks other variables. In addition, VAA data differ in recent years, particularly with relation to from NISSA data in that the former can be the number of people with food and liveli- disaggregated at the livelihood zone level and hood deficits, the geographical areas most the latter by administrative level (for example, affected, and the amounts of cash grants by community councils and villages), but no needed to support affected households. methodology is now in place for matching community councils with livelihood zones. As NISSA has played a central role in recent a result, any response based on the VAA data shock responses. A database that covers cannot be designed at levels lower than the virtually all households in the country district level. and contains demographic and socioeco- nomic data can be very useful for targeting The national early warning system provides responses. The use of NISSA data in recent limited information for postdisaster support vertical and horizontal expansions has by DRM or social protection. According to encountered some challenges, however: the early warning system manual, it is the role of the Lesotho Meteorological Services ● NISSA’s MIS was not designed to pro- to undertake climate-related hazard map- vide multiple payments to households. ping, to identify the geographical areas that This caused substantial delays in the may be affected, and to inform DRM actions. response to COVID-19. While line ministries like the Ministry of Agri- ● As noted above, NISSA’s interoperabil- culture and Food Security and the Ministry ity with other databases and systems of Water also produce early warning system is very limited. During the response data, the Lesotho Meteorological Services to COVID-19, problems caused by the should be the agency generating meteorolog- NISSA MIS in delivering multiple pay- ical data that can inform response actions. Its ments prompted officials to consider capacity is very limited, however; its human using Old Age Pension software, which resources, observational infrastructure, and does allow for such transfers. The lack of data are largely assessed as nascent. interoperability between the two systems was an obstacle, however. U S E O F P R E E X I S T I N G D ATA ● Because the capacity of the MOSD’s TO I N FO R M ADAPTIVE information technology department is S O C IAL P R OTECTI O N limited, accessibility to NISSA data is not RESPONSES always timely. The use of preexisting data can enable rapid Although NISSA and VAA data have been response to shocks while new data are col- used to inform recent ASP responses, pro- lected. The existing data may include social tocols and memorandums of understanding protection data (for instance, from social for information sharing are lacking. Data registries) and DRM data (from vulnerabil- exchanges are mostly ad hoc. The absence ity assessments), among others (Barca and of protocols affects not only the efficiency of Beazley 2019). data sharing but, more importantly, data pro- tection and security. Both VAA and NISSA data have informed recent horizontal and vertical expansions. VAA 4 :   L esotho 101 NISSA’s registry was not designed to inform POSTDISASTER NEEDS ASP actions, and it is not entirely fit for this A S S E S S M E N T S A N D D ATA purpose. The usefulness of existing social C O L L ECTI O N protection information systems in shock Depending on the type and scale of the response is determined by a number of fac- shock and the response strategy, the use tors. Table 4.5 presents five main dimensions of preexisting data may need to be comple- of data quality—completeness, relevance, mented by ex post data collection exercises. currency, accessibility, and accuracy—and This would aim to improve and complement assesses the NISSA system against these. initial targeting, which relied on ex ante data TA B L E 4 . 5   Utility of NISSA data for informing shock responses Dimension NISSA characteristic Implications Completeness: number of Virtually all households Registries can extend or update records compared with what their coverage to include regions would be perceived as a full set affected by recurrent shocks ■ Does not include variables that can help predict vulnera- ■ Operational information is rel- bility to shocks (e.g., “climate-smart”) evant for any type of shock Relevance: contains the vari- ables required for the intended ■ Some operationally relevant information is missing (e.g., ■ Socioeconomic data may purpose georeferencing, bank account details) be more relevant for certain types of shocks (slow onset, ■ Includes socioeconomic data to estimate poverty levels recurrent) ■ Data were collected between 2016 and 2019; many Rapid-onset disasters may cause records are likely to be outdated and not reflective of widespread internal displace- current circumstances (more so after shocks that have ment, split up households, and affected the country) significantly change their mate- Currency: degree to which data rial circumstances are up to date ■ CGP data can be updated more easily than NISSA data, through program operations ■ No mechanisms for updating data; system sustainabil- ity is at risk ■ NISSA is digitally maintained and stored, and information Challenges of accessing a data- is shared on an ad hoc basis and with substantial delays base are compounded in a Accessibility: ease of access rapid-onset natural disaster ■ Few data-sharing agreements are in place for potential users ■ Whether NISSA portal will enhance accessibility remains to be seen ■ Processes for continuously verifying and validating exist- ing data are missing Relevant for every type of shock Accuracy (integrity): data are ■ Data users including the DMA have raised concerns considered accurate if free about NISSA’s accuracy (GOL 2020d) from errors and omissions— that is, the data can be trusted ■ In poverty-targeted programs like the CGP, high errors of inclusion and exclusion can affect perceived accuracy and usability of data for responses Source: Based on Barca and Beazley 2019 and Barca and O’Brien 2017, except where otherwise noted. 102 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA R E C O M M E N D AT I O N S F O R D ATA A N D that do not describe ex post conditions and/ I N F O R M AT I O N S Y S T E M S I N L E S O T H O or to reach those left behind because they are not included in the registries used (Barca The recommendations for this building block focus on making and Beazley 2019). NISSA more accessible, interoperable, and risk informed, as well as on improving its data. Recommendations are also provided for The government of Lesotho’s capacity to other registries and for collecting data after a shock occurs: conduct postdisaster household needs assessments at a large scale and to ensure 1. Ensure that social protection programs (beyond the CGP) the data collected inform social protec- use NISSA data and feed data back to NISSA. In line with the tion responses is very limited. The Disaster second National Strategic Development Plan, which envisions Management Act of 1997 gave the DMA a NISSA as a key tool for improving efficiency, effectiveness, and mandate to lead a national targeting exercise integration, social assistance programs should use NISSA for to identify households in need of support in their eligibility assessments and for informing policy and pro- times of crisis. While the DMA thus takes the gram design. Accessibility to NISSA information needs to be lead in preparedness and response, sectoral extended beyond the social protection sector. lead institutions are responsible for undertak- 2. Develop mechanisms for updating NISSA data. Although a pilot is ing assessments in their respective sectors. under way, what data need to be updated, for what population(s), The exercise is organized at the district level and how frequently remain to be defined. Despite the merits of by district disaster management teams, which having a registry that covers the whole population, collecting and use community-based targeting approaches. updating data for all the variables for every household may not be The DMA and other actors apply a cutoff sustainable. A combination of approaches may be required. based on the percentage of households esti- 3. Develop protocols and policies for data sharing, data protec- mated by the LVAC to be in need of support tion, and privacy, and use the national ID as a unique identifier. and administer a survey to those below the Data exchanges so far have been ad hoc, posing risks in terms cutoff. The district teams verify samples of of inefficiency and security. Using the national ID as a unique these questionnaires in people’s houses. This identifier can enable greater interoperability and integration information is then shared with the interna- both within and beyond the social protection sector. tional or national organizations that respond 4. Ensure the data collected by NISSA are relevant for the responses to the crisis (Kardan, O’Brien, and Masasa envisioned. It may be worth collecting additional information to 2017). Despite the merits of this approach measure vulnerability and exposure to shocks, along with oper- (relying on local structures, prioritizing com- ationally relevant information, such as contact details. munity perspectives), resource constraints heavily impede its effectiveness in practice, in 5. Invest in adapting NISSA’s MIS to ensure it can perform the tasks terms of both data collection and data shar- envisioned for shock responses—that is, allow for multiple pay- ing. The targeting exercise can take months ments, multiple recipients, and different delivery mechanisms. when the scale of the shock is large, and the 6. Social assistance programs should develop digital registries DMA does not have its own database or MIS and MISs to improve efficiency and transparency and potential to store, manage, and share the data Kardan, responsiveness to shocks. O’Brien, and Masasa 2017). 7. Strengthen the government’s capacity to conduct postdisas- ter needs assessment at large scale, possibly in partnership with humanitarian and civil society actors. This effort should be undertaken in collaboration with social protection to ensure the data collected are useful for ASP actions and that data can be exchanged. 4 :   L esotho 103 BUILDING BLOCK 4: TA B L E 4 . 6   Scorecard for finance in Lesotho FINANCE Indicator Score Lesotho does not have financing mecha- Composite country rating ● nisms in place to support ASP; of the four Foundational indicators building blocks, this is the least developed. Social protection spending Social protection spending is already very high in the country, and the current con- Disaster risk financing policies and strategies strained fiscal situation limits its further Adaptive indicators expansion. Efficiency gains are needed to Quantification of the expected costs for social protection increase fiscal capacity. In the meantime, responses ● funding for many key investments in social Disaster risk financing instruments linked to social protection ● assistance (and DRM capacity) has had to Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is come from development partners. Critically, between two stages of development. each of the recent vertical and horizon- tal expansions has also been financed by development partners. The government of F I G U R E 4 . 9   Social assistance spending as a percentage of Lesotho does not have a disaster risk financ- GDP for SACU and other countries ing strategy in place and has no financial 6.4% 7 commitments to fund the scale-up of social protection when need arises. Although some 6 Percentage of GDP initiatives to quantify future ASP costs have 5 been undertaken, including the creation of 4 a costing tool and scalability framework, the 3.3% 2.9% 3 2.6% findings have not yet been fully incorporated UMI – 2.7% into policies or plans. 2 1.7% HI – 1.9% OECD – 1.6% 1 LMI – 1.4% Following are detailed findings and rat- 0 ings justifications for this building block Botswana Eswatini Lesotho Namibia South Africa (table 4.6). Sources: Spending data (based on administrative data) are from latest available years in ASPIRE (Atlas of Social Protection Indicators of Resilience and Equity) Database (Botswana: 2014–16; Eswatini: 2010–11; Lesotho: 2014–17; South Africa: 2015). Regional S O C IAL P R OTECTI O N averages are from World Bank 2018. Data for OECD countries refer to 2013 and are based on the social expenditure database. SPENDING Note: HI = high income; LMI = lower-middle income; OECD = Organisation of Economic Co-operation and Development; UMI = upper-middle income. Economies are divided Social protection spending in Lesotho is very among income groups according to 2016 gross national income per capita, calculated using the World Bank Atlas method. The groups are as follows: LMI = $1,006–$3,955; high. At 6.4 percent of gross domestic prod- UMI = $3,956–$12,235; HI = $12,236 or more. uct (GDP), Lesotho’s spending is over twice that of its neighbors (World Bank 2021b; see figure 4.9). While in theory high spend- social assistance programs covered in this ing could be associated with more robust report, most of the spending is allocated to administrative systems, the service deliv- the Old Age Pension (2.03 percent of GDP), ery constraints described under Building followed by the National School Feeding Block 2 for both social protection and ASP program (0.56 percent of GDP). In contrast, have meant that spending has not neces- very little is spent on programs that reach sarily led to stronger systems. Among the the poorest. The CGP only accounts for 104 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA 0.15  percent of GDP (based on 2017 fig- DISASTER RISK FINANCING ures); the program’s coverage is relatively P O L I C Y A N D S T R AT E G I E S small, and benefit levels are low. Although The government of Lesotho does not have this percentage has probably increased with a disaster risk financing strategy in place. the scale-up of the program in recent years, It does, however, have a few ex ante and it remains low. Similarly, the Public Assis- ex post financing instruments (described in tance program accounts for 0.11 percent of box 4.3) it can utilize for risk layering—an GDP. The limited spending on the poor and approach that allows countries to combine vulnerable, beyond the Old Age Pension, financing instruments to match needs aris- implies weak support for resilience building. ing from different types of disasters. In risk layering, relatively inexpensive instruments Lesotho’s current fiscal situation is con- are used as a first line of defense against strained, limiting further expansion of social high-frequency, low-severity events, with protection spending. According to the World more expensive instruments used only when Bank (2021b) this was already the case the cheaper sources of financing have been before the COVID-19 pandemic, which may exhausted to meet needs associated with have constrained fiscal resources even low-frequency, high-severity events (World further. Consequently, an expansion of Bank 2014). Such an approach enables programs supporting the poor and an adjust- countries to match the timing and amount of ment of benefit values would have to be their resource expenditures with their needs. funded by reallocating resources and with efficiency savings. Fiscal capacity for con- tingent funding for the scale-up of social Q U A N T I F I C AT I O N O F P O S T - protection in response to shocks is thus very S H O C K ADAPTIVE S O C IAL limited. P R OTECTI O N C OSTS An exercise was conducted to quantify future The growth of the social assistance sector in ASP needs, but the results have not yet been Lesotho—in particular of the CGP and of ASP incorporated into plans or strategies. As part initial investments—has been substantially of an ECHO-funded project in 2019, the cost funded by donors. While the government of scaling up the CGP in response to droughts fully funds its regular cash transfer pro- was quantified. Box 4.4 summarizes the grams, system strengthening investments estimated costs included in the scalability are primarily funded by donor partners. This framework developed, as well as the limita- includes the funding of consultants to work tions of the approach used. This scalability in the MOSD, of firms to collect NISSA data plan was not backed by any financial com- and develop operations and policy manuals, mitments to guarantee the availability of and of researchers to conduct evaluations, a predefined amount of resources during as well as funding for responses to recent crises, and it has not yet been adopted by shocks. Although regular transfers are the MOSD, despite the ministry’s active par- funded by the government and the programs ticipation in its development (GOL 2020h). (except for the National School Feeding pro- gram) are implemented with government structures, this continuing dependence on external support will affect future ASP programming. 4 :   L esotho 105 B OX 4.3   Lesotho’s disaster risk financing instruments Lesotho has a few financing instruments in place Lesotho does not have a contingent credit to meet that the government can combine in a risk layering disaster-related needs or sovereign disaster risk approach to disaster risk financing. These include both insurance. ex ante and ex post instruments. Ex post disaster risk financing instruments. The govern- Ex ante disaster risk financing instruments. Ex ante ment does not use risk transfer instruments to mobilize disaster risk financing instruments can help govern- financing after disasters, although sovereign insur- ments respond to different types of disasters in a ance options are available in the Sub-Saharan Africa timely manner. Several instruments could be used for subregion (World Bank 2019b). The two main ex post this purpose, such as contingency funds, contingent financing strategies used by Lesotho are budget real- lines of credit, and risk transfer instruments. locations and humanitarian appeals. Although such ad hoc government budget reallocation can be a way Lesotho has two contingency funds, one managed of meeting urgent financial needs following disasters, by the Ministry of Finance and the other, the Disaster it can undermine planning and jeopardize the govern- Management Fund, by the DMA. Generally, contin- ment’s development objectives as funding is taken gency funds are used to cover the needs associated away from other important projects. with recurrent high-frequency, low-intensity events (World Bank 2019d). The Ministry of Finance’s con- Lesotho has received a yearly average of $7.28 million tingency fund is not specifically designated to cover in total donor assistance and $4.15 million in human- disaster-related needs, but rather is used to meet itarian assistance for food security over the period needs for a broad range of different purposes, such as 2002–18 (World Bank 2019b, 2021b). While donor additional expenses of ministries. As it is not a revolv- assistance is free from the government’s point of view, ing fund, the allocations to it must be spent within the amount that will be made available is uncertain, one year. In Lesotho, the rainy season falls toward and the provision of assistance can be slow, leading to the end of the fiscal year, so by the time droughts or delays in the response. To date, Lesotho has not used floods strike, very few resources generally are left in an emergency credit to meet disaster-related needs the contingency fund (World Bank 2019b). Meanwhile, (World Bank 2019d). allocations to the Disaster Management Fund have been decreasing, and its size is inadequate for Leso- tho’s risk profile. DISASTER RISK FINANCING protection when the need arises. The use M EC HAN I S M S FO R ADAPTIVE of social protection in response to shocks S O C IAL P R OTECTI O N has relied mostly on external financing and mechanisms developed ex post. Despite the recurrence of certain shocks— such as droughts—no financial commitment is currently in place to fund scale-up of social 106 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA B OX 4.4   Contingency planning for CGP scale-ups in response to droughts In 2019, a scalability framework for ASP—consisting of ■ Duration. According to the same calculations, the a set of principles, parameters, and rules for scaling up monthly assistance in 2019/20 should have cov- social protection in response to a shock—was devel- ered six months. oped as part of an ECHO-funded project. The project was led by the MOSD and the DMA and took a consul- Under the framework, reaching the annual average tative approach. number of households in survival deficit (47,313) during the past five years with a monthly transfer of M 756 Under the framework, scale-up will be triggered by ($51) lasting six months (and excluding operations the annual VAA report, which is the main, most reli- costs) should have cost M 214,612,448 ($14,379,034). able, and most timely source of information about the effects of droughts in the country. The report is avail- Although similar to some of the annual contingency able every year around July, before the beginning of the plans developed by the DMA and the MOSD, the drought season, and it provides projected estimates of approach taken under this framework is not backed by the number of people in survival and livelihood defi- any financial commitments and is probably impossi- cits and of food and livelihood gaps. It also estimates ble to afford with domestic resources. The cost noted the duration in months of the drought season for the above, for example, represents 2.3 times the annual coming year. VAA estimates are based on survey data CGP budget (approximately M 75 million [$4.5 million], and other sources, and they can be disaggregated at transfer only). the district or livelihood zone level. Two lessons emerged from the development of the Three key parameters drive the cost of a scale-up: scalability framework with respect to ASP financing. coverage, transfer value, and duration. The scalability framework considers them as follows: ■ The lack of financial commitments for future scale-ups, and therefore of a budget ceiling, makes ■ Coverage. The framework targets households with contingency planning difficult. survival deficits in districts where more than 20 per- ■ Although covering the entire deficit gap has obvious cent of the population is in survival deficits. The merits, it is also important to assess whether it is the yearly average number of eligible households in role of a social protection scale-up to do so. Survival 2014–18 was 47,313. deficits in Lesotho are caused by chronic and acute ■ Transfer value. According to LVAC calculations, needs. Despite the severity of shocks, some people each household in survival deficit should have are always affected because of chronic conditions. received a monthly transfer of approximately M 756 The scalability framework is designed to respond to ($51) to cover the food gap in 2019/20. acute needs and may not be the right tool for cover- ing the entire food gap. Source: Based on GOL 2020i. 4 :   L esotho 107 R E C O M M E N D AT I O N S F O R F I N A N C E I N L E S O T H O As the least developed ASP building block in Lesotho, finance requires more attention and prioritized investments. The government of Lesotho should undertake to do the following: 1. Identify ways to finance the expansion of social protection coverage and the increase of benefit amounts through efficiencies, budget reallocations, and strategic support from development partners. 