COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience Yoonyoung Cho and Doug Johnson Source: https://www.facebook.com/dswdserves COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 1 March 2022 The World Bank PHILIPPINES 26th Floor, One Global Place 5th Ave. corner 25th St. Bonifacio Global City, Taguig City Philippines 1634 T: +63 2-465-2500 F: +63 2-465-2505 W: www.worldbank.org/en/country/philippines This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, 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 of the data included in this work. 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 considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 1 This note was prepared by Yoonyoung Cho (Senior Economist) and Doug Johnson (Social Protection Consultant) at the World Bank. The work is based on various studies, reports, and survey results, which document COVID-19 policy measures and their impacts. The authors are grateful to Ruth Rodriguez, Yasuhiro Kawasoe, and Ma Laarni Revilla (Philippine Social Protection team at the World Bank) for their valuable feedback and input. Ugo Gentilini (Global Lead of Social Safety Net at the World Bank) and Aniceto Orbeta (President of the Philippine Institute of Development Studies [PIDS]) provided excellent comments as peer reviewers. Robert Palacios (Lead Specialist), Jonathan Marskell (Senior ID Specialist), Nadia Belghith (Senior Poverty Economist), Moira Enerva (Communication Consultant), and Pol Villanueva (Graphic Designer) supported finalization of the note. The team also thanks Ndiame Diop (Country Director) and Yasser El- Gammal (Practice Manager) for their guidance. Collaboration from the Department of Social Welfare and Development is acknowledged and greatly appreciated. This note benefited from generous support from the Department of Foreign Affairs and Trade (DFAT) of the Government of Australia and G2Px Trust Fund. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 2 Table of Contents Executive Summary 4 1. Introduction 5 2. Background 6 2.1 Pre-Pandemic Social Protection Context 6 2.2 Evolving COVID-19 Environment 7 2.3 COVID-19 Social Protection Response 9 Social Amelioration Program 10 Localized Ayuda 11 Worker Protection Programs 12 3. Assessment of SA Performance 13 3.1 Coverage and Adequacy 13 3.2 Targeting 18 3.3 Financing and Delivery 22 4. Policy Recommendations 25 References 27 Annex A. Additional indicators reflecting targeting accuracy 30 COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 3 Executive Summary The COVID-19 pandemic in the last two years has activities gradually resumed when quarantine gone through multiple phases in the Philippines restrictions were relaxed, but recovery of and its impact is still unfolding. The magnitude of household incomes was much slower. A significant disruptions in jobs and livelihoods due to lockdown increase in the poverty rates is expected. measures as well as government social protection responses have varied across time. In the same • The targeting accuracy of social assistance period, the country experienced natural disasters programs varied greatly. 4Ps continued to be including Typhoon Odette, which triggered the relatively well targeted, though targeting accuracy declaration of a state of calamity and underscored continued to decline. SAP also demonstrated the importance of an adaptive social protection (SP) progressivity in targeting. Other programs were not response. as well-targeted and, in some cases, regressive. While food aid reached households faster than This note describes the phases of the COVID-19 cash due to the country’s well established relief SP response in the Philippines and assesses the mechanism developed to cope with frequent performance with a focus on social assistance (SA). natural disasters, it was not well targeted, the We provide estimates of the coverage, adequacy, monetary value was small, and it was also costly to targeting accuracy, timeliness, and payment delivery deliver. of COVID-19 social assistance in the Philippines based on several household surveys and studies. • SAP payments were delivered through three We pay special attention to the Philippines’ flagship different channels: cash cards for current 4Ps social protection program, the Pantawid Pamilyang beneficiaries, physical delivery by LGU officials, Pilipino Program (4Ps), and the COVID-19 emergency and banks and e-money issuers. This indicates cash assistance program, the Social Amelioration that the government-to-person (G2P) payment Program (SAP). Salient findings are summarized as mechanisms in the country have mixed levels of follows: development. The initial attempt at large scale digital payments using banks and e-money issuers • Overall SA coverage was very high during the for the second tranche of SAP presented both pandemic. In the first half of 2020, 96 percent of current challenges and future opportunities for Filipino households reported receiving some form leapfrogging the country’s G2P mechanisms. of social assistance. The most common forms of assistance were local relief (mainly food aid The experience provides some important lessons. distributed by local government units or LGUs). First, it is critical to have well-established SP policies, SAP also reached a high share of households— programs, and systems upon which emergency close to 60 percent of households based on a operations can be built. Second, SP delivery systems nationwide survey—and about 23 million require regular updates and modernization beneficiaries based on administrative data. investment and efforts. Third, digital tools and technological advances should be actively used for • Despite the large coverage of SA programs, SP delivery. Fourth, timely and agile policy their ability to mitigate the negative impact adaptations building on the lessons learned from of the pandemic was modest. Timely support each shock can help strengthen overall SP systems provided to 4Ps beneficiaries helped them cope and delivery for future resilience. with the shock better. Delivery of cash assistance to beneficiaries beyond 4Ps was challenging due The note proposes the following priorities for adaptive to weak delivery systems. The overall impact and resilient SP program delivery: adoption of the of SA was modest given the magnitude of the national ID system for SP delivery; enhancement of shock and the prolonged pandemic as well as the the targeting system; development of digital platforms required amount and timeliness of assistance. and tools; continued innovation in digital G2P Food insecurity rose sharply, and many households payments; strengthening of contingency financing were forced to employ negative coping strategies. mechanisms and readiness for disaster response; and Household employment rebounded as economic shift of resources from in-kind to cash assistance. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 4 1. Introduction COVID-19 has caused an unprecedented health discretion of local government units (LGUs) and economic crisis in the Philippines. Between continued and increased in the later phase of the the start of the crisis and March 29, 2022, about 3.7 pandemic. million Filipinos contracted COVID-19 and over 59,000 died from the disease;1 the GDP growth rate In December 2021 amid the pandemic, Typhoon fell from 6.0 percent in 2019 to -9.5 percent in 2020 Odette hit the country and heavily affected six then rebounded to 5.6 percent in 2021;2 and the regions. A state of calamity was declared in the poverty rate that was on the downward trend prior regions, the third such declaration in the country in to the pandemic increased from 16.6 percent in 2018 the last two years. This new state of calamity to 23.7 percent in the first semester of 2021 (based overlapped with the nationwide state of calamity on estimates of the Philippine Statistics Authority).3 due to COVID declared in March 2020 and extended Survey evidence and official statistics indicate that over time. It was also in addition to the earlier state the crisis has led to a number of secondary negative of calamity that was declared in May 2021 due to the effects, including hunger, stress and anxiety, and African swine flu. While the coverage, populations domestic violence; and simulations suggest that affected, and severity of the shocks varied, the there are likely large additional unmeasured events highlighted the country’s frequent exposure negative effects, such as high learning loss among to natural disasters and the importance of adaptive students (Rivas 2021; Tee et al. 2020; Calleja 2020; social protection for resilience. Gayares 2021). In this note, we use representative household The Government of the Philippines acted quickly survey and administrative data to discuss the to provide assistance to households affected by coverage, adequacy, targeting accuracy, the crisis especially during the early period of the timeliness, and payment delivery of COVID-19 SA pandemic by adjusting existing social assistance in the Philippines. We synthesize results from a programs and launching new ones. Conditions for variety of data sources, including the Annual the country’s flagship conditional cash transfer Poverty Indicator Surveys (APIS), Family Income program, the Pantawid Pamilyang Pilipino Program and Expenditure Surveys (FIES), Social Weather (4Ps), were lifted during the initial period of the Stations (SWS) surveys, Labor Force Surveys (LFS), pandemic;4 in-kind food aid was distributed to many World Bank Household Panel and Economic (HOPE) households; and a new and large unconditional Survey,5 and World Bank High Frequency cash transfer program aiming to cover over three Monitoring (HFM) Survey.6 Where possible, we also quarters of Filipino households, the Social compare different social assistance programs and Amelioration Program (SAP), was launched. With different methods for delivering social assistance to narrowing fiscal space as the pandemic continued, provide insight into the combination of social and the COVID situation varying by locality, assistance programs and delivery channels that is nationwide social assistance (SA) response was likely to be most effective in future emergencies. phased out whereas assistance based on the Photo by: Joham Kirby Datoy 1 https://ourworldindata.org/coronavirus/country/philippines 2 National Accounts. Philippine Statistics Authority. 