2. Develop a disaster risk financing strategy, including provision for ASP. The strategy, based on a risk layering approach that combines financing instruments to match needs arising from different disasters, should put forth a plan and a commitment to fund ASP actions. It could be based on the CGP scalability framework already devel- oped, taking into account the lessons from recent ASP responses and including the roles of other programs, such as the Public Assistance program. 3. The scalability framework should be further developed on the basis of agreements on how to respond to shocks, when to respond, who should receive support, and what support should be given. Defining these parameters beforehand will allow responses to be implemented more rapidly and on the basis of transparent criteria and procedures. “Frontloading” all these decisions requires strong commitment and leadership, however—and, in particular, funding commitments. In fact, although the development of the framework includes aspects of all four building blocks, it is included under finance to emphasize the importance of funding commitments. PRIORITY for investment are related to institutional capacity and coordination, the delivery of INTERVENTIONS cash payments, and the development of FO R L E S OTH O disaster risk financing mechanisms. The analysis presented in this case study Table 4.7 presents priority actions for both shows that ASP in Lesotho is still between foundational social protection and ASP. Its the nascent and emerging stages, which recommendations are worth prefacing with constrains the ability of social protection some policy reflections. to respond to shocks. Several elements highlighted by the traffic light scorecard ● Beyond responses: enhancing resilience represent strong foundations and adaptive and adaptation. VAA data show that in measures. The inclusion of ASP provisions years when no major disaster or shock in both social protection and DRM legisla- occurs, chronic needs are more predom- tive frameworks and policies in Lesotho—in inant than acute ones. That is, because addition to the country’s high spending on of unresolved underlying poverty and social protection, the existence of a social vulnerability issues, households lack the registry that covers virtually all of its pop- adaptive capacity to cope with minor ulation, recent experiences in expanding adverse events. While the scale-up of the CGP and other programs vertically and social assistance programs can support horizontally, and strong collaboration with people with chronic needs temporarily, development partners—is critical to the a sustainable and long-term approach strengthening and further development of would offer social protection, resil- its ASP capacity. The most strategic areas ience building, climate adaptation, and 108 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA economic and financial inclusion policies ● Finding a balance between planning and programs. and flexibility. Contingency plans and ● A government-led approach. The support scalability frameworks, in which most rel- of development partners and donors has evant response decisions are predefined, been fundamental in the expansion of the can enable timely and more transparent sector; examples include the scale-up responses. They are especially suitable and improvement of the CGP, the devel- for the types of shocks to which Lesotho is opment of NISSA, the development of highly exposed—namely, high-frequency, policies and plans, and the responses to low-severity events (recurrent droughts) recent shocks. Given the limited capac- and medium-frequency, high-severity ity of the government of Lesotho and events (El Niño–induced droughts), the constraints on its provision of social where needs can be predicted with protection and ASP support, these part- greater accuracy. In addition to plan- nerships are vital. Policies, plans, and ning and pre-agreed commitments, it intersectoral forums do not necessarily is important for the system to have the translate into stronger ownership, lead- flexibility to respond to low-frequency, ership, and coordination. The DMA’s high-severity events like the COVID-19 constraints, for instance, are rooted in pandemic. This flexibility should be built profound issues of political economy across the four building blocks. that are complicated to resolve. 4 :   L esotho 109 TA B L E 4 . 7   Summary of all recommendations by building block for Lesotho Building block Foundational social protection Adaptive social protection ■ Empower the DMA and strengthen its capacity to ■ Integrate ASP provisions into DRM policies and plans lead and coordinate the sector more effectively Institutional ■ Provide adequate resources to the MOSD’s infor- ■ Put into practice the memorandum of understand- arrangements mation technology department ing between the MOSD and the DMA for harmonizing and partnerships ■ Strengthen MOSD and DMA capacity to engage coordination structures with local authorities and communities ■ Expand the coverage of poverty-targeted pro- ■ Develop contingency protocols for business continu- grams to reach the intended population ity during crises ■ Adjust benefit amounts in line with program ■ Design protocols for scale-ups objectives ■ Complement protocols with adequate training and ■ Roll out digital payment mechanisms, in combi- resources nation with cash-in-transit delivery where still ■ Design new digital payment mechanisms that are needed flexible and scalable (e.g., standby agreements with Program design ■ Establish and document transparent criteria and service providers) and delivery processes for targeting and service delivery, ■ Revise the amounts of cash responses to ensure they including recertification mechanisms are adequate and agree on amounts with all partners ■ Ensure national IDs are issued to all beneficia- ■ Design and implement a mechanism for regular ries, and use NISSA as a tool for informing further cash support to households exposed to recur- ID coverage expansions rent droughts (e.g., top-ups at the beginning of the drought season) and accompanying interventions to support adaptation (including, e.g., with an emphasis on livelihood diversification and economic inclusion) ■ Develop mechanisms for updating NISSA data ■ Ensure the data collected for NISSA are relevant for based on a clear understanding of what data the responses envisioned; it may be worth collecting need to be updated, for what population, and how additional information to measure vulnerability and frequently; a combination of approaches may be exposure to shocks and operationally relevant infor- needed mation, such as contact details ■ Ensure social protection programs (beyond the ■ Adapt NISSA’s MIS to ensure it can perform the tasks CGP) use NISSA data for eligibility assessments envisioned in shock responses (i.e., allow for multiple and for informing policy and program designs payments, multiple recipients, and different payment ■ Increase accessibility of NISSA information providers) Data and infor- mation systems ■ Develop protocols and policies for data sharing ■ Strengthen the government’s capacity to conduct and data protection and privacy postdisaster needs assessments at large scale, possibly in partnership with humanitarian and civil ■ Use the national ID as a unique identifier society actors; liaise with social protection to ensure ■ Promote greater interoperability and integration the data collected are useful for ASP actions and within and beyond the social protection sector that data can be exchanged ■ Develop digital beneficiary registries and MISs ■ Enhance the capacity to exchange data with other for social assistance programs government sectors and with nongovernmental ■ Develop an integrated MIS for all MOSD pro- actors grams that links to central and district offices Develop a strategy for funding social protec- ■ Develop a disaster risk financing strategy based on a tion coverage expansions, benefit increases, and risk layering approach that combines financing instru- capacity-strengthening efforts with efficiency ments to match needs arising from different disasters savings, budget reallocations, and support of devel- ■ Develop a plan for ASP based on funding commit- Finance opment partners in strategic areas ments (i.e., a scalability framework) ■ Outline a strategy for financing ex ante household resilience building and adaptation interventions, making the case for linkages to climate adaptation financing SOUTH AFRICA A D A P T I V E S O C I A L P R O T E C T I O N A S S E S S M E N T C A S E S T U DY 5 times higher than for black Africans (Statis- OVERVIEW tics South Africa 2019). Level of education OF RISK AND affects poverty, as it affects income and job HOUSEHOLD opportunities. Poverty is also consistently higher among female-headed households, VULNERABILITY the less educated, the unemployed, and large families (Sulla and Zikhali 2018). POVERTY Poverty has a strong spatial dimension in Although South Africa is classified as an South Africa, with rural areas having a much upper-middle-income country based on higher concentration of poor inhabitants average per capita gross domestic product, (65 percent) than urban areas (25 percent). extremely high inequality masks widespread Eastern Cape, KwaZulu-Natal, and Limpopo poverty. Between 2005 and 2010, the per- were consistently the three poorest prov- centage of South Africa’s population below inces between 2006 and 2015. Gauteng the poverty line for an upper-middle-in- consistently had the lowest poverty rates percent, come country fell from 68 to 56  percent in 2015), and, at 26  (19  percent but subsequently trended slightly upward (2015), KwaZulu-Natal had the largest share percent in 2015 and was projected to 57  of the poor, reflecting in part the high popu- to reach 60 percent in 2020 (World Bank lation of this province. 2021f). Deep pockets of deprivation exist across the country, with an estimated 18.7 percent of South Africans in absolute SHOCKS AND poverty (that is, living on less than $1.90 per DISASTERS day). South Africa is generally seen as a nation at relatively low risk of disaster. It ranks 92nd Poverty varies demographically and geo- out of the 180 countries in the 2019 World graphically in the country. The legacy of Risk Index (Bündnis Entwicklung Hilft 2016). apartheid means that poverty in South Nonetheless, as table 5.1 shows, South Africa is highly racially skewed. Between Africa experienced more than 100 disaster 2006 and 2015, the annual median expen- events between 1952 and 2019, resulting diture for white Africans was more than 10 111 112 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 5 . 1   Incidence and impact of disaster events in South occurring once every two years, and they Africa, 1952–2019 result in the highest death tolls and economic impacts (table 5.1). While droughts occur less People Economic frequently (once every six years), they affect Hazard Occurrences Deaths affected losses ($) extremely high numbers of people and also Storm 34 299 651,895 1,855 million result in significant economic losses. Evi- Flood 46 1,378 589,856 2,938 million dence indicates extreme weather events in Drought 11 0 20,925,000 2,585 million South Africa are increasing, with tempera- Earthquake 5 37 3,112 136 million tures rising, dry spells getting slightly longer, Wildfire 11 137 12,880 863 million and rainfall intensity increasing (GOSA Landslide 1 34 n.a. n.a. 2017a). These changes result in more cli- Extreme 3 63 20 n.a. matic shocks happening more frequently. temperature Table 5.1 and figure 5.1 show the implications Epidemic 7 336 112,385 n.a. of this in terms of the financial costs of disas- Total 118 2,284 22,295,148 8,378 million ters. The compound effect of multiple shocks Source: Adapted from World Bank 2021e. occurring at the same time is also becoming Note: n.a. = not available. Aggregate impacts across different disaster events. apparent; 2017 stands out as the costliest year on record, with a combination of storms, floods, wildfires, and drought having pro- bil- in total economic losses of over $8.4  duced economic losses of over $2.5 billion. lion. Note that because the table does not extend through 2020, it does not include the The human impact is also increasing. In South COVID-19 crisis, the full impact and costs of Africa, disasters are estimated to result in an which are yet to be fully assessed. average of 63 deaths and affect more than 500,000 people each year (i.e., those who Climatic events dominate disasters in terms are injured, become homeless, or require of cost and human impact. Floods and storms food assistance). Drought in particular has are the most frequent disaster events, million South Africans affected nearly 22  F I G U R E 5.1   South Africa: financial cost of disasters, 1968–2019 2,600 2,400 Insured loss 2,200 Uninsured loss 2,000 1,800 1,600 Million $ 1,400 1,200 1,000 800 600 400 200 0 2019 1968 1974 1978 1981 1984 1986 1988 1991 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 Source: World Bank 2021e. 5 :   S outh A frica 113 in the past decade. The deadliest year was M A P 5 . 1   Hazard levels for various types of flooding in South 1987, in which a flood in KwaZulu-Natal killed Africa, 2021 506 people and resulted in economic losses million. In 2004, floods in Cape of $765  a. River flooding b. Urban flooding c. Coastal flooding Town resulted in 15,000 homeless, and a widespread drought across KwaZulu-Natal, Eastern Cape, Northern Cape, North West, Mpumalanga, Limpopo, and Free State affected over 15 million. In aggregate, drought has the most extensive impact, while flood and storm have the most intensive impact. Source: ThinkHazard! South Africa, Global Facility for Disaster Reduction and Recovery. Note: = high; = medium; = low; = very low. The frequency and impact of different cli- mate hazards vary in different parts of South Africa. The geographical areas most MAP 5.2  South Africa standard precipitation index, 2016 affected by the primary climate hazards are outlined below. F LO O D S South Africa is considered at high risk for all three types of flooding—river, urban, and coastal. Map 5.1 shows that the Northern Cape, Eastern Cape, and KwaZulu-Natal provinces have the greatest exposure. DROUGHTS Nearly every part of South Africa is highly vulnerable to drought, and the frequency and severity are worsening over time. (See Source: Agricultural Research Council 2016. map  5.2.) A national state of disaster has been declared twice in the last three years due to drought. Four of the country’s nine provinces—Limpopo, Northern Cape, West- WILDFIRES ern Cape, and Eastern Cape—have declared repeated states of disaster after prolonged Wildfires occur regularly in South Africa and droughts between 2015 and 2019. Although often cause significant damage to property drought has direct impacts on rural pop- and critical infrastructure, such as power ulations, which rely on agriculture (both lines. The combination of expanding human commercial and subsistence/smallholder), settlements with climate and vegetation urban populations are increasingly affected. characteristics that encourage fires results Chronic drought affects urban water sup- in substantial wildfire risk across the coun- plies and can undermine power supplies, try, particularly in eastern regions. In these particularly from hydroelectric production. areas, informal settlement dwellers are 114 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA MAP 5.3  Days when fire danger is “very likely” in South common with the governments of many coun- Africa tries, the government of South Africa has instigated a range of public health measures, including several complete and partial lock- downs, to control the spread of the disease. The economic impact of these measures is still to be fully assessed, but latest estimates indicate gross domestic product (GDP) in 2020 was likely to have contracted by 7 per- cent (UNDP 2020). The impacts are being felt throughout the country, and unemploy- Source: Schulze 2016. ment reached over 30 percent in 2021. The Note: Based on present and future climate data. lockdown measures most severely affected low-income urban populations that rely heavily on ad hoc daily labor and petty trad- generally the most vulnerable. Map 5.3 indi- ing. A major adaptive social protection (ASP) cates that the situation will only worsen as a response was implemented and is described result of climate change. in more detail in the following section. South Africa thus faces multiple disaster risks, many of which are likely to be exacerbated by ASSESSMENT OF climate change. To date, the country’s overall disaster risk has been assessed as relatively ADAPTIVE S O C IAL low in comparison to other African nations. P R OTECTI O N Due in part to the size of its economy, gov- ernment capacity to respond, and private BUILDING BLOCK 1: insurance, the cost and impact of individ- ual shocks has been largely absorbed. The INSTITUTIONAL increasing frequency and combination of ARRANGEMENTS AND hazards, however, will work to compound PA R T N E R S H I P S disaster risk in the coming decades. This In general, robust and relevant social pro- increasing risk comes in the wake of the tection and disaster risk management policy COVID-19 pandemic, which is clearly demon- and legislative frameworks are in place in strating the very real threat disasters pose to South Africa. The right to social security, South Africa’s ambitions to reduce poverty including appropriate social assistance, is and inequality and increase employment. enshrined in the South African Constitution and Bill of Rights (1996) and guaranteed South Africa was among the countries hit in law. Similarly, the Disaster Management hardest by COVID-19 in Sub-Saharan Africa. Act (2002), Disaster Management Frame- million cases and over 51,000 Over 1.5  work (2005), and National Climate Change deaths were recorded as of January 2021.1 In Adaptation Strategy (2018) provide the legal authority for a comprehensive approach to disaster preparedness, reduction, mitiga- 1 Source: South Africa Department of Health tion, and response. COVID-19 web page. 5 :   S outh A frica 115 Two recent amendments (2020 and 2022) because it straddles several ministries. This to the Social Assistance Act of 2004 have poses a challenge for coordination, mapping, provided a mandate for the use of the Social and referral of clients and beneficiaries. DRM Relief of Distress (SRD) grant program in capacity varies geographically, depend- response to disasters. Historically, the SRD ing on the resources allocated by provincial program has mostly provided in-kind support and municipal authorities, and coordination to temporarily meet needs from an idiosyn- between social protection and DRM delivery cratic shock to households, such as death agencies is limited to nonexistent. of the primary breadwinner or a housefire. Recognizing the shortcomings of the social Following are detailed findings and rat- protection system to support households ings justifications for this building block negatively affected by natural disasters, the (table 5.2). government has formalized the use of the SRD grant program to also provide support P O L I C Y A N D L E G I S L AT I V E to people affected by natural disasters. To FRAMEWORKS formally define the role of the SRD in disas- ter response, an amendment to the 2004 South Africa’s social protection sector is Social Assistance Act was made in Decem- supported by a strong and comprehensive ber 2020 to “provide for social relief of statutory legal and policy framework. The distress in the event of disaster.” A further system is enshrined in the Constitution and amendment was introduced in May 2022 Bill of Rights, which guarantee South African that built on this disaster response mandate citizens the right to “social security, includ- with specific directions on the nature of the ing, if they are unable to support themselves SRD support to those affected by disasters, and their dependents, appropriate social including with high-level directions on the assistance” (GOSA 1996, article  27). The processes to verify eligibility as well as ben- efit amounts and periodicity. This represents an important step in clarifying the mandate TA B L E 5 . 