3 https://psa.gov.ph/poverty-press-releases 4 Conditions of 4Ps were brought back in October 2020. 5 See policy notes based on the HOPE survey focusing on low-income households: https://www.worldbank.org/en/country/philippines/brief/covid-19-impacts-on-low-income-families-in-the-philippines 6 See overview, survey findings, and policy notes using the high frequency monitoring survey – firms, households, and communities: https://www.worldbank. org/en/country/philippines/brief/monitoring-covid-19-impacts-on-firms-and-families-in-the-philippines COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 5 2. Background 2.1 Pre-Pandemic Social Protection Context them, it was also found that 4Ps has reduced gender- The SP Policy Framework in the Philippines has based violence among the poor. The program, the seen steady progress with synergistic initiatives largest SP program in the country implemented by the and legislations. The Philippine Development Plan Department of Social Welfare and Development 2017-2022 recognizes SP as an important pillar in (DSWD), serves about 4.2 million poor and vulnerable poverty reduction and inclusive growth. The Pantawid households objectively identified by the country’s Pamilyang Pilipino Program Act (4Ps Act) was targeting system, Listahanan. enacted in 2019, building on over a decade’s success of the program and institutionalizing it as a national The Unconditional Cash Transfer (UCT) Program poverty reduction strategy. The law was the basis for under the Tax Reform for Acceleration and the SP Operational Framework, which was Inclusion (TRAIN) Law was introduced as a tax complemented by the SP Plan for 2020-2022. Along reform mitigation program in 2018. The intention with the 4Ps Act, other key legislations and reforms to was to support households that may not benefit from support the poor and vulnerable populations were the lower income tax rates but may be adversely established.7 These laws, operational framework and affected by rising prices. It covers about 2.2 million plans helped establish the institutional arrangement additional poor households identified through and coordination mechanisms among various SP Listahanan in addition to current 4Ps beneficiaries (4.2 agencies. Among various SP instruments–SA, social million) and non-contributory social pension (SocPen) insurance, social services and active labor market beneficiaries of indigent senior citizens (3.4 million). programs–SA is most prominent with a wide coverage Thus, close to 10 million households receive benefits in the Philippines. from the TRAIN-UCT, although the extent to which the three different target groups overlap is unclear. Since The achievements of 4Ps are noteworthy and the TRAIN-UCT program was introduced to widely recognized. Numerous studies show that the temporarily mitigate the impact of the tax reform, the program has contributed to reducing poverty, program expired in 2021. 4Ps beneficiaries received improving human capital, and supporting women’s additional TRAIN-UCT benefits of ₱300 [US$6] per empowerment, among others. 4Ps is a conditional month while the program lasted, in addition to their cash transfer (CCT) for promoting health, nutrition, regular health and education grants supplemented by and education for the poor. The 4Ps Act increased the rice subsidies. SocPen beneficiaries received their benefit amount, and the program now transfers ₱750 regular benefits (₱500 [US$10]) and TRAIN-UCT (US$15) for monthly health grant and ₱300 to 700 benefits (₱300 [US$6]) per month. (US$6 to 14) for monthly education grants depending on the level of grades up to 10 months a year. In contrast to social assistance which has high Additional benefits include rice subsidy (₱600 [US$12] coverage, social insurance is accessible to only a per month) and PhilHealth coverage.8 The positive fraction of the workforce, mostly in the formal impacts of 4Ps are evident across various indicators. sector. Two institutions administer contributory social Estimates suggest that a quarter of overall poverty insurance/pension schemes for the public and private reduction in the country from 2006 to 2018 could be sectors: the Government Service Insurance System attributed to government transfers driven by 4Ps (GSIS) and the Social Security System (SSS), (World Bank 2018). Likewise, 4Ps children were more respectively. The SSS includes a partially funded likely to stay in school than others from a similar scheme for informal workers,10 with de-jure worker socioeconomic background (e.g., junior high school participation in the pension schemes mandatory; enrolment rate at 83.4 percent for 4Ps versus 78.2 however, actual or de-facto coverage rates are low percent for non-4Ps) (Orbeta et al. 2014). 