2   Scorecard for institutional arrangements and of social protection in disaster response, partnerships in South Africa albeit in reference to only one flagship disas- ter response program. Implementation of the Indicator Score legislation will be important in translating Composite country rating ● this mandate into a program that is capable Foundational indicators of and adequately prepared to respond to Social protection policy and legislative frameworks ● large natural disasters at scale. Disaster risk management policies and legislative frameworks ● Institutional capacity to deliver social pro- Social protection institutional capacity and coordination ● tection and DRM programs and responses, Disaster risk management institutional capacity and coordination ● despite gaps and imperfect coordination, Adaptive indicators is relatively strong compared with neigh- boring countries. The South Africa Social Clear mandate for social protection response ● Security Agency (SASSA) provides compre- Multisectoral coordination for social protection responses ● hensive national coverage for the key social Partnerships with nongovernmental actors — assistance programs. SASSA is not respon- Note: ● = nascent; ● = emerging; ● = established; — = not available. Mixed colors imply sible for all social protection interventions the rating is between two stages of development. 116 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA 2004 Social Assistance Act provided for the adaptation interventions for their respective creation in 2006 of SASSA, which sits within jurisdictions. Federal, provincial, and munic- the Department of Social Development and ipal administrations are required to develop, is mandated to ensure the “effective and regularly update, and implement disaster efficient administration, management and management plans that address disaster and payment of social assistance and social climate change risks. The National Climate security” (SASSA 2020). The National Change Adaptation Strategy (GOSA 2019) Development Plan for 2030 (GOSA 2012) embeds DRM actions in the achievement of reaffirms commitments to the provision of a many of its objectives. social security system for all working people, with social protection for the poor and other I N S T I T U T I O N A L C A PA C I T Y groups in need, such as children and people A N D C O O R D I N AT I O N living with disabilities. Limited funding for DRM means capac- The role for social protection in shock ity to implement the Disaster Management response was recently made explicit in leg- Act—particularly at the provincial and munic- islation. An amendment to the 2004 Social ipal levels—is limited and variable, which Assistance Act was made in December 2020 undermines its effectiveness. The Disaster to “provide for social relief of distress in the Management Act established the National event of disaster.” The amendment specified Disaster Management Centre as well as a net- the SRD grant program be used in response work of disaster management centers across to natural disasters. A further amendment the country at the provincial and munici- in May 2022 built on this disaster response pal levels. These centers are mandated to mandate with specific directions on the coordinate and promote integrated disas- nature of the SRD support to those affected ter management, build capacity, and improve by disaster. the resilience of communities and infrastruc- ture to reduce risk. They are often poorly The Constitution of South Africa recognizes resourced, however, particularly at the munic- the role of the state in responding to natu- ipal level, and are co-located within different ral disasters. As mentioned, the Disaster sectoral offices. With limited funding and Management Act (GOSA 2002) and National poor capacity, disaster management centers Disaster Management Framework (GOSA cannot effectively fulfill the requirements set 2005) are the two most relevant legislative out by the National Disaster Management and policy documents for DRM. They set Framework. These include the development out policy and institutional arrangements (or revision) of high-quality disaster manage- for disaster management at different levels ment plans, which feed into local integrated of government, including prevention, miti- development plans. As a result of the lack gation, emergency preparedness, response, of funding, capacity, and clarity of roles and and postdisaster recovery and rehabilita- responsibilities for the disaster management tion. The Disaster Management Amendment centers, elements of the DRM policy are often Act of 2015 (GOSA 2015) links DRM more taken forward by different departments or directly to climate change adaptation, agencies—among them the Department of assigning responsibility to national, pro- Agriculture, the Department of Forestry and vincial, and local governments to invest in Fisheries, the Department of Water and Sani- disaster risk reduction and climate change tation, and the South African Weather Service. 5 :   S outh A frica 117 The institutional capacity of government provide a basis—and precedent—for further agencies to deliver social protection pro- cooperation and integration as part of the grams in South Africa is well developed. National Integrated Social Protection Infor- SASSA has a clear management structure mation System (NISPIS) project (discussed and has developed strong systems for target- under Building Block 3). ing, case management, data administration, and payment for all social protection grants. DRM programming should, by its nature, The agency has annual objectives and tar- involve cross-sectoral integration and coop- gets on which it is audited, and it produces eration. The Disaster Management Act annual reports. It has a network of 9 pro- provided for the creation of the National vincial offices, 46 district offices, 389 local Disaster Management Advisory Forum, which offices, 1,163 service points, and 1,740 pay brings heads of national sector departments points nationwide (SASSA 2019). Despite together to coordinate actions on matters this extensive infrastructure, complaints relating to disaster management. This struc- persist that beneficiary access points for ser- ture should be replicated at the provincial vices are insufficient, primarily due to lack of and municipal levels. In reality, however, the staff. Limited financing for recruitment has disaster management role is often assigned resulted in an estimated 55 percent of posi- to civil defense structures (that is, fire and tions remaining unfilled (SASSA 2019). police services). This means the focus remains on ad hoc disaster response rather than mitigation and preparedness. M U LT I S E C T O R A L C O O R D I N AT I O N The recent increase in disaster events and SASSA is not responsible for the adminis- the renewed focus on DRM in climate change tration of all social protection grants and adaptation efforts require greater collabo- programs, which raises coordination issues. ration and coordination. Each local disaster A recent review of social assistance programs management center should be supported by (World Bank 2021f) concluded that integra- a multistakeholder forum of advisors from tion across social protection programs and different line departments or service provid- government levels is not particularly strong. ers. That said, there appears to be limited (if This referred to the lack of integration of or any) cross-sectoral linkage or coordination coordination among the main social protec- between social protection and DRM agen- tion grants, which focus on children and the cies on any level. Until the recent COVID-19 elderly, and on the two main public works ASP response, South Africa’s social protec- programs that target working-age adults tion system had only limited dealings with (discussed in Building Block 2). Improving large-scale crises. coordination among these different minis- tries and departments remains a challenge, Given the extensive coverage and capacity which was further highlighted during the of South Africa’s social protection system, COVID-19 crisis. The ASP response to the the role of nongovernmental agencies in pro- pandemic (discussed in more detail below) viding emergency response is much more necessitated urgent cross-governmental limited than elsewhere in southern Africa. work with the South Africa Revenue Ser- Additionally, the government of South Afri- vice to screen applicants applying for the ca’s own disaster response capacity (for COVID-19 SRD grant. This effort could non-ASP interventions) is well established. 118 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA This has left United Nations agencies and BUILDING BLOCK 2: international nongovernmental organiza- PROGRAM DESIGN AND tions with a minimal role in the provision of D E L IVE RY general disaster-related response in South Africa. While civil society, on the other hand, South Africa’s social protection systems does play a big role in providing safety net are mature by African standards, but the support to needy households on a localized COVID-19 crisis exposed gaps in them. and ad hoc basis, the on-demand nature of The social protection systems for vulnera- most social assistance grants means they ble groups offer national coverage of social supplement rather than fill gaps in the pro- assistance grants, which are well targeted to vision of cash for the core vulnerable groups the poorest populations. The main safety net of women, children, people living with dis- grants are available on demand to all who abilities, and older persons. meet the eligibility requirements. Payments represent a significant boost to consumption for the poorest households; over one-third of households benefit from at least one payment. Most payments administered by R E C O M M E N D AT I O N S F O R I N S T I T U T I O N A L SASSA, which has advanced payment and A R R A N G E M E N T S A N D PA R T N E R S H I P S I N delivery systems. These include the social SOUTH AFRICA grants payment system (SOCPEN), which is a digital management information system 1. Use momentum built from the social protection response to (MIS) with links to an automated pay- COVID-19 to establish a high-level working group in SASSA to roll system. The widespread coverage of review, enhance, and advocate for the role of social protection national IDs facilitated SASSA’s issuance of in shock response. The committee should be actively engaged debit or Mastercards that have significantly in the government’s ongoing dialogue centered on the just increased financial inclusion rates in South transition, including by making the case for the role of social Africa. protection broadly and the SRD program specifically in sup- porting climate adaptation. South Africa lacks a national ASP response 2. Building on the recent provisions for the SRD program, inte- for the most common and frequent natu- grate a clear mandate for social protection into DRM and ral disasters. Instead, local municipal and climate change policy and legislation on the specific roles of provincial authorities are responsible for social protection actors and interventions in responding to providing relief, which can result in differen- shocks, including natural disasters. tial and ad hoc provision depending on local 3. Integrate SASSA and the role of social protection programs for- priorities and resources. The SRD grant or mally into DRM coordination mechanisms at the national and assistance is the one program with explicit local levels, while empowering the National Disaster Manage- criteria for disaster-affected individuals and ment Centre and its network of disaster management centers households. But it is designed primarily for at the subnational level to fulfill their coordination mandate idiosyncratic shocks, such as the death of through increased resource allocation. the main income earner, and its coverage is limited. 5 :   S outh A frica 119 The response to COVID-19 was South Afri- system (such as unemployment payments ca’s first significant ASP response, and its and the Road Accident Fund) are designed rollout was impressive in terms of scale, cov- to support individuals through idiosyncratic erage, and timeliness. The response involved shocks. These forms of social protection are significant vertical expansion of existing pro- contributory and therefore apply to wealthier grams, the creation of the new SRD grant to individuals and those employed in the formal scale social protection horizontally to new sector. adult beneficiaries, and a remodeling of unemployment benefits for formal sector Social assistance interventions represent a workers. The ASP response to the COVID-19 direct safety net for the poorest. These inter- crisis has already provided many lessons. ventions address absolute deprivation and The experience should be used to develop the vulnerabilities of the poorest, aiming to clearer protocols, systems, and guidelines lift recipients out of poverty and increase for future ASP responses and for collabora- their resilience to shocks, and they are the tion and coordination with other sectors and main focus of this assessment. The review of ministries, particularly for DRM. programs below also mentions wider social protection interventions to support the poor, The South African social protection system including public works programs. The work is one of the most comprehensive on the programs are due to be expanded as part continent. Figure 5.2 illustrates the various of the post-COVID Presidential Employment elements of the system, which comprises Stimulus package (GOSA 2020). three pillars: social assistance, mandatory social insurance, and voluntary private insur- Following are detailed findings and rat- ance. Most of the statutory and voluntary ings justifications for this building block funds within South Africa’s social protection (table 5.3). FIGURE 5.2  South Africa’s social protection system Employers and Workers Financing General Tax Revenues Road Users Social Assistance Statutory Funds Voluntary Funds Social Grants Public Works Other Employment Other Health Retirement Programmes Child support Disability Older persons Expanded National Unemployment Road Accident Medical Retirement Grant-in-aid Public Works School Insurance Funds Fund Schemes funds Care depend. (Covid-19) Nutrition War veterans Community Programme Compensation Foster care Work Funds Children Working ages Older persons Other groups Intended Beneficiaries Source: World Bank 2021f. 120 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 5 . 3   Scorecard for programs and delivery systems in account for 94 percent of total spending on South Africa social protection grants. Indicator Score In addition to the regular grants, the gov- Composite country rating ernment of South Africa operates a range of Foundational indicators temporary or in-kind social protection assis- tance to other target groups. The SRD grant Benefit adequacy ● was established to provide temporary “assis- Eligibility criteria and targeting ● tance for persons in such dire need that Coverage of the main social assistance program ● they are unable to meet their or their fami- Financial inclusion and national ID ● lies’ most basic needs” (SASSA 2020). The SRD grant generally provides support for up Social protection payment systems to three months and is mostly in-kind. The Social protection delivery systems ● Expanded Public Works Program, admin- Adaptive indicators istered by the Department of Public Works and Infrastructure, targets working-age Experience of vertical expansion ● adults and aims to provide income, work Experience of horizontal expansion ● experience, and training to the unem- Protocols and contingency plans ● ployed. Beneficiaries can receive up to 100 Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is days of work experience and/or on-the-job between two stages of development. training in four sectors: infrastructure, non- state (supporting employment in nonprofit organizations), environment, and culture S O C IAL P R OTECTI O N and social. The Community Work Program, PROGRAMS OVERVIEW although part of the Expanded Public Works Program, is coordinated by the Department This overview of social protection programs of Cooperative Governance and Traditional in South Africa focuses on social assistance Affairs. This is because the Community Work grants, which account for most government Program is implemented by nonprofit agen- social protection expenditures. Descriptions cies, which must offer work that “contributes of the core grants and programs are fol- to the public good or improves the quality lowed by summaries of how they have been of life in communities” (World Bank 2021f, adapted or scaled in response to shocks. 18). The National School Nutrition Program is operated by the Department for Basic Edu- Social protection grants form the backbone cation and is estimated to provide meals to of the government’s noncontributory social million school-going children on a over 9  protection programs. Since 1994, coverage regular basis. Nearly 85 percent of children has increased from 2.7 million to 17.81 mil- percent attend schools in the poorest 20  lion, representing nearly one-third of the that operate the program. population. Table 5.4 lists the various social assistance programs in South Africa. Cover- age is dominated by three categorical grants: the Child Support Grant (12.7 million); the Older Persons Grant (3.55 million); and the Disability Grant (1.05 million). Together, they 5 :   S outh A frica 121 TA B L E 5 . 4   Social assistance grants and programs in South Africa Eligibility Number Program Beneficiary/age means tested Benefit amount enrolled Older Persons Grant Persons 60 and above Yes: income and ~R 1,860 ($125)/person/month 3.6 million assets Child Support Grant Children 18 or younger (up Yes: income ~R 415 ($30)/child/month 12.7 million to maximum of 6 children) Disability Grant Disabled persons 18–59 Yes: income and ~R 1,860 ($125)/person/month 1 million assets Care Dependency Grant Children under 18 Yes: income ~R 1,860 ($125)/person/month 154,735 Foster Care Grant Children under 18 No ~R 970 ($70)/person/month 355,609 War Veterans Grant Persons 60 or over (or Yes: income and ~R 1,860 ($125)/person/month 62 disabled) assets Grant-in-aid Normally adults over 18 No: implied ~R 415 ($30)/person/month 273,922 Total regular grant recipients 18,290,592 Temporary or in-kind programs Social Relief of Distress Individuals and fami- No: implied No set value, usually in-kind 344,482a lies experiencing undue (e.g., food parcels or school uni- hardships forms or supplies) Expanded Public Works Working-age adults Provides up to 100 days of work 267,000b Program experience Community Work Program Working-age adults Provides up to 100 days of work 280,000 experience/community service National School Nutrition Schoolchildren Operated by the Department of 9.1 millionc Program Basic Education Source: World Bank 2021f. Note: Enrollment figures are for fiscal year 2018/19 unless otherwise noted. a. 2019/20 figures. b. Refers to full-time equivalent; more were reached with shorter-duration opportunities. c. 2014 figures. S O C IAL P R OTECTI O N paid to 33.1 percent of individuals, with the C O V E R A G E , TA R G E T I N G , result that an estimated 64 percent of the A N D B E N E F I T A D E Q UACY population lived in households with some- one in receipt of a social protection grant Social protection payments in South (World Bank 2021f). This is four times higher Africa cover large portions of the pop- than the average for Sub-Saharan Africa ulation and seem to be well targeted to and around one-third higher than the global the poorest households. The main social average. Figure 5.3 shows the distribution of protection grants are operated on a contin- the main social protection grants by quintile uous, on-demand basis and are therefore (World Bank 2021f). available to anyone who meets the cate- gorical and income eligibility requirements. Monthly social protection grant payments In 2014/15, social protection grants were represent a significant contribution to 122 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA F I G U R E 5 . 3   Distribution of social protection grants by wealth S O C IAL P R OTECTI O N quintile in South Africa, 2014/15 P AY M E N T A N D D E L I V E R Y SYST E M S 5.0 Older persons Disability Child support Although SASSA operates advanced admin- Millions of social protection grants 4.0 istration and payment systems for all regular social protection grant programs, last-mile delivery remains a challenge. Applications 3.0 for social grants take place in person at 360 local SASSA offices.2 The face-to-face grant 2.0 application process enables local social workers to ensure applicants apply for the 1.0 most appropriate grant(s) and clarify or check supporting documentation. To meet the eli- 0 gibility criteria for most grants, this includes Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 a national ID card and documents related Source: World Bank 2021f. to children, disability, residency, assets and income, and marital status, among others. All applications are registered on SOCPEN the incomes of the poorest households. (see Building Block 3). While registration for Figure 5.3 shows that the poorest 40 percent a grant can be done immediately, the appli- of the population receives over 62  per- cation process takes longer, as cross-checks cent of social protection cash transfers. are undertaken on supporting documenta- Although the aggregated average indi- tion. The latest SASSA annual report states cates social protection transfers represent that nearly 99 percent of grants in the pre- only 7.3 percent of individuals’ total expen- vious year were processed within 10 days diture, income inequality in South Africa (SASSA 2020). The applicants are informed masks exceptionally large variations in by letter about the status of their applica- income across all households. Based on tions, and, if accepted, they are added to the pretransfer incomes in 2014/15, social assis- SOCPEN pay lists for the social protection tance transfers accounted for an average programs for which they qualify. of 61 percent of household expenditure in quintile 1 and 32 percent in quintile 2 (World SASSA operates a highly automated payroll Bank 2021f). Various studies have exam- system, which makes it possible to pay mil- ined the poverty-reducing effects of South lions of grants quickly and in a timely manner Africa’s social protection grants. Most have into social protection beneficiary accounts concluded that the grants make a consid- every month. Social protection grants are erable contribution to poverty reduction paid by SASSA. Approximately one-third of (World Bank 2021f). Additionally, a recent percent) are made directly payments (29  analysis by the World Bank found that the into beneficiaries’ personal bank accounts, progressive nature of South Africa’s social with the remaining beneficiaries getting assistance system has had a substantial access to their grants through their SASSA impact on inequality, reducing the Gini coef- ficient by 10.5 percent (World Bank 2021f). 2 The recent COVID-19 grant is the exception. 5 :   S outh A frica 123 debit or Mastercard accounts. The issuance the full amount of their funds on payday. This and use of SASSA cards are facilitated by can overwhelm pay points, resulting in long the fact that over 90 percent of South Afri- lines and problems with liquidity. cans have a national ID. A unique 13-digit identification number is issued at birth reg- SASSA has developed comprehensive pro- istration, which makes issuance of national cedures and protocols for the operation of smart ID cards to individuals from the age of each of the social protection grants it admin- 16 very straightforward. This ID is also used isters. SASSA is an established public when enrolling children (and caregivers) for agency operating programs for millions of the Child Support Grant. ID cards contain South Africans and employing thousands of basic information and a photo but no bio- staff. Efficient delivery at this scale requires metric data. They significantly speed up the highly standardized systems and guidelines, registration and processing of social pro- which SASSA has in place. tection grant applications. They also reduce the incidence of duplication and fraud and EXP E R I E N C E I N ADAPTIVE enable individuals to open bank accounts. S O C IAL P R OTECTI O N The distribution of SASSA cards and bank P R O G RAM D E L I V E RY accounts opened by social protection grant recipients has contributed to a large increase Since the establishment of South Africa’s in the “banked” population—from 46 percent social protection system in 2004 until 2020, of adults in 2004 to 77 percent in 2016 (Fin- instances of responding to shocks with Mark Trust 2017). social protection have been limited and local- ized. The most frequent shocks or disasters Although payments to beneficiary accounts with the greatest impacts in South Africa and debit cards are highly digitized, bene- (outlined at the beginning of this chapter) ficiaries’ access to cash is hampered by a generally affect specific geographical areas, lack of point-of-sale payment points. While which leaves immediate responsibility for SASSA cards can be used as debit cards, response with the provincial or municipal automated teller machines (ATMs) and authorities. Since most of the social protec- retailers that accept debit cards are scarce tion systems—specifically those dedicated in townships and rural areas. Consequently, to grants—are administered federally by few poor South Africans use debit cards, as SASSA, the modification of national social the majority shop at local and small conve- protection grants for local disasters has nience stores. Known as Spaza shops, these received little attention. tend not to accept digital or debit payments because of the relatively high cost of having The only social protection grant explic- a point-of-sale system and the fees incurred itly designed to respond to shocks is the by traders on payments (approximately SRD grant. Local SASSA offices can accept 3.5 percent of transacted value). To retrieve applications for SRD grants from eligible the cash the shopowners prefer, most ben- individuals and households. While “distress” eficiaries without bank accounts have to normally refers to an idiosyncratic shock travel to the nearest Post Bank pay points, affecting an individual or household, such which may be some distance away. To min- as the death of the main income earner, the imize the cost of these trips in time and grants can also be used to assist house- money, the beneficiaries tend to withdraw holds affected by wider disasters, per the 124 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Disaster Management Act of 2002. Unlike these payments represent a high proportion other grants, the SRD is rarely provided as of consumption costs for the poorest. The cash. Instead, it is usually given in the form new COVID-19 SRD grants provide a much of food parcels or vouchers or other in-kind lower level of assistance: R 350 ($24) per assistance (such as school uniforms). SRD month. This is likely to represent a much assistance is normally approved for a max- lower proportion of consumption than the imum of three months but can be extended main grants and is far too low to compensate in exceptional circumstances. In response to for the loss in income experienced by many localized disaster events, such as flooding, as a result of lockdown measures. Providing it is usually a one-off distribution. A recent a more generous grant that more accurately flood disaster in Limpopo, for example, reflected the loss of income due to the crisis saw 78 families provided with food parcels, would probably have been financially infeasi- dignity packs, blankets, and other basic ble given the widespread nature of the crisis. amenities. According to the SASSA annual Most poor households would have benefited report, over 344,000 SRD applications were from a number of grants. Further research is approved in 2019/20. Only 2,315 of the ben- required to better understand the adequacy eficiaries were provided with cash, while of the overall increase in social protection most (194,397) were allocated food parcels grant payments in response to COVID-19. (SASSA 2020). The COVID-19 response was relatively In contrast to the years preceding, 2020 saw timely, given that it involved the development the most remarkable and unprecedented of a completely new grant payment program. efforts to adapt South Africa’s social pro- Planning for the grant started soon after tection system since its inception in 2004. the national lockdown on March 26, 2020. The scale and ambition of the social protec- Online applications opened on May  11, and tion response to COVID-19 was impressive over 116,000 grants were paid by the end (see box 5.1). It was also deemed essential, of the month. By June 18, 2020—less than given the scale of the crisis. Key elements million six weeks later—approximately 1.2  of the existing social protection system COVID-19 SRD grants had been paid. By were expanded both vertically and hori- November 2020, over 6 million people were zontally. The vertical expansion increased registered and had received the grant. Ini- mil- payments to almost all existing (17  tially, SASSA’s efforts were hampered by lion) social protection grant recipients limited capacity, poor and conflicting commu- and increased unemployment benefits to nication regarding required documentation, formal sector workers who lost jobs due to and a far greater number of applications the pandemic or lockdown. The horizontal than had been envisioned. Nonetheless, as expansion resulted in establishment of the the agency worked to address these chal- COVID-19 SRD, expansion of the existing lenges, systems for processing applications SRD grant to an additional 13,000 benefi- improved. It should be noted that resources ciaries, and establishment of the Temporary to cover the COVID-19 SRD grant were not Employer-Employee Relief Scheme (TERS). inexhaustible; determining the number of people who were eligible for the grant was The vertical expansion in response to tied closely to budget limitations. COVID-19 provided additional payments to caregivers of children. As highlighted above, 5 :   S outh A frica 125 P AY M E N T A N D D E L I V E R Y using this program to move to a system of SYST E M S D U R I N G S H O C K automated cash payments could be con- RESPONSE sidered, but it is unlikely to be financially feasible. The vertical expansion of payments to exist- ing beneficiaries during the COVID-19 crisis Individual insurance offers an alternative, pri- was made through SASSA payment accounts marily private, form of social protection, which as part of regular payments. The additional is already well developed for wealthier South distribution costs incurred through this Africans but could offer far greater protec- approach were minimal—one of the main tion to poorer households. Box 5.2 outlines advantages of an electronic payment system. the main areas in which microinsurance has In the horizontal expansion of payments to the potential to protect the poorest urban new recipients, most of the COVID-19 SRD and rural households from some of the most grants (over 4 million) were made though the prevalent natural disasters. Post Bank accounts with an accompanying debit card (the default option). The remain- der (1.7 million) were made directly into bank accounts. R E C O M M E N D AT I O N S F O R P R O G R A M D E S I G N AN D D E L IVE RY I N S O UTH AF R I CA The use of mobile payment systems in South South Africa’s social protection is mature and well developed in its Africa remains relatively low, despite over coverage, implementation capacity, and systems. While this facili- 95 percent of adults having mobile phones. tated an impressive response to the pandemic, South Africa’s prior In 2020, payments via mobile money experience in using social protection to respond to more frequent were piloted for the first time to 100,000 shocks was limited to occasional use of the SRD grant. Although COVID-19 SRD recipients, enabling them to disaster response capabilities exist at all levels, response tends to withdraw cash from an ATM without a debit be ad hoc and ex post. The following actions are recommended to card. The number of recipients reached improve it: through mobile payments was much lower than expected, primarily because of the low 1. Evaluate implementation of the COVID-19 SRD, capturing les- penetration of mobile money usage, par- sons learned and using them as a basis for establishing formal ticularly among the poorest wealth groups. contingency protocols for future, large-scale social protection Only 45 percent of recipients reported using shock responses through the regular SRD program. mobile money to purchase goods or services 2. In line with the recent amendments to the Social Assistance in 2019 (FinMark Trust 2020). Addition- Act, fully implement the legislative amendment to expand ally, most applicants did not have phones access to the SRD in response to disasters. or mobile money applications registered in 3. When doing so, pivot from a primarily in-kind intervention pro- their names. This difficulty highlighted the vided through local disaster relief teams to an SASSA-led, challenges of expanding social protection cash-based relief, building on the delivery innovations of the payment systems horizontally to populations COVID-19 SRD. with limited financial inclusion. Regular SRD shock-response assistance is provided by local disaster relief teams as in-kind assistance. Given the much smaller scale on which cash assistance is provided, 126 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA B OX 5.1   South Africa’s social protection response to COVID-19 South Africa’s social protection response to the Africa, vertical scaling to existing beneficiaries was COVID-19 pandemic included both vertical expansion relatively easy and limited only by the resources avail- and horizontal scaling. Each is detailed below. able. The top-up payments are detailed in table B5.1.1. Most social protection grants are well targeted to the Vertical expansion. Given the nature and comprehen- poorest. sive package of social protection programs in South TA B L E B 5 . 1 . 1   Adaptions of South Africa social assistance grants and programs to COVID-19 Program Beneficiary/age Enrolled COVID-19 adjustments Older Persons Persons 60 and above 3.6 million Grant increased by R 250 ($17)/beneficiary (May– Grant Oct. 2020) Child Support Grant Maximum of six children 18 12.7 million Payment increased by R 300 ($20)/child (May–Oct. or younger 2020) Caregivers also paid R 500 ($33)/month (June–Oct.) Disability Grant Disabled persons 18–59 1 million Grant increased by R 250 ($17)/beneficiary (May– Oct. 2020) Care Dependency Children under 18 154,735 Grant increased by R 250 ($17)/beneficiary (May– Grant Oct. 2020) Foster Care Grant Children under 18 355,609 Grant increased by R 250 ($17)/beneficiary (May– Oct. 2020) War Veterans Veterans 60 or older (or 62 Grant increased by R 250 ($17)/beneficiary (May– disabled) Oct. 2020) Grant-in-aid Adults 18 or older 273,922 Grant increased by R 250 ($17)/beneficiary (May– Oct. 2020) Social Relief of Individuals and fami- 344,482 Temporarily superseded by new COVID-19 SRD Distress lies experiencing undue grant (see below) hardships COVID-19 SRD Unemployed adults 18–59 6 million Monthly payment of R 350 ($24) for six months grant with no other social protec- tion assistance Expanded Public Working-age adults 267,000 Not expanded during COVID-19 lockdown but Works Program expansion planned as part of post-COVID stimulus Community Work Working-age adults 280,000 Not expanded during COVID-19 lockdown but Program expansion planned as part of post-COVID stimulus National School Schoolchildren 9.1 million Unclear how this continued during school closure Nutrition Program (2014 figures) period Temporary Working-age adults in formal 3.87 million Between R 3,500 and R 6,370 ($235 and $428)/ Employer-Employee employment employee/month Relief Scheme Source: Adapted from World Bank, “Where We Work” Botswana web page. 5 :   S outh A frica 127 B OX 5.1   South Africa’s social protection response to COVID-19 (continued) Horizontal scaling. Early in the crisis, it became clear the post-lockdown recovery. Thus, the Department that large numbers of individuals not normally targeted of Employment and Labor issued Notice 240/2020 by the regular suite of social protection programs were to create the Temporary Employee-Employer Relief affected by the lockdown and not eligible for exist- Scheme (TERS). ing social protection grants. In particular, the crisis affected informal sector workers, who suffered an The TERS scheme paid between R 3,500 and R 6,370 almost complete loss of income during the lockdown. ($235 and $428) per employee per month to firms to Many were not co-residents with other grant recipients support furloughed employees during the first three (Bhorat et al. 2020). It also became apparent that the months of the lockdown. The program was imple- government, which does not capture data on informal mented under and financed through Unemployment workers, had no real way of identifying these individu- Insurance Fund systems. Firms applied for the funds als through any of its existing databases. each month through an online portal, received the grants through direct deposit to bank accounts, and Consequently, the government took the bold step of were required to make the furlough payments to their reinventing the SRD grant in response to the COVID-19 employees. Firms that received the payments were not crisis and establishing a customized ASP program to permitted to dismiss their workers. Workers who were address the gaps. SASSA led the development of the dismissed from their jobs were eligible to receive the COVID-19 SRD grant, which was established and oper- standard Unemployment Insurance Fund payment, ational in just over one month. The grant provided which paid a lower rate than the TERS for low-income R 350 ($24) per month for six months to those who workers. met the eligibility criteria: working-age adults (18–59 years); not in formal employment; not paying unem- The TERS was designed, piloted, and implemented ployment insurance fund deductions; not receiving during the COVID crisis. In April 2020 alone, an esti- an old age grant, disability grant, or grant-in-aid. By mated 351,000 firms received TERS payments November 2020, over 6 million people were registered million employees. The program on behalf of 3.87  and being paid through the COVID-19 SRD grant. improved on the Unemployment Insurance Fund system in several ways. First, it instituted new protocols to con- South Africa’s Unemployment Insurance Fund is a trol for duplicate and ineligible applications. It included contributory social protection program that offers a transparency mechanism—all firm recipients were unemployment benefits to those who have made con- named through a website—to provide workers with tributions while working. It therefore primarily covers information to monitor their receipt of funds. Finally, workers in the formal sector. During the COVID-19 it developed grievance procedures, including the use lockdown and associated economic downturn, the gov- of independent investigators to follow up on employee ernment feared significant firm closures and a potential and employer complaints. An additional 650,000 work- loss of 8.4 million jobs—a risk it was keen to mitigate by ers who were dismissed from their jobs received the supporting firms in furloughing (rather than laying off) standard unemployment insurance payment, equiva- workers to avoid mass unemployment and facilitate lent in total to R 3 billion ($201 million). 