4Ps (only about 38 percent of wage workers contribute to households showed significantly lower incidence of SSS). Further, the distribution of pension income is hunger,9 and greater expenditure for children (e.g., concentrated in higher-income households (World education, clothing and footwear) in part due to Bank 2018). The SSS faces a challenge in achieving payments being transferred mostly to mothers. By long-term sustainability considering the aging profile transferring resources to women and empowering of contributors and retirees. 7 These include the Universal Health Care Act, Social Security Act, Magna Carta of the Poor, National Commission of Senior Citizens Act, Community-Based Monitoring System Act, and Philippine Identification System Act. 8 See Cho et al. 2020a. 9 Share of households reporting hunger: 13.1% for 4Ps versus 18.7% for non-4Ps. In the 3rd impact evaluation based on 2017-2018 data. 10 Contribution rates are 8.5% for employers and 4.5% for employees, or 13.0% total for both as of 2021. The self-employed pay the full 13.0 percent contribu- tion. Options to contribute to SSS for voluntary and non-working spouse members as well as OFWs also exist. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 6 Despite active implementation of various SP (Figure 1). The last quarter of 2021 saw a huge programs especially SA,11 SP delivery systems are decline in the number of cases as if the pandemic underdeveloped in the country. Only 4Ps and social was over in the Philippines, but in the beginning of insurance programs use digital channels to transfer 2022, the country again experienced an resources to individuals, but most other programs unprecedented surge, in part due to the Omicron tend to rely on physical (over-the-counter) cash variant. delivery. When the pandemic happened, the national targeting system (Listahanan) that was intended to The pandemic period can be roughly divided into help identify and verify eligible beneficiaries, was not four phases, characterized by varying severity of up-to-date. Moreover, the national ID system quarantine and social protection measures, levels (PhilSys) was at a nascent stage and was not of compliance, and global environment (Table 1).13 operational. Phase 1 of the pandemic covered Q1 and Q2 of 2020, when the lockdown was most stringent and mobility 2.2 Evolving COVID-19 Environment was significantly limited. Business and consumer COVID-19 spread to the Philippines in early 2020, confidence plummeted, and the economy contracted with the first official case in late January and the by 16.9 percent from Q2 2019 to Q2 2020. Little was first official case of community transmission in known about the virus worldwide, and lockdown March 2020. To reduce the spread of the virus, the measures were considered the only option to “flatten Government of the Philippines declared a state of the curve” and avoid the collapse of the healthcare calamity in March 2020 and imposed strict system. During this time, compliance with lockdown lockdowns on many areas of the country, restricted measures was high but consequently, many jobs travel, closed businesses, and prohibited large were lost with the skyrocketing unemployment rate gatherings.12 The number of daily COVID cases was (World Bank, 2021). The Philippine government relatively low in 2020, but fluctuated widely in 2021 introduced a large social protection program, SAP, to with an average level a lot higher than that of 2020 support pandemic-affected populations, including displaced workers. Figure 1. New daily cases of COVID-19 in the Philippines 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 0 0 0 1 1 1 0 1 1 1 20 1 2/ 0 4/ 0 6/ 0 9/ 0 1 2 21 21 21 2/ 2 2/ 1 0 20 20 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 /2 1/ 0/ 1/ 1/ 0/ 1/ 31 31 31 31 30 30 30 31 28 30 30 30 31 31 31 31 28 29 30 31 /3 /3 /3 /3 /3 /3 3/ 5/ 8/ 7/ 1/ 1/ 3/ 5/ 8/ 4/ 6/ 9/ 7/ 1/ 10 12 12 11 10 11 new_cases 7 day average Source: Our World in Data 11 Apart from programs providing regular cash transfers at scale, there are several other programs that address community needs or respond to disasters. The Sustainable Livelihood Program and National Community Driven Development Program support poor, vulnerable, and marginalized households or communities in building their assets and promoting self-help through community projects. Various Cash [Food]-for-Work, Cash [Food]-for- Training, Cash [Food]-for-Caring support, assistance for emergency shelter, and most recently, the Emergency Cash Transfer, are inplace to provide emergency assistance. Assistance to individuals in crisis situations (AICS) provides temporary cash assistance based on on-demand requests in case of exposure to personal shocks. The AICS program was a small scale support with an on-demand application, but during the pandemic, a large number of beneficiaries received support from this program. 