128 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA B OX 5.2   Microinsurance as a shock-responsive social protection option Microinsurance protects low-income individuals disaster-linked property microinsurance is now avail- against specific perils in exchange for regular premium able, the microinsurance market is dominated by life payments proportionate to the likelihood and cost of insurance and funeral coverage. There is a case for the risks involved. In countries with more developed piloting and carrying out further cost–benefit analysis of insurance markets, such as South Africa, microinsur- subsidized property microinsurance schemes in highly ance has the potential to play a key role in covering the disaster-affected, low-income urban settlements. Such losses incurred by poor households from natural disas- schemes would need to incorporate targeted financial ters, thereby increasing their resilience and reducing education elements to help people understand these the burden on the state to respond. forms of insurance. Microinsurance could be extremely valuable for Agricultural insurance also is well developed in South low-income residents in informal settlements, who are Africa but is focused on high-turnover, commercial-scale particularly vulnerable to damage from extreme weather farmers. A wide range of crop, livestock, and forestry events and fire. Currently, almost all these households insurance products is available, but coverage is low. are completely uninsured. This is a reflection (1) of a lack The most widespread product, multi-peril crop insur- of affordable insurance products and (2) of low interest ance, has even seen coverage decline significantly in in or understanding of insurance products. Only one-fifth recent years. This is a result of expensive premiums of South Africans have any form of short-term insurance driven by unsustainably high claims—driven in turn coverage, the majority of which relates to automobiles by the increasing frequency and intensity of climate or mobile phones. Studies on financial literacy show a events, particularly drought, excessive rainfall, and large variation in insurance awareness and adoption flooding. Extremely few smallholder farmers buy agri- by educational attainment, wealth, and province (Rob- cultural insurance, which leaves them highly exposed erts et al. 2018). A regulatory framework introduced in to natural disaster risk. Expanding the coverage of 2017 paved the way to expand more affordable micro- these schemes to smallholder farmers could fill signifi- insurance products to the lower-income market (that cant gaps in their income and consumption in the face is, households in the poorest seven deciles). Although of increasingly severe drought-induced losses. BUILDING BLOCK 3: regular social protection assistance to the D ATA A N D poorest households and referrals for post- disaster support (for example, the recently I N F O R M AT I O N S Y S T E M S announced post-COVID Presidential Employ- South Africa has a highly advanced elec- ment Stimulus package; GOSA 2020). They tronic registration and payment MIS in place would also facilitate identification of those for its social protection programs and recip- households vulnerable to disaster risk or ients. While SOCPEN is sophisticated and that may require immediate assistance when comprehensive, the COVID-19 crisis has disasters strike. The proposed development highlighted a need for greater interop- of NISPIS is a welcome step in this regard. erability among government databases. Data-sharing protocols would facilitate a The links between social protection and data more coordinated approach for providing on risk, vulnerability, and disaster are weak 5 :   S outh A frica 129 and reflect the limited role ASP plays in TA B L E 5 . 5   Scorecard for data and information systems in disaster response in South Africa. Currently, South Africa multiple early warning systems are operat- Indicator Score ing in South Africa, focused on forecasting and monitoring weather and fires. National Composite country rating early warning systems offer little analysis of Foundational indicators household vulnerability and exposure to the Data and information on disaster risk and household vulnerability most common disasters in terms of impact on household livelihoods and consump- Beneficiary registry ● tion or food security. Although government Social registry departments and agencies at all levels are Adaptive indicators required to undertake and update risk and Use of preexisting data for shock response ● vulnerability assessments as part of DRM or climate change initiatives, it is not clear Post-shock household needs assessments ● that they attempt to assess (and quantify) Data-sharing protocols with internal/external partners the household-level impacts of disasters or Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. support planning for appropriate responses, including ASP. Following are detailed findings and rat- the Department of Home Affairs ID system ings justifications for this building block and a number of other government data- (table 5.5). bases, such as those of the government’s payroll system, the Unemployment Insur- B E N E F I C I A R Y R E G I S T R AT I O N ance Fund, the National Treasury, and the SYST E M S Department of Basic Education (a database of students). It does not, however, interface SASSA has developed a comprehensive dig- with the South African Revenue Service itized beneficiary registration and payment system database, which contains income system, SOCPEN. This MIS is used to information. As most grants are means tested, cross-referencing income and asset ● Process applications for the old age, dis- information on applications with revenue ability, war veterans, child support, foster system data can take longer. care, and care dependency grants; ● Generate monthly payrolls for the In addition to SOCPEN, the Department for approximately 17 million grants; and Social Development operates the Social ● Automatically produce lists of beneficia- Development Integrated Case Manage- ries due for reassessment. ment System. This database comprises all households with access to other social ser- The latest SASSA annual report states that vices, such as the Child Protection Register; nearly 99 percent of grants in 2019/20 were probation case management; victim empow- processed within 10 days. Once accepted, erment; and adoptions and the Register of applicants are added to the SOCPEN pay Adoptable Children and Prospective Adop- lists for the respective social protection pro- tive Parents. grams. The SOCPEN is interoperable with 130 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Relatively good protections are in place to and beneficiaries but also those who receive ensure the privacy and security of beneficiary other kinds of social assistance. The latter data. South Africa’s Protection of Personal might be those participating in the Expanded Information Act, which became effective July Public Works Program or other programs in 1, 2020, provides a constitutional right to pri- the recent Presidential Employment Stimu- vacy through the safeguarding of personal lus package (GOSA 2020). The new system information and stipulates the ways in which could also play an important role in referring personal information may be processed social grant recipients to the employment and shared. While this means individuals’ and work experience services provided by data in government MISs are commendably the Department of Employment and Labor well protected, it does create challenges to and the Presidential Youth Employment facilitating data exchange and interopera- Interventions. bility between government departments. In processing the COVID-19 SRD grants, for P O S T- S H O C K example, SOCPEN data could not be used R E G I S T R AT I O N / to cross-reference the same individuals’ tax M A N A G E M E N T I N F O R M AT I O N data on the South Africa Revenue Service SYST E M MIS without high-level approval and mem- orandums of understanding in place. The The COVID-19 SRD grant was the first program lessons learned can be applied to develop- to use online registration, which was seen as ing NISPIS. the best way to maintain social distancing given public health concerns. Although effi- cient, online application presents a barrier to S O C IAL R E G I ST RY access by many of the poorest who are dig- South Africa lacks a fully functioning social itally excluded (only 64 percent of adults in registry. This presented a challenge during South Africa have smartphones). To address the COVID-19 crisis when millions of citi- this, social workers with phones and/or tab- zens were required to register to receive the lets were dispatched to communities in some COVID-19 SRD grant. The National Social areas to help people prepare and submit Development Strategic Plan (2015–20) pro- their electronic COVID-19 grant applications. posed the development of NISPIS, however. This demonstrates the importance of main- Efforts are now under way to link SOCPEN taining alternative application processes for and the Social Development Integrated Case the more marginalized and digitally excluded Management System with the databases of groups. the Departments of Basic Education, Health, and Home Affairs. The longer-term aim is While some interoperability occurs between to also link to the databases of the Depart- SASSA and other government departments, ments of Rural Development, Employment systems and protocols for sharing informa- and Labor, Public Works and Infrastructure, tion and for wider cooperation are not clearly Transport, and the Cooperative Governance defined. Once a COVID-19 SRD grant appli- and Traditional Affairs. Provinces and dis- cation is submitted, eligibility is determined tricts would have access to NISPIS. by cross-referencing multiple govern- ment databases. These include SASSA’s NISPIS has the potential to lead to better SOCPEN and databases of the Unemploy- tracking, not just of social grant recipients ment Insurance Fund, the South African 5 :   S outh A frica 131 Revenue Service, the Government Employ- There are multiple early warning systems ees Pensions Fund, and the National and organizations involved in forecasting and Student Financial Aid Scheme. In developing monitoring climatic events in South Africa. the grant, SASSA set a target of seven days Technically, responsibility for early warning from application to payment. This was overly in the country lies with the National Disas- ambitious, and initial applications took much ter Management Centre, which, according longer for a variety of reasons. The process to the National Disaster Management has facilitated thinking on data-sharing pro- Framework, “is responsible for ensuring the tocols among key government departments. technical identification and monitoring of hazards and facilitating the development of SASSA and the Department of Social Devel- standard early warnings by national organs opment have invested significantly in the of state” (GOSA 2005). Table 5.6 lists some new electronic application and management of the systems in place to track different cli- system for the COVID-19 SRD grant. Although matic events and disasters. Generally, the the grant is temporary, the investments made Department of Agriculture and agricultural in the electronic application system and reg- research agencies tend to focus on drought istration of 6 million new payees will likely monitoring, while the Department of Science prove invaluable in the future. Additionally, and Innovation maintains the South African these beneficiary lists will be useful should Risk and Vulnerability Atlas, and the South further scale-ups be required or to support African Weather Service provides general post-COVID recovery programs—such as the weather warnings. rollout of wage payments to around 800,000 temporary workers proposed as part of the In general, the national (and local) disas- Presidential Employment Stimulus package. ter management centers tend to focus on disaster-event tracking followed by assess- ment, rather than on systems that predict or D ATA A N D I N F O R M AT I O N O N preempt disasters. The current approach R I S K , V U L N E R A B I L I T Y, A N D to multihazard early warning systems has DISASTERS weaknesses that go beyond the lack of The focus on climate change, reinforced by coordination. In particular, processes to the 2018 National Climate Change Adaptation disseminate information are not well devel- Strategy, has renewed awareness of disas- oped, which means adoption by users is ter risk. As mentioned in Building Block 1, limited. Tools and mechanisms by which legislation is in place that has required gov- forecasts or warnings translate into practi- ernment structures at all levels to put in cal information or actions for end users are place disaster management plans and adap- also limited, as is the use of methodologies tation strategies that address disaster risk. to assess household-level vulnerability and The climate change strategy reinforces the risk in terms of welfare and consumption. importance of undertaking risk and vulner- ability assessments as a first step. In reality, Renewed impetus to improve early warn- such disaster management plans are not ing systems has been driven by climate comprehensively in place. Where they exist, change legislation. The Department of Envi- they focus on identifying disaster risks, map- ronmental Affairs coordinates a high-level, ping geographical areas at risk, and planning cross-government committee on climate contingency responses. change policy, and the National Climate 132 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA TA B L E 5 . 6   Early warning systems in place for most frequent shocks in South Africa Hazard Early warning Decision support system Flood ■ Severe Weather Forecasts Demonstration project ■ South African Flash Flood Guidance System South African Weather Service ■ South Africa River Flow Flood Forecasting System ■ Seasonal forecasts Fire National Fire Danger Rating System Advanced Fire Information System severe weather alerts Drought Umlindi ■ Drought Monitoring Desk at South African Weather Services ■ Seasonal forecasts Storm surge National Storm Surge Early Warning System ■ South African Navy Hydrographic Office ■ Council for Scientific and Industrial Research for the National Ports Authority ■ South African Weather Service Marine Office Source: CSIR 2014. Change Adaptation Strategy (2019) proposes events affecting South Africa (outlined at the creation of an improved and coordinated the beginning of this chapter), current DRM national and provincial early warning system. and climate change policies make little or no This will involve the National Disaster Man- mention of ASP interventions in response agement Centre, the South African Weather to the most common disasters. It would Service, the National Space Agency, the take some coordinated planning to design a Department of Environmental Affairs, and system that enabled households affected by the Department of Health, among others. The the most frequent and localized disasters to aim is to create an system that offers accurate be paid through the national SASSA system. climate information, such as historical trend The process for identification and targeting data, seasonal predictions, and projections for highly localized shocks, such as floods, and early warning of extreme weather and which require short-term and immediate other climate-related events, as provided by assistance would have to be very efficient. existing and new monitoring and forecasting Social protection responses paid via SASSA facilities and networks. Plans for the system to longer-term shocks, such as prolonged include maintenance and enhancement; its drought, do not appear to have been con- intended end use is to inform adaptation sidered, however. This is interesting in light planning and disaster risk reduction. of the immediate and significant social pro- tection response identified in the COVID-19 crisis. USING DISASTER RISK M A N A G E M E N T I N F O R M AT I O N Postdisaster assessments, led by the disas- SYST E M S TO I N F O R M ter management centers, may identify ADAPTIVE S O C IAL emergency relief for affected households. P R OTECTI O N R E S P O N S E Postdisaster assessments form a key part The DRM data and postdisaster assess- of DRM information systems, given the ments of early warning systems are not fast-onset nature of many climate crises linked to ASP systems in South Africa. (particularly floods and storms). The National Despite the increasing frequency of disaster Disaster Management Centre has developed 5 :   S outh A frica 133 significant guidance on undertaking these to provide temporary emergency assistance assessments, which concentrate primar- to locally affected households, but no coor- ily on mapping the physical damage caused dination or linkage with SASSA is apparent by disasters to infrastructure and property. as part of this process. Furthermore, it is The mapping of human impact and response not apparent if the practice of mapping focuses mainly on the provision of immedi- household impact and response is clearly ate in-kind assistance, such as food parcels or systematically included in guidelines and or temporary shelter. As highlighted, author- protocols for postdisaster assessment and ities may identify the need for SRD grants response at local or higher levels. R E C O M M E N D AT I O N S F O R D ATA A N D I N F O R M AT I O N S Y S T E M S I N S O U T H AFRICA Although South Africa has a highly advanced beneficiary registration and payment MIS, gaps in interoperability with wider government MISs mean no social registry is currently in place. The following recommendations focus on this lack of coordination between and across social protection and DRM departments and agencies: 1. Accelerate the development of the proposed NISPIS, which can become a good foundation for informing future shock responses. NISPIS should enable interoperability of SASSA, the Department of Social Development, and other key government departments delivering social protection programs. Beyond creating day-to-day delivery efficiencies, this would also enhance the ability of these departments to respond to shocks using a more uni- fied and comprehensive database of current and potential beneficiaries. Linking NISPIS to the databases of the Departments of Rural Development, Employment and Labor, Public Works and Infrastructure, Cooperative Gov- ernance and Traditional Affairs, and of the Revenue Service would further facilitate selection, verification, and payment processes for the expansion of temporary social protection assistance to individuals not currently eli- gible for regular social protection grants. NISPIS should also be made accessible to relevant departments at the provincial and municipal levels to support localized disaster response. 2. Investigate the feasibility of institutionalizing the innovative electronic application and management system cre- ated for the COVID-19 SRD grant. 3. Appraise the ways in which national (and international) disaster risk and early warning information and house- hold vulnerability data can be used to estimate and forecast the impact of disasters on poor households in South Africa, informing any future SASSA-led responses to natural disasters. 134 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA BUILDING BLOCK 4: immediate response may vary depending on the shock and the financial resources of the FINANCE responding agency and/or local authority. South Africa’s national expenditure on social The COVID-19 crisis has been a significant protection is high by international standards exception that highlights the role of microin- and comprises a regular and key element of surance in reducing the disaster risk of poor annual government spending. While this is to households. be commended, financing of ASP responses is less developed. Several pieces of national Utilization of existing risk and vulnerability legislation encourage government depart- mapping or early warning tools to inform or ments and authorities to pre-plan the model the costs of disasters to poor house- allocation of resources for disaster response holds in terms of consumption and welfare is as well as undertake contingency planning limited. Given the disconnect between DRM and invest in other DRM preparedness and and ASP, there has been little examination mitigation measures. None of this legisla- of its financing. Recent work on agricultural tion explicitly relates to social protection insurance, however, is a promising start. responses or interventions, however. This In undertaking a review of ASP (as recom- reflects the fundamental disconnect, high- mended above) the costs of, and options for, lighted in the Building Block 1 discussion, financing any identified response must also between DRM and ASP in South Africa. be examined to inform decision making. The cost of the ASP programming implemented The quantification of disaster costs in in response to COVID-19 will have brought South Africa is almost exclusively under- attention to disaster risk financing generally taken post-shock and tends to focus on the and to this aspect in particular. impacts on property and infrastructure. The impacts of shocks on households are usually Following are detailed findings and rat- identified as part of immediate postdisaster ings justifications for this building block needs assessment, which focuses on emer- (table 5.7). gency relief. The resources available for SPENDING ON SOCIAL P R OTECTI O N TA B L E 5 . 7   Scorecard for finance in South Africa South Africa is one of the biggest spenders Indicator Score globally on social assistance as measured Composite country rating as a percentage of GDP. Government expenditure on social assistance programs Foundational indicators represents 3.3 percent of GDP and 16.2 per- Social protection spending ● cent of total government expenditure (World Disaster risk financing policies and strategies ● Bank 2021f). This is the fourth highest per- Adaptive indicators centage in Sub-Saharan Africa and the tenth highest in the world. South Africa spends Quantification of the expected costs for social protection responses ● more proportionately than many wealthier middle-income countries, including Argen- Disaster risk financing instruments linked to social protection ● tina (2.1 percent), Mexico (1.7 percent), and Note: ● = nascent; ● = emerging; ● = established. Mixed colors imply the rating is between two stages of development. percent)—an the Russian Federation (1.9  5 :   S outh A frica 135 impressive accomplishment, given that slug- budgets of four municipalities to higher levels gish economic growth in recent years has of government, as they failed to implement flattened tax revenues and reduced fiscal infrastructural projects and demonstrated capacity for government expenditures. Most poor governance (GOSA 2017b). of the spending goes to the Older Persons Grant (51 percent) and the child support and Two grants are designed to support local associated grants (37 percent). government postdisaster expenditures: the Municipal Disaster Relief Grant and the Municipal Disaster Recovery Grant. The total DISASTER RISK FINANCING amount paid out to local communities under P O L I C I E S A N D S T R AT E G I E S these grants in 2019/20 was R 335 million The Disaster Management Act of 2002, ($22.8 million) and R 194 million ($13.2 mil- amended in 2015, is the main act referring lion), respectively. Both grants are allocated specifically to DRM and risk financing in based on strict criteria, and municipalities South Africa. Section 52 of the act requires must fulfill specific conditions to qualify. The all levels of government, including municipal- Disaster Relief Grant is designed to enable ities, to prepare contingency strategies that a timely response to immediate needs after include pre-planned measures to finance a a disaster has occurred; it finances such postdisaster response. The aim is to encour- expenditures as emergency repair of criti- age lower levels of government to invest cal infrastructure and emergency provision in disaster risk reduction to reduce disas- of goods and services. Although envisioned ter liability. These strategies have failed to as rapid-response instruments, the bureau- materialize in practice, with DRM expendi- cracy associated with both grants acts as a tures mostly taking place post-shock. disincentive to application. Consequently, the finance available for disaster response The Public Finance Management Act No. 1 of usually depends on the capacity of differ- 1999 governs how public money is spent to ent municipalities to raise resources. Large manage disaster response in South Africa. metropolitan authorities may have sufficient This act permits national and provincial reserves or other receipts they can use or governments to authorize extrabudgetary reallocate to support response. Some can expenditures in response to disasters up to also obtain loan finance, particularly for a threshold limit of 2 percent of their annual reconstruction projects. budgets. The Municipal Finance Manage- ment Act (2003) gives municipalities the Q U A N T I F I C AT I O N O F P O S T - power to authorize extrabudgetary spending S H O C K ADAPTIVE S O C IAL beyond this level. In reality, municipalities are P R OTECTI O N C OSTS quite limited in how far they can reallocate revenues received from the national govern- The quantification of disaster costs in South ment in the form of conditional grants. The Africa is almost exclusively undertaken Municipal Finance Management Act also post-shock. Disaster costs are generally gives national and provincial governments incurred and quantified in several phases. the authority to stop allocations to munici- The first phase relates to the costs of rescue palities and transfer disaster response funds and relief operations associated with imme- to higher levels of government. In 2017, the diate disaster and humanitarian response Minister of Finance transferred the allocated and may include helicopter airlifts, food 136 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA baskets, soup kitchens, tents, and ablu- agricultural losses estimated at R 5.9 billion tion facilities, among others. The needs ($40 million; World Bank 2021e). Such anal- for these types of assistance are assessed yses also noted that the drought triggered and mobilized very quickly. At the municipal a loss of 30,000 agricultural jobs, but it is level, disaster management and emergency not clear how this translated into individual control units are responsible for coordi- losses or reductions in consumption at the nating immediate relief efforts. While they household level. have some operating budget, it is usu- ally far less than required. The immediate In short, assessment of disaster costs in costs of responding to major floods in eThe- South Africa is rarely conducted from a kwini metropolitan municipality in 2019, social protection perspective. Support to for example, were estimated at R  47  mil- individuals and households focuses on lion ($3.2 million), but the operating budget immediate and short-term relief assistance. of the local disaster management unit was Some of this—for example, food parcels and only R 7 million ($0.5 million). A wider pro- vouchers and temporary accommodations— vincial assessment estimated the cost of could be deemed “social protection,” with repairing or replacing damaged infrastruc- the amounts and durations locally defined ture and government services at R 1.8 billion based on the nature of the disaster and the ($125  million) and included only loss and resources available. No ex ante aggregated damage to government assets and property. analysis seems to be conducted of individ- ual- or household-level impacts or the costs The second phase involves much larger of disasters for affected populations. assessments of damage to property and infrastructure, particularly government ser- The COVID-19 response has proved a major vices. These assessments can take months and significant exception to this lack of atten- or years to complete, as they are subject to tion to social protection and appears to be repeated review and negotiation as munic- the only disaster event for which response ipalities struggle to identify resources, has been based on a recognition of eco- relying heavily on reallocations from line nomic losses of and impacts on individuals department budgets. Although private insur- and households. This desk review has not ance mechanisms do cover a significant been able to establish what financial analy- share of disaster loss in South Africa (dis- sis or modeling was used to set the final level cussed in the next section), settling claims for a significant ASP response. is not always straightforward. DISASTER RISK FINANCING The costs of some disasters have been MECHANISMS FOR SOCIAL assessed from a macroeconomic perspec- P R OTECTI O N tive to quantify the impact on government Pre-agreed disaster risk financing instru- tax and other receipts and the drops in GDP ments in place for South Africa are limited, associated with commercial or agricultural and none is specifically designed to finance a losses. The macroeconomic analysis of social protection response. Nationally, South the 2018 drought’s impact on the Western Africa maintains a contingency reserve Cape, for example, estimated a 25 percent fund that is allocated R 5 billion ($341 mil- reduction in export volumes of key prod- lion) annually. The fund can be used for ucts (such as fruit and wine), with total 5 :   S outh A frica 137 immediate disaster response without having Property insurance is the most widely avail- to reallocate other funds or resort to borrow- able form of insurance that covers natural ing. Notably, the fund is not earmarked for disaster risk, including earthquake, strong natural disasters and is frequently used to winds, flood, hailstorms, landslide, and sub- support other “emergencies”—typically the sistence. As discussed in box 5.2, it could bailouts of parastatals. In recent years, this provide valuable protection to poor house- has included payments to South African Air- holds at high risk of damage and loss from ways and the public broadcasting agency. disasters. As no regulatory requirements exist for property owners (including land- The national government also relies on some lords) to have natural disaster insurance, pre-agreed, postdisaster borrowing. South only 4 percent of the population is estimated Africa has access to and has used the Inter- to have property insurance. The bulk of nat- national Monetary Fund’s Rapid Financing ural disaster risk coverage is on residential Instrument to respond to the COVID-19 property in the wealthiest parts of the coun- crisis. In 2020, upon approval from the Inter- try and the upper-income neighborhoods national Monetary Fund, South Africa drew within them. Residential property accounts against its full allocated amount of $4.3 bil- for 56 percent of the total insured value, with lion—the largest single disbursement by the more than half of natural disaster coverage International Monetary Fund to a country located in Gauteng Province. Similarly, in affected by the pandemic. In addition, the rural areas, agricultural insurance products African Development Bank issued a loan have been pitched to the needs of larger, of $288 million. The government of South commercial farmers, leaving the far more Africa has not yet reached any prefinancing numerous smallholder farmers excluded. agreements with other international lenders, such as the World Bank. Pilots and plans are in development to expand access and coverage of microin- Insurance represents a major disaster risk surance for natural disaster risk. The World financing option in South Africa, with sig- Bank is supporting the government of South nificant potential to provide greater social Africa in fully exploring the development protection support. South Africa has the of two insurance programs for smallholder largest and most mature insurance market and medium-scale farmers. Specifically, in Africa, with total gross written premi- this exploration pertains to Area Yield Index ums of $47.24 billion in 2018. Life insurance Insurance and Satellite-Based Pasture dominates the insurance market, but at Drought Index Insurance for subsistence and least 92 other insurance products are avail- livestock producers that rely on rangelands. able, including those provided by three These index-based insurance solutions work state-owned insurers. Nonlife or short-term by providing protection to farmers from insurance products primarily include prop- losses against an average area yield index erty, auto, personal accident, and health in a defined geographical area, rather than insurance. The nonlife insurance market is paying out against yield losses in individ- well capitalized and has sufficient capacity ual farmers’ fields. This approach makes the to offer financial protection against natural products much more affordable. Such pro- disasters. grams normally involve the collaboration of private sector insurance partners to share risks and responsibilities. 138 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Financial management systems for on the fiduciary standards of the local body post-shock assistance at the household responsible for managing relief and dis- level depend largely on the nature of the pro- tribution. By contrast, expansion of social vider. Direct support to affected households protection grants as part of the COVID-19 is currently ad hoc and varied, depending on response was administered by SASSA and the nature of the crisis and the resources of the grants issued through SOCPEN into indi- the responding district or municipal author- vidual payment accounts. This minimized ity. To account for post-shock assistance in opportunities for payments to be diverted cases of localized crisis, financial manage- and enabled automatic reconciliation. ment systems thus are likely to vary based R E C O M M E N D AT I O N S F O R F I N A N C E I N S O U T H A F R I C A The government of South Africa’s expenditure on social protection reflects the high political and fiscal priority placed on expenditure in this sector. Since the government has not prioritized disaster risk financing efforts to the same degree, ex ante consideration of the economic and wider costs of disasters has been limited—and almost none has come from a social protection perspective. This means pre-agreed financing instruments to reduce the disaster risk of the poorest households, particularly in response to localized disasters, are also limited. The following recommen- dations are intended to strengthen finance for ASP in South Africa: 1. Undertake further analytical work to model and quantify ex ante the costs of disasters in South Africa. This anal- ysis should specifically address consumption, asset loss, and the financial resilience of poor households and can comprise an element of the review of ASP recommended under Building Block 2. Such an analysis should be used to inform the nature of social protection responses that would most effectively build the disaster resilience of poor households. Modeling should include a range of the most likely potential disasters and consider how the costs of response and impact vary in relation to the magnitude and frequency of each. 2. Develop a national disaster risk financing strategy that outlines a process to expand existing disaster risk financ- ing instruments. Ensure that some of these instruments are designed to finance social protection responses. 3. Improve systems for the release of disaster response financing to lower-level authorities for social protection (and all) interventions. Consider how standard protocols covering triggers, payment amounts, frequency, and coverage could be used to overcome accountability and financial management concerns. 4. Support public–private partnerships (with subsidies as appropriate) for piloting microinsurance programs for poor households in informal urban settlements. For the most frequent natural disasters in urban areas, micro- insurance against loss of property or other assets may be more efficient and have more impact than temporary social protection payment programs. 5. Expand public–private pilots in rural areas for providing agricultural insurance to smallholder and medium-scale farmers. Such pilots provide an opportunity to examine the role and impact of such programs in the adoption of better climate change adaptation strategies versus other ASP interventions. 6. Develop interventions to increase the financial literacy of poor households on the value and benefits of microinsurance. 5 :   S outh A frica 139 PRIORITY The most strategic areas for investment are related to institutional capacity and coor- INTERVENTIONS dination, delivery of cash payments, and FOR SOUTH development of disaster risk financing mech- anisms. Policy reflections of note include the AFRICA following: This analysis has shown that, overall, ASP is at an emerging stage in South Africa, reflect- ● Address the disconnect between social ing a mixture of well-established dimensions protection and DRM. Despite good pol- and significant gaps. The traffic light score- icies being in place for both social card highlights areas where South Africa’s protection and DRM as well as strong social protection and DRM systems are institutions with national coverage, both strong and well established. Robust and rel- sectors effectively operate in siloes with evant policy and legislative frameworks are no overlap. The COVID-19 response in place in both sectors, and South Afri- demonstrated the critical role appro- ca’s social protection system is mature and priately designed social protection well funded. The main government agency, responses can play in mitigating the SASSA, has strong institutional capacity, impacts of disasters. As the pandemic including advanced management informa- crisis subsides, this is the ideal opportu- tion and payment systems to deliver social nity to establish a high-level committee assistance grants to large portions of the to lead in strengthening the role of social poorest populations. These strengths pro- protection in shock response as part vided a valuable foundation from which of national and local DRM and climate to launch an impressive COVID-19 ASP change adaptation policy, strategy, response, which has also provided useful implementation, and financing. lessons. ● Develop appropriate ASP programs and interventions that most effectively South Africa has a relatively strong DRM reduce the disaster risk of poor house- capacity, with disaster management centers holds. Heavy reliance on categorical located across the country at provincial and targeting means large swaths of the pop- municipal levels. This is necessary, given the ulation are not covered by existing social localized nature of the most frequent disas- protection systems and remain vulnera- ters. A range of well-established national ble to disasters. In light of the COVID-19 early warning systems are in place to monitor response, all existing social protection the primary climatic disasters, particularly programs should be examined to assess floods, storms, and wildfires. The emer- their potential to adapt (vertically and gence of climate change adaptation policies horizontally) in the face of the suite of and efforts in recent years has placed a disasters more regularly affecting South renewed focus on DRM and highlighted the Africa. The SRD grant should be devel- need for greater collaboration and coordina- oped into a specific shock-responsive tion among government agencies. social protection grant that can be uti- lized in the face of both national and Table 5.8 summarizes the priority actions for localized disasters. Clear protocols and both foundational social protection and ASP. 140 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA guidelines are required to support SRD ● Expand disaster risk financing for ASP. implementation. The disaster risk financing instruments ● Improve data collection and analysis in place to finance disaster response of disaster risk and household vulner- are limited, and even fewer are avail- ability. The disconnect between social able for social protection responses. The protection and DRM means capabil- COVID-19 crisis has shown the value of ity is limited to collect and analyze early social protection as a response but has warning, disaster risk, and vulnerability also demonstrated its cost. The urgency data and information to develop a better and political will are now greater to understanding of the impact of disasters develop a national disaster risk financ- on poor households in South Africa— ing strategy, which must make provision which in turn is needed to understand for instruments designed to finance which social protection interventions are social protection responses. Significant most appropriate for which shocks; the potential exists to expand the provision coverage, scale, and duration of ASP of appropriate microinsurance products required; and the short- and long-term for the rural and urban poor that would impacts on poverty and disaster resil- reduce the impact of and loss from disas- ience. Cross-sectoral coordination and ters—in particular, index-based products collaboration are required to identify that pay out automatically in response which departments or agencies should to climatic triggers for shocks, such as be responsible for data collection and drought. analysis. 5 :   S outh A frica 141 TA B L E 5 . 