12 According to the Oxford COVID-19 Government Response Tracker, the lockdown in the Philippines was one of the most severe lockdowns in the world and significantly more severe than in most countries in the East Asian and Pacific region.. 13 This part is adopted and updated from the Philippines Economic Update – December 2021 edition. See World Bank (2021). COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 7 Phase 2 was between Q3 of 2020 and Q1 of 2021 and uncertainties and inconsistencies in community when the economy was slowly rebounding and quarantines and government support prevailed. jobs were coming back, in part with the help of Similar patterns were repeated in the surge in August the end-of-year seasonal effect. Better knowledge 2021 due to the Delta variant. The government’s about the disease and active vaccine development health sector efforts focused on providing vaccines and deployment14 occasioned cautious optimism. throughout the country and addressing vaccine The implementation of SAP during this phase was hesitancy and misinformation. By the end of the year, concentrated in areas still under the enhanced the number of cases rapidly declined and remained community quarantine (ECQ). Meanwhile, overall very low in November and December 2021 as if the business and consumer confidence was slowly pandemic was coming to an end. returning, and government support focusing on the labor market (e.g., wage subsidies and public works) Phase 4 in Q1 2022 is again seeing an was introduced and enhanced. During this phase, unprecedented surge of the Omicron variant, with measures to control the spread of the virus were the alert system enhanced and stricter measures diversified from lockdowns to contact tracing, testing, implemented in many parts of the country. With an and isolation. estimated 63 million people (58 percent of population) fully vaccinated,15 the Philippine Phase 3 was marked by a surge of COVID-19 cases government’s efforts are focused on last mile delivery in Q2 2021 that led to another round of strict of vaccines and mass vaccination for booster shots. lockdown measures in the National Capital Region While hospitalization and death rates may be lower, (NCR) and surrounding regions. Since then, the sheer number of cases continues to raise localized lockdowns depending on the number of concerns about the health care system. Despite new cases were imposed. Along with the decisions many uncertainties, an increasing number of on lockdown measures, SA support also became countries are acknowledging a COVID endemic decentralized, relying on the discretion of local scenario where the virus threat would become less governments. As the cash assistance distributed as serious after a large-scale inoculation or infection. In part of the national government’s initiative dwindled, such a case, even if the number of cases increases, local ayuda consisting mostly of food and non-food reduced hospitalization or death rates can occasion a in-kind packages were provided to locally identified cautious shift to a period of living with the virus. populations. During this period, lockdown fatigue Table 1. Characteristics of the four phases of the pandemic Phase 1 Phase 2 Phase 3 Phase 4 Characteristics Beginning – Q2 2020 Q3 2020 – Q1 2021 Q2 2021 – Q4 2021 Q1 2022 Stringent community Relaxed community Mixed community Localized alert systems Lockdown quarantines, high quarantines quarantines, measures compliance lockdown fatigue Huge uncertainty; heavy Modest confidence; Growing confidence; Unprecedented COVID reliance on lockdowns marked by need for marked by need for large surge due to variants; Health policy tracing, isolating, and vaccine procurement and booster vaccination drive; environment treating, in addition to distribution marked by uncertainty lockdowns SAP 2nd tranche (cash, Major COVID-19 SA Local assistance or ayuda Local assistance or ayuda SAP 1st tranche (cash, large scale, focusing on programs including for (cash and in-kind per LGU (cash and in-kind per large scale, nationwide) NCR and neighboring displaced workers discretion) LGU discretion) regions) Limited wage subsi- Expanded cash for work, Continued programs, dies and cash for work, continuous repatria- renewed commitment Major COVID-19 labor increased demand for tion and reintegration through the National Continued programs market programs repatriation support of of OFWs; training and Employment Recovery OFWs livelihood assistance Strategy Vaccine optimism mixed Global Economic rebounding/ Acknowledging a COVID External enviroment with caution against vari- recession/uncertainty vaccine optimism endemic scenario ants’ surge 14 COVAX, a global vaccine facility for COVID-19 vaccine development and deployment was launched in July 2020. 