8   Summary of all recommendations by building block for South Africa Building block Foundational social protection Adaptive social protection ■ Establish a high-level working committee to review ■ Integrate ASP provisions into DRM policies and Institutional and enhance the role of social protection in DRM plans arrangements ■ Integrate DRM provisions into social protection ■ Formally integrate SASSA and the role of social and partnerships policies and plans protection programs into DRM coordination mecha- nisms at national and local levels Evaluate implementation of the COVID-19 SRD, ■ In line with recent amendments to the Social capturing lessons learned and using them as a basis Assistance Act, fully implement the legisla- for establishing formal contingency protocols for tive amendment to expand access to the SRD in Program design future, large-scale social protection shock responses response to disasters and delivery through the regular SRD program ■ Pivot from a primarily in-kind intervention provided through local disaster relief teams to an SASSA-led, cash-based relief, building on the delivery innovations of the COVID-19 SRD ■ Accelerate efforts to develop the proposed NISPIS ■ Have NISPIS link SOCPEN with databases of the Departments of Rural Development, Employment ■ Ensure NISPIS is interoperable with SASSA, the and Labor, Public Works and Infrastructure, Coop- Department of Social Development, and other key erative Governance and Traditional Affairs, and government departments delivering social protec- the Revenue Service to form a core foundation for tion programs to increase their ability to adapt in future social protection shock response response to shocks ■ Investigate the feasibility of institutionalizing the ■ Ensure NISPIS will be accessible at the provincial Data and infor- new, innovative electronic application and manage- and municipal levels mation systems ment system for a revised SRD grant ■ Improve analysis of early warning, disaster risk, and vulnerability data and information to gain a better understanding of the impact of disasters on poor households in South Africa ■ Clarify which departments or agencies should be responsible for data collection and analysis to inform ASP responses ■ Improve systems for the release of disaster ■ Undertake further analytical work to model and response financing to lower-level authorities for quantify the ex ante costs of disasters in South social protection (and all other) interventions Africa associated with the consumption, welfare, and financial resilience of poor households ■ Support pilots (with subsidies as appropriate) of microinsurance programs for poor households in ■ Develop a national disaster risk financing strategy Finance informal urban settlements that outlines a process to expand the pre-agreed disaster risk financing instruments in place for ■ Expand pilots providing agricultural insurance to South Africa, ensuring that some are designed to smallholder and medium-scale farmers finance social protection responses ■ Develop interventions to increase the financial lit- eracy of poor households on the value and benefits of microinsurance APPENDIX Dimension Nascent Emerging Established Institutional arrangements and partnerships SP policy and legislative frameworks: No national SP policy or National SP policy or SP policy and strategy Examines existence and relevance of strategy strategy in place but up to date, with legal national SP policies and strategies and outdated, without legal foundations, and fully their adequacy in terms of frameworks, basis, or only partially operationalized goals, provision, and roles operationalized SP institutional capacity and coordination: Extremely weak SP Constrained SP capacity, Adequate SP capac- Examines SP coordination structures and capacity (e.g., inade- with some coordina- ity and well-coordinated overall administrative capacity (e.g., ade- quate staffing) and SP tion among SP agencies, program delivery, with quacy of human, physical, and financial delivered in silos, with no governed by quasi-legal limited overlap and resourcing) or limited coordination structures effective information Foundational indicators among SP agencies sharing, governed by formal legal structures DRM policies and legislative frameworks: ■ DRM policy absent DRM policy in place and DRM policy current and Examines existence and relevance of or outdated and not relatively up to date but relevant, and legisla- national DRM policies in terms of how highly focused on cur- legislation and/or bud- tion in place to ensure well they reflect disaster risks affecting rent disaster risks and gets only partially in the policy’s provisions the country and put in place the legal and impacts place to support its are resourced and imple- financial frameworks to ensure DRM pro- implementation mented at all levels of ■ No clear backup by visions are implemented and enforced government legislation or gover- nance structures DRM institutional capacity and coordina- No or extremely weak DRM institutional lead- DRM strong, with coor- tion: Assesses the institutional capacity DRM government body ership effective, despite dination structures well to lead and coordinate DRM at different or agency in place to capacity constraints and resourced and support- levels in terms of adequate resourcing, deliver policy and legisla- coordination structures ive of the widespread authority, coordination structures, etc. tion and coordinate DRM with limited resources or implementation of activities mandates for risk reduc- ex post and ex ante DRM tion/mitigation activities activities 143 144 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Dimension Nascent Emerging Established Clear mandate for SP response: Exam- ■ No mention of ASP in ■ Some mention of ASP ASP a clear element of ines the extent to which the use of SP in SP or DRM policies, in SP and/or DRM SP and/or DRM policy response to or in advance of shocks is legislation, or strategy policy documents, and strategy at the explicitly supported by national SP and/ documents but with unclear or national and local levels or DRM policies and strategies nonspecific refer- ■ SP policy documents ences to roles and do not articulate a role responsibilities and for SP in response to implementation shocks arrangements ■ Role of subnational actors unarticulated Adaptive indicators and ASP absent from subnational policies Multisectoral coordination for SP No clear Some processes/guide- Clear cross-sectoral responses: Examines the extent to which cross-governmental pro- lines and coordination guidance in place on DRM, SP, or other coordination mecha- cesses or structures in structures in place, but respective responsibili- nisms are in place to support effective place in support of shock with no SP participa- ties of departments and vertical and horizontal coordination of response coordination tion or SP inadequately decision-making struc- ASP actions represented tures, including the role of SP in response to shocks Partnerships with nongovernmental Nongovernmental and Some ad hoc Role of nongovernmen- actors: Examines the mechanisms and government SP interven- partnerships with non- tal actors in response agreements in place for government tions delivered in parallel governmental actors to shocks (in terms of coordination with humanitarian actors, after a shock, with little when responding to financial, technical, and development partners, and civil society, or no coordination shocks supplemental implemen- as well as recent partnership experiences tation capacity) defined ex ante in protocols and MOUs A ppendix 145 Dimension Nascent Emerging Established Programs and delivery Coverage of the main social assistance Low coverage of the Reasonable coverage of Broad coverage of the program: Examines the coverage of SP main poverty reduction the main poverty reduc- main poverty reduction and social assistance programs, and SP program among the tion SP program among SP program among the in particular of programs that provide poor the poor poor support to poor households especially vulnerable to shocks, with higher cover- age providing a better platform for vertical and horizontal expansions and indicative of stronger administrative capacity Benefit adequacy: Examines the extent Regular SP ben- Regular SP benefits Regular SP benefits ade- to which current SP assistance meets efits inadequate, fairly adequate for con- quate for consumption households’ consumption needs or other with no effective sumption smoothing smoothing; adequacy gaps to achieve poverty reduction and consumption-smoothing level and/or the provi- resilience-building objectives impact sion of accompanying measures may also sup- port household resilience building Eligibility criteria and targeting: Examines Reliance on subjec- Reliance on objective Targeting based on Foundational indicators ■ the extent to which targeting processes tive assessments and/ and transparent meth- objective and trans- are based on clear, objective eligibil- or undocumented pro- ods for some programs; parent methods for all ity criteria and robust data collection cesses for targeting subjective assessments programs methods or undocumented pro- ■ Eligibility criteria and cesses for others targeting that account for vulnerability to recurrent/predictable shocks Financial inclusion and national ID: Exam- Low financial inclusion Financial inclusion and High financial inclusion ines national ID coverage and financial and national ID cover- national ID coverage and national ID coverage inclusion as key enablers of payment age in comparison to similar to relevant inter- according to interna- delivery, information sharing, and other relevant international national comparators tional standards processes comparators SP payment systems: Examines the Cash or food trans- Cash transfers mostly Cash transfers delivered modality through which cash trans- fers mostly delivered delivered electronically completely electroni- fers are delivered to social assistance manually but limited to a single cally, with options for beneficiaries delivery mechanism benefit delivery to pro- vide beneficiary choice SP delivery systems: Examines the No documented deliv- Documented delivery Well-documented deliv- extent to which administrative systems ery processes for regular processes but with oper- ery processes, with are strong and effective, focusing on programs ations manuals that are operations manuals reg- processes and systems used for iden- rudimentary and/or pos- ularly updated to reflect tification, eligibility assessment, and sibly out of date new procedures registration 146 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Dimension Nascent Emerging Established Experience of vertical expansion Never implemented Some experience with Experience with (increased benefits): Examines the extent vertical expansions, with large-scale and/or multi- to which SP programs have previously limited effectiveness in ple expansions effective provided additional temporary support terms of timeliness, cov- in terms of timeliness, during or after shocks to existing bene- erage, and adequacy coverage, and adequacy ficiaries; also considers how effectively this was done, including in terms of time- liness and adequacy of the response Adaptive indicators Experience of horizontal expansion Never implemented Some experience, with Experience with (increased coverage): Examines the limited effectiveness in large-scale and/or multi- extent to which SP programs have terms of timeliness, cov- ple expansions effective been expanded to new beneficiaries in erage, and adequacy in terms of timeliness, response to shocks and the ability to coverage, and adequacy scale horizontally beyond routine bene- ficiaries (which requires more advanced systems and planning) and assesses the timeliness, adequacy, and coverage of such recent experiences Protocols and contingency plans: Exam- No contingency pro- Protocols and contin- Protocols and plans in ines the availability of protocols and tocols for business gency plans for one or place for most programs plans for ASP that establish clearly how continuity and/or for more programs in place that have been tested SP should respond to shocks scaling up after a shock but not yet tested or and operationalized and operationalized are regularly updated Data and information systems Beneficiary registry: Examines the avail- No beneficiary registry Digital beneficiary reg- All SP beneficiary ability of digital registries to manage SP or, if in place, paper- or istry in place only for registries digitized programs and the level of interoperability Excel-based individual program(s), and integrated and and integration, with scoring also reflecting with limited integration interoperable data quality in terms of completeness, accu- and interoperability racy, accessibility, relevance, and currency Social registry: Examines the availabil- No social registry Social registry in place, Social registry in place ity of a social registry; its degree of but supports a single that is current, dynamic, interoperability and integration; and its program, is static, or has and covers a large share Foundational indicators data quality in terms of completeness, low coverage of the pop- of the population and accuracy, accessibility, relevance, and ulation or those most at those most at risk from currency risk shocks, with a good degree of integration and interoperability Data and information on disaster risk and ■ Very limited early ■ Some early warn- ■ High-quality and com- household vulnerability: Examines the warning and/or ing and household prehensive early extent to which data to identify and map household risk and risk and vulnerabil- warning data actual and potential shocks (e.g., early vulnerability data ity data and analysis ■ Risk and vulner- warning data) and household-level expo- in place but with some ■ Data not well analyzed ability analysis in sure and vulnerability are available and of limitations in terms of or used to inform deci- place and used to good quality (highly disaggregated/gran- quality and timeliness sion making inform planning and ular and up to date) ■ Data not well used to implementation inform shock responses or as an input to contin- gency planning A ppendix 147 Dimension Nascent Emerging Established Use of preexisting data for shock Preexisting data not ■ Some preexisting data Most relevant data response: Examines the extent to which used to inform ASP used to inform ASP sources systematically preexisting data are used to inform SP responses responses but on an utilized to inform ASP responses, including, critically, benefi- ad hoc basis shock responses ciary and registry data and, in the case of ■ Not all relevant infor- certain shocks, early warning information mation sources utilized for responses (e.g., early warning information) Post-shock household needs assess- Extremely low capacity ■ Needs assessments ■ Needs assessments ments: Examines the capacity to conduct to conduct postdisaster conducted at large conducted at large Adaptive indicators postdisaster household needs assess- needs assessments at scale, relying on scale, relying on elec- ments where preexisting data are limited, large scale, with serious paper-based methods tronic data collection incomplete, or out of date; where addi- delays in conducting the methods ■ SP shock responses tional eligibility criteria are required assessments, duplication informed by ad hoc ■ Databases linked to SP that are not readily available; or when across ministries, and/or assessments (not information systems household needs are expected to have overreliance on humani- based on protocols, substantially changed after the shock, tarian actors ■ Protocols and mech- standardized ques- requiring reassessment, and/or to be anisms in place for tionnaires, etc.) highly heterogeneous within communities post-shock needs assessments to inform SP responses Data-sharing protocols with internal/ No protocols or MOUs Some protocols or MOUs Data shared with proper external partners: Examines the capac- in place; data either not in place; data still shared legal and technical back- ity, in terms of protocols, MOUs, and shared or shared without without adequate legal ing through MOUs, technical solutions, to share data across proper legal and techni- and technical backing protocols, etc. government sectors and with external cal backing partners Finance Government spending on SP: Examines SP spending lower than SP spending similar to SP spending compares the proportion of GDP spent on SP and in comparable regions/ comparable regions/ favorably with compa- social assistance as a good indication of countries, heavily reliant countries, mostly reli- rable regions/countries, government commitment to SP and the on external financing, able, and domestically domestically funded, Foundational indicators maturity of the system and/or unpredictable funded consistent, reliable, and sustainable DRF policies and strategies: Examines No existing DRF strategy ■ Some existing DRF Comprehensive (risk whether a government proactively man- or policy documents policies or strategies layered) DRF strategy ages disaster risk by putting in place DRF but not backed by leg- in place for a range of policy and legislation, including legal pro- islation or aligned shocks, with support- visions and frameworks and instruments financial instruments ing legal and financial across ministries and at different levels of instruments ■ Work currently under government way to translate risk financing diagnostics into DRF strategies 148 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA Dimension Nascent Emerging Established Quantification of the expected costs for Costs of responding to Basic costing under- ■ Government systems ASP responses: Examines whether the shocks estimated with taken for using SP to and capacity present government has made efforts or put in SP not quantified respond to the primary for cost analysis place systems or models to calculate the shock faced by the coun- ■ Analyses undertaken costs of providing (or not) a minimum SP try but not to inform a for a range of possible response during or after disasters, as an financing strategy for Adaptive indicators shock types, and costs important prerequisite to developing DRF ASP estimated to define a mechanisms for ASP financing strategy for ASP DRF instruments linked to ASP: Assesses No specific financing Some financing instru- Range of pre-agreed the extent to which the government’s instruments earmarked ments identified and instruments in place to DRF mechanisms, where present, are for SP, with SP response earmarked for SP quickly cover the antici- linked ex ante to SP responses fully dependent on response, but normally pated cost of responding budget reallocation, inadequate or expected with SP to shocks of appeals, and external aid to be inadequate based different types and on quantification of severities post-shock costs Note: ASP = adaptive social protection; DRF = disaster risk financing; DRM = disaster risk management; GDP = gross domestic product; MOU = memorandum of understanding; SP = social protection. G LO S S A RY Adaptive social protection (ASP): As defined directives, organizations, and operational by the World Bank, refers to helping “to build skills and capacities to implement strategies, the resilience of poor and vulnerable house- policies, and improved coping capacities to holds by investing in their capacity to prepare lessen the adverse impacts of hazards and for, cope with, and adapt to shocks: protect- the possibility of disaster. ing their wellbeing and ensuring that they do not fall into poverty or become trapped in Disaster risk reduction (DRR): According to poverty as a result of the impacts” (Bowen the United Nations International Strategy et al. 2020). Often used interchangeably for Disaster Reduction, “aimed at prevent- with shock-responsive social protection, ing new and reducing existing disaster risk which tends to refer to the adaptation of and managing residual risk, all of which routine social protection programs following contribute to strengthening resilience and large-scale shocks. Both can include ex ante therefore to the achievement of sustainable actions, such as building shock-responsive development.” systems, plans, and partnerships to pre- pare better for emergency response; or ex Early warning system (EWS): The set of post actions, such as supporting households capacities needed to generate and dis- once the shock has occurred. seminate timely and meaningful warning information to enable individuals, commu- Beneficiary registry: Registry or database nities, and organizations threatened by a that tracks information on beneficiaries hazard to prepare for and act appropriately and benefits in one specific social protec- and in sufficient time to reduce the possibil- tion program. Beneficiary registries support ity of harm or loss. program management and implementation (payments, case management, condition- Horizontal expansion: Temporary inclusion alities monitoring, and grievance redress). of new beneficiaries from disaster-affected See also social registry and integrated communities in an existing social protection beneficiary registry. program. Disaster risk management (DRM): The sys- Integrated beneficiary registry: Social pro- tematic process of using administrative tection database that operates as a data 149 150 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA warehouse. Contains information from dif- pensions; health, unemployment, or disaster ferent social protection programs and their insurance; and funeral assistance. It can be benefits administration systems, allowing for provided formally through a bank or employer monitoring and coordination of who receives or informally through a community-based what benefits and for identifying intended or pooled fund. Social insurance is strongly unintended duplications across programs. linked to the formal labor market, meaning See also beneficiary registry and social coverage is often limited to formal workers. registry. Social protection: Government systems that Interoperability: The ability to exchange help individuals and families—especially the and make use of information among differ- poor and vulnerable—cope with crises and ent databases and registries. Interoperability shocks, find jobs, invest in the health and entails setting up legal, organizational, tech- education of their children, and protect the nical, and operational procedures to ensure aging population. Social protection includes data exchanges are secure, timely, effective, social assistance and insurance and other and efficient. interventions targeted to both poor and non- poor individuals and households, depending Management information system (MIS): A on specific needs at different stages of the system that transforms data retrieved from life cycle and/or the level of insecurity they a database or registry into information that experience as a result of external shocks. can be used for program management and The definition of social protection is not policy decisions. In social protection litera- fixed or universal. ture, an MIS is associated with program- or ministry-level information management. Social protection information system: The overarching system that enables the flow Shock-responsive social protection: See and management of information within the adaptive social protection (ASP). social protection sector and sometimes beyond to other sectors. It encompasses Social assistance: Direct, regular cash or registries and databases; software applica- in-kind transfers to poor and vulnerable tions; procedures for collecting, analyzing, individuals or households. Transfers are non- and sharing data; and human resources. contributory—that is, those in need are not asked to make contributions to be entitled Social protection system: Includes the policy to receive assistance. Some transfers are and legislative framework for social protec- targeted on the basis of categories of vulner- tion, including budget and financing, and the ability, and some are targeted to low-income specific social protection programs and their groups. corresponding delivery systems. See also systematization. Social insurance: Contributory programs in which participants make regular pay- Social registry: Management information ments to a plan that will cover costs related system containing information on all reg- to life-course events such as pregnancy, istrants, regardless of whether they are unemployment, or illness. Sometimes costs deemed eligible for or enrolled in a specific are matched or subsidized by the provider. social program. Social registries support pro- Social insurance includes contributory cesses of outreach, intake and registration, G lossary 151 and assessment of needs and conditions Public works program: A type of social to determine eligibility for multiple social assistance program that provides benefi- programs. Assessment usually considers ciaries with jobs on infrastructure projects measures of socioeconomic status, cate- in exchange for cash or food. Public works gorical factors, or a combination of both. programs are sometimes classified as labor See also beneficiary registry and integrated market interventions, depending on whether beneficiary registry. their function is primarily poverty alleviation, job creation, or social protection. Social transfer: Noncontributory, publicly funded, direct, regular, and predictable Vertical expansion: Temporary increase in resource transfer (in cash or in kind) to a poor the benefit value or duration of an existing and/or vulnerable individual or household. social protection program for some or all Transfers are intended to reduce recipients’ beneficiaries. deficits in consumption, protect them from economic and climatic shocks, and—in some cases—strengthen their productive capacity and human capital. Systematization: The idea that social pro- tection instruments can be integrated into a more comprehensive system of policies and programs that not only tackle poverty and vulnerability over the life cycle but also strengthen pro-poor and inclusive economic growth and social development. REFERENCES Abay, K. A., G. Berhane, J. F. Hoddinott, and K. Barca, V., and C. O’Brien. 2017. “Factors Affecting Tafere. 2020. “COVID-19 and Food Security the Usefulness of Existing Social Protec- in Ethiopia: Do Social Protection Programs tion Databases in Disaster Preparedness and Protect?” IFPRI Discussion Paper 1972. Inter- Response.” Policy brief. Oxford Policy Man- national Food Policy Research Institute, agement, Oxford, UK. Washington, DC. V., S. Brook, J. Holland, and M. ­ Barca, ­ Otulana. Agricultural Research Council. 2016. “Standard 2015. “Qualitative Research and Analysis of Precipitation Index.” Institute for Soil, Climate the Economic Impacts of Cash Transfer Pro- and Water, Republic of South Africa, Pretoria. grammes in Sub-Saharan Africa.” Synthesis report. Food and Agriculture Organization of Alahacoon, N., M. Edirisinghe, M. Simwanda, the United Nations, Rome. E. Perera, V. R. Nyirenda, and M. Ranaga- lage. 2022. “Rainfall Variability and Trends Beazley, R., and A. Williams. 2021. “Adaptive Social over the African Continent using TAMSAT Protection in the Caribbean: Building Human Data (1983–2020): Towards Climate Change Capital for Resilience.” 360° Resilience Back- Resilience and Adaptation.” Remote Sensing ground Paper. World Bank, Washington, DC. 14 (96): 1–27. Bhorat, H., T. Köhler, M. Oosthuizen, B. Stan- Anderson, S., H. Matthew, and M. Murray. 2017. wix, F. Steenkamp, and A. Thornton. 2020. “Building the Role of Social Protection in The Economics of Covid-19 in South Africa: Disaster Response and Resilience in Lesotho.” Early Impressions. DPRU Working Paper No. World Bank, Washington, DC. 202004. Development Policy Research Unit, University of Capetown. Asfaw, S., and B. Davis. 2018. “Can Cash Transfer Programmes Promote Household Resilience? Bowen, T., C. del Ninno, C. Andrews, S. Coll-Black, Cross-Country Evidence from Sub-Saharan U. Gentilini, K. Johnson, Y. Kawasoe, A. Krye- Africa.” In Climate Smart Agriculture, L. ziu, B. Maher, and A. Williams. 2020. Adaptive Lipper, N. McCarthy, D. Zilberman, S. Asfaw, Social Protection: Building Resilience to and G. Branca, eds. Springer. Shocks. International Development in Focus. Washington DC: World Bank Group. Barca, V., and R. Beazley. 2019. “Building on Gov- ernment Systems for Shock Preparedness Bündnis Entwicklung Hilft. 2016. World Risk and Response: The Role of Social Assistance Report 2016. Berlin: Bündnis Entwicklung Data and Information Systems.” Common- Hilft and United Nations University–Institute wealth of Australia, Department of Foreign for Environment and Human Security. Affairs and Trade, Canberra. CIMA (International Centre on Environmental Mon- itoring) Research Foundation. 2018. “Disaster Risk Profile: Botswana.” CIMA, Savona, Italy. 153 154 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA CSIR (Council for Scientific and Industrial GOB (Government of Botswana). 2009. “National Research) 2014. “Climate Information and Disaster Risk Management Plan.” National Early Warning Systems for Supporting the Disaster Risk Management Office, Office of Disaster Risk Reduction and Management the President. Sector in South Africa under Future Climates.” Presentation. GOB (Government of Botswana). 2016. “National Development Plan 11: Volume 1 April 2017– Dammert, A., J. J. de Hoop, ­E. Myukiyehe, and March 2023.” Rosati. 2018. “Effects of Public Policy F. C. ­ on Child Labor. Current Knowledge, Gaps, GOB (Government of Botswana). 2018a. “National and Implications for Program ­Design.” World Social Protection Framework.” Ministry of Development 110: 104–23. Local Government and Rural Development, Government of Botswana. del Ninno, Carlo, Sarah Coll-Black, and Pierre Fal- lavier. 2016. Social Protection Programs for GOB (Government of Botswana). 2018b. Africa's Drylands. Washington, DC: World “Botswana Multi-Topic Household: Survey Bank. Report 2015/16.” Statistics Botswana, Gov- ernment of Botswana. Dhemba, J. 2021. “Eswatini’s Social Policy Response to Covid-19: Temporary and GOB (Government of Botswana). 2019a. Donor-Driven.” CRC 1342 Covid-19 Social “Summary Report for the Drought and Policy Response Series No. 30. Bremen. Household Food Security Vulnerability Assessment 2018/19.” Ministry of Local Gov- Drechsler, M., K. Johnson, B. Maher, and J. Boko. ernment and Rural Development. 2019. “Eswatini Shock-Responsive Social Protection.” Unpublished. World Bank. GOB (Government of Botswana). 2019b. “Vision Botswana 2036.” ECHO (Directorate-General for European Civil Protection and Humanitarian Aid Operations). GOB and UNDP (Government of Botswana and 2021. “Strengthening Disaster Preparedness United Nations Development Programme). and Response Systems in Lesotho: Final 2013. “National Disaster Risk Reduction Report.” Strategy 2013–2018: Beyond Vulnerability— Towards Resilience.” EPRI (Economic Policy Research Institute). 2020. “Shock Responsive Social Protection Com- GOB and UNDP (Government of Botswana and munications and Advocacy Strategy.” Final United Nations Development Programme). draft. 2020a. “National Social Protection Recovery Plan; Part 1—Analysis and Recommendations.” EPRI (Economic Policy Research Institute). 2021. Ministry of Local Government and Rural “Update National Social Protection Strategy Development and UN. 2021–2025.” Strategic Dialogue Workshop Report. GOB and UNDP (Government of Botswana and United Nations Development Programme). Fauchereau, N., S. Trzaska, M. Rouault, and 2020b. “Quarterly Multi-Topic Survey: Labour Y. Richard. 2003. “Rainfall Variability and Force Module Report Quarter 1: 2020.” Sta- Changes in Southern Africa during the 20th tistics Botswana, Government of Botswana. Century in the Global Warming Context.” Nat- ural Hazards 29: 139–154. GOE (Government of Eswatini). 2010. “National Social Development Policy.” FinMark Trust. 2017. Annual Report 2017: Making Financial Markets Work for the Poor. Mid- GOE (Government of Eswatini). 2012. “National rand, South Africa. Multi Hazard Contingency Plan 2012 to 2013.” FinMark Trust. 2020. FinScope South Africa 2019 GOE (Government of Eswatini). 2015. “Swaziland: Pocket Guide. Midrand, South Africa. National Progress Report on the Implemen- tation of the Hyogo Framework for Action Gentilini, U., M. Almenfi, I. Orton, and P. Dale. (2013–2015)—Interim.” 2021. “Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country Measures.” World Bank, Washington, DC. REFERENCES 155 GOE (Government of Eswatini). 2018. “Orphans GOSA (Government of South Africa). 2002. and Vulnerable Children Cash Transfer Pilot: “Disaster Management Act no. 57 of 2002.” Impact Evaluation.” Deputy Prime Minister’s Pretoria. Office (DPMO), Government of Botswana. GOSA (Government of South Africa). 2005. GOE (Government of Eswatini). 2019a. “National “Disaster Management Act: Policy Framework Disaster Risk Management Operational for Disaster Risk Management in South Framework.” Africa.” Pretoria. GOE (Government of Eswatini). 2019b. “National GOSA (Government of South Africa). 2012. Operational Guideline for Emergency “National Development Plan 2030. Our Management.” Future Make It Work.” Pretoria. GOE (Government of Eswatini). 2020. “Food GOSA (Government of South Africa). 2015. Response Plan for COVID 19.” “National Disaster Management Amendment Act 16 of 2015.” Pretoria. GOE (Government of Eswatini). 2021. “National Social Assistance Policy.” GOSA (Government of South Africa). 2017a. “South Africa’s Third National Communica- GOL (Government of Lesotho). 2014. “National tion under the United Nations Framework Social Protection Strategy (NSPS), Convention on Climate Change.” Department 2014/15–2018/19.” of Environmental Affairs, Pretoria. GOL (Government of Lesotho). 2018. “Second GOSA (Government of South Africa). 2017b. National Strategic Development Plan.” “Stopping and Re-Allocation of Municipal Infrastructure Grant Allocation and Transfer for Disaster in Terms of Division of Revenue GOL (Government of Lesotho). 2020a. “Disaster Act.” National Treasury, Pretoria. Risk Reduction and Management Act.” Disas- ter Management Agency. GOSA (Government of South Africa). 2019. “National Climate Change Adaptation GOL (Government of Lesotho). 2020b. “An Inte- Strategy.” Pretoria. grated Policy Framework for Early Warning and Early Action.” GOSA (Government of South Africa). 2020. “Building a Society that Works; Public GOL (Government of Lesotho). 2020c. “Lesotho Investment in a Mass Employment Strategy Multi-Hazard Contingency Plan 2020–2023.” to Build a New Economy.” Pretoria. Disaster Management Agency. Hallegatte, S., M. Bangalore, Mook, L. Bonzanigo, GOL (Government of Lesotho). 2020d. “Methods M. Fay, T. Kane, U. Narloch, J. Rozenberg, D. for Updating the National Information System Treguer, and A. Vogt-Schilb. 2016. “Shock for Social Assistance.” Waves: Managing the Impacts of Climate Change on Poverty, Climate Change and GOL (Government of Lesotho). 2020e. Development.” World Bank, Washington, DC. “Multi-Hazard Early Warning System Manual for Disaster Risk Management.” Disaster Hill, R., E. Skoufias, and B. Maher. 2019. “The Management Agency. Chronology of a Disaster: A Review and Assessment of the Value of Acting Early on GOL (Government of Lesotho). 2020f. “National Household Welfare.” World Bank, Washing- Early Warning–Early Action Plan.” ton, DC. GOL (Government of Lesotho). 2020g. “National IFRC (International Federation of Red Cross Early Warning Strategic Action Plan.’ and Red Crescent Societies). 2021. “Final Report. Southern Africa (Botswana, Eswatini, GOL (Government of Lesotho). 2020h. “Social Lesotho, Namibia): Drought/Food Insecurity Protection Scalability Framework.” Emergency Appeal.” GOSA (Government of South Africa). 1996. “Constitution of the Republic of South Africa.” No. 108. Pretoria. 156 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA IMF (International Monetary Fund). 2020a. “IMF Roberts, B., J. Struwig, S. Gordon, and T. Radebe. Executive Board Approves US$110.4  million 2018. “Financial Literacy in South Africa: in Emergency Support to the Kingdom of Results from the 2017/18 South African Eswatini to Address the COVID-19 Pandemic.” Social Attitudes Survey Round.” Report pre- Press release no. 20/274. pared by the Human Sciences Research Council on behalf of the Financial Sector IMF (International Monetary Fund). 2020b. Conduct Authority, Pretoria. “Kingdom of Lesotho: Requests for Disbursement under the Rapid Credit Facility SADC (Southern African Development Commu- and Purchase under the Rapid Financing nity). 2020. “Synthesis Report on the State Instrument.” Press release, IMF staff country of Food and Nutrition Security and Vulnera- report, and statement by the Executive Direc- bility in Southern Africa.” Food, Agriculture tor for the Kingdom of Lesotho. and Natural Resources Directorate and the Disaster Risk Reduction Unit, SADC. IMF (International Monetary Fund). 2020c. “South Africa: Request for Purchase Under the SASSA (South Africa Social Security Agency). Rapid Financing Instrument.” Press release, 2019. “Annual Performance Plan 2019/20.” IMF staff report, and statement by the Execu- Republic of South Africa, Pretoria. tive Director for South Africa. SASSA (South Africa Social Security Agency). IPC (Integrated Food Security Phase Classifi- 2020. “2019/20 Annual Report.” Republic of cation). 2021. “Eswatini: IPC Acute Food South Africa, Pretoria. Insecurity Analysis July 2021– March 2022.” Schubert, B. 2018. “Quantitative Assessment of Kardan, A., C. O’Brien, and M. Masasa. 2017. the Social Assistance System in Eswatini.” “Shock-Responsive Social Protection United Nations Children’s Fund. Systems Research: Case Study—Lesotho.” Oxford Policy Management, Oxford, UK. Schulze, R. E., ed. 2016. “Thematic Booklet 14. Hazards and Climate Change in South Africa: Knippenberg, E., and J. F. Hoddinott. 2017. Fire Danger Rating under Natural Conditions.” “Shocks, Social Protection, and Resilience: Department of Agriculture, Forestry and Fish- Evidence from Ethiopia.” ESSP Working Paper eries, Republic of South Africa, Pretoria. 109. International Food Policy Research Insti- tute and Ethiopian Development Research Statistics South Africa. 2019. “Inequality Trends Institute, Addis Ababa. in South Africa: A Multidimensional Diagnos- tic of Inequality.” Republic of South Africa, Köhler, T., and H. Bhorat. 2020. “COVID-19, Social Pretoria. Protection, and the Labour Market in South Africa: Are Social Grants Being Targeted at Sulla, V., and P. Zikhali. 2018. “Overcoming the Most Vulnerable?” DPRU Working Paper Poverty and Inequality in South Africa: An 202008. Development Policy Research Unit, Assessment of Drivers, Constraints and University of Capetown. Opportunities.” World Bank Group, Washing- ton, DC. Longhurst, Daniel, Sophie Evans, Dayna Connolly, Felix Lung, Anna McCord, Stephanie Allan, Tlakane, L. 2020. “Second Round of CGP Vertical and Michele Plichta. 2021. “What Are Future Expansion Emergency Cash Top-Up Trans- Financing Options for Shock Responsive fers in the Southern Districts.” Unpublished Social Protection? A Technical Primer.” Social field monitoring report. Protection Approaches to COVID-19 Expert Advice Service (SPACE), DAI Global UK Ltd. Ulrichs, M., and M. Mphale. 2016. “Lesotho Country Case Study Report: Strengthening Raju, Dhushyanth, and Stephen D. Younger. 2021. Coherence between Agriculture and Social “Social Assistance Programs and Household Protection.” Overseas Development Institute, Welfare in Eswatini.” Social Protection & Jobs London. Discussion Paper No. 2106. World Bank, Washington, DC. UNDP (United Nations Development Programme). 2020. “Study on the Socio-Economic Impacts of Covid-19 in South Africa.” UNDP, Pretoria. REFERENCES 157 UNDRR (United Nations Office for Disaster Risk World Bank. 2019a. “Adaptive Social Protection Reduction). 2015. “Sendai Framework for (ASP) in Botswana.” World Bank South Africa Disaster Risk Reduction 2015–2030.” Office. Unpublished. UNDRR (United Nations Office for Disaster Risk World Bank. 2019b. “Lesotho: Disaster Risk Reduction). 2020. “Botswana: Risk-Sensitive Financing Diagnostic.” World Bank, Washing- Budget Review.” UNDRR Regional Office for ton, DC. Africa World Bank. 2019c. Lesotho Poverty Assessment: UNICEF (United Nations Children’s Fund). 2017. Progress and Challenges in Reducing “Social Protection Budget Brief 2017/2018.” Poverty. Washington, DC: World Bank. UNICEF (United Nations Children’s Fund). 2019. World Bank. 2019d. “Toward a Shock-Responsive “Evaluation of the Vulnerable Groups Feeding Social Protection Systems in Lesotho.” World Program.” Botswana Institute for Develop- Bank, Washington, DC. ment Policy Analysis and UNICEF, Gaborone. World Bank. 2020a. “Botswana Macro-Economic UNICEF (United Nations Children’s Fund). Covid Outlook.” World Bank, Washington, DC. 2020. “Socio-Economic Impact Analysis of COVID-19 in Botswana— Analysis Brief No. 1.” World Bank. 2020b. “Post-Disaster Household Assessments and Eligibility Determination WFP (World Food Programme). 2018. “Vulnerability for Post-Disaster Social Protection Benefits.” Analysis and Mapping: Food Security Analy- Guidance Note. World Bank, Washington, sis at the World Food Programme.” Saving DC. Lives, Changing Lives. November. World Bank. 2020c. “Universal Financial Access WFP (World Food Programme).2019. “End of 2020: Botswana.” Online. Season Update for 2018/19 and Overview of the Food Security Situation in 2019/20.” WFP World Bank. 2021a. “Climate Risk Country Profile: Regional Bureau for Southern Africa. Botswana.” World Bank, Washington, DC. WFP (World Food Programme).2020a. “Eswatini World Bank. 2021b. “Lesotho: Social Protection Annual Country Report 2020.” Programs and System Review.” World Bank, Washington, DC. WFP (World Food Programme).2020b. “WFP Lesotho Country Brief, April 2020.” World Bank. 2021c. “Macro Poverty Outlook: Botswana.” World Bank, Washington, DC. WFP (World Food Programme).2020c. “Southern Africa in Throes of Climate Emergency with World Bank. 2021d. “Revised Estimates of the 45 million People Facing Hunger across the Impact of Climate Change on Extreme Region.” Press release. January 16. Poverty by 2030.” Working Paper. World Bank, Washington, DC. World Bank. 2013. “Botswana Social Protection Assessment.” Botswana Institute for Devel- World Bank. 2021e. “South Africa: Disaster Risk opment Policy Analysis and World Bank, Finance Diagnostic.” Draft. World Bank, Washington, DC. Washington, DC. World Bank. 2014. “Financial Protection against World Bank. 2021f. “South Africa: Social Natural Disasters: An Operational Framework Assistance Programs and Systems Review.” for Disaster Risk Financing and Insurance.” Washington, DC: World Bank. World Bank, Washington, DC. World Bank. 2021g. “Stress Testing Social World Bank. 2015. “Identity Management Systems Protection: A Rapid Appraisal of the Analysis: Botswana Country Report.” World Adaptability of Social Protection Systems Bank, Washington, DC. and their Readiness to Scale-Up.” World Bank, Washington, DC. World Bank. 2018. “The State of Social Safety Nets 2018.” World Bank, Washington, DC. 158 ADAPTIVE S O C IAL P R OTECTI O N I N S O UTH E R N AF R I CA World Bank. 2021h. “Updated Estimates of the World Bank. 2022b. “Eswatini: Disaster Risk Impact of COVID-19 on Global Poverty: Finance Diagnostic.” World Bank, Washing- Turning the Corner on the Pandemic in ton, DC. 2021?” Data Blog, June 24. World Bank, Washington, DC. World Bank. 2022c. “Inequality in Southern Africa: An Assessment of the Southern Africa World Bank. 2022a. “Botswana Social Protection Customs Union.” World Bank, Washington, Programs and Systems Review.” Washington, DC. DC: World Bank.