15 As of February 27, 2022. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 8 2.3 COVID-19 Social Protection Response measures were implemented in the later phases of The government’s decision to support households the pandemic in 2021. As the central government’s affected by the pandemic was swift. In late March fiscal and human resources for additional SA 2020, the Bayanihan16 to Heal as One Act (Bayanihan programs diminished and the prospect for potential I) was enacted within two weeks of the declaration of Bayanihan III became dim despite multiple bills, local a state of calamity in mid-March. The law included government units took the lead in providing ayuda in far-reaching initiatives to help firms and households a decentralized manner starting in April 2021 (Phase cope with the effects of the crisis which included, 3). DSWD expanded its specialized program for along with several other smaller programs, a new emergency response, the AICS, to provide support to and large unconditional cash transfer program called beneficiaries on an on-demand basis and also Social Amelioration Program or SAP. 17 Conditions of through LGUs.20 Unlike the centralized system of 4Ps associated with school attendance, health facility cash assistance, localized ayuda can be either cash visits, and family development sessions were waived or in-kind and mostly untargeted. The challenging on the basis of force majeure, and 4Ps beneficiaries part of this approach is the lack of coherent data that automatically became priority recipients in SAP. indicate the coverage and size of the transfers, which vary widely by localities. Meanwhile, voluntary In September 2020, the government passed the “community pantries”21 from grassroot organizers Bayanihan to Recover as One Act or Bayanihan II. became part of high-profile phenomena in the This enabled expanded support for labor market country, indicating both the bayanihan spirit of interventions while continuing efforts for SP as citizens and the daunting needs that are beyond envisaged in Bayanihan I. Labor market measures government’s support. through Bayanihan II include expansion of a cash- for-work program called TUPAD or Tulong The Philippines spent roughly 2.5 percent of GDP Panghanapbuhay sa Ating Disadvantaged/Displaced on the overall fiscal response to COVID-19, the Workers18 that helped displaced informal sector majority of which was for social protection. When workers; and access to capital with concessional measured by growth rate of real GDP per capita in rates, especially to micro, small, and medium 2020, it is clear that the Philippines experienced enterprises, through the Small Business Corporation, larger shocks than regional peers. While the Land Bank of the Philippines (LBP), Development Philippines’ response to COVID-19 was large by Bank of the Philippines, and PhilGuarantee (World historical standards and received justification Bank, 2021). While it was encouraging that the overall considering the magnitude of the shock, it was direction of policy providing support beyond SA smaller as a share of GDP compared with many other included labor market interventions through countries’ responses (Figure 2). For example, Bayanihan II, the budget allocation was not as Thailand, which experienced a smaller economic extensive as Bayanihan I (World Bank, 2021).19 contraction than the Philippines, spent over 8 percent of GDP on its COVID-19 response. Fiscal After SAP implementation, SA support became space and debt-GDP ratios were important decentralized as localized lockdown and alert considerations for determining the size of response. Source: https://www.facebook.com/dswdserves 16 “Bayanihan” is a Filipino word that roughly translates as “spirit of civic unity.” 17 In official documents SAP is referred to as the “Emergency Subsidy Program.” We use the more widely used “SAP” in this paper. 18 TUPAD is implemented by the Department of Labor and Employment. 19 Based on the Department of Budget and Management (DBM) data on COVID-19 releases as of September 30, 2021, Bayanihan I had more than ₱387 billion total budget allocation to implement the law while for Bayanihan 2, about ₱214 billion was allocated. 20 DSWD reports that AICS supported 2.8 million individuals in 2020 to mid-2021. https://www.pna.gov.ph/articles/1153520 21 In April 2021, local entrepreneur Ana Patricia Non created a small “community pantry” in her neighborhood of Quezon City and encouraged local residents to “give according to your ability, take according to your need.” Non’s community pantry concept attracted large media attention and resulted in numerous similar community pantries. However, it was not received well by some quarters in government (Gozum et al. 2021). COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 9 Figure 2. Magnitude of COVID-19 shock and fiscal response as share of GDP Source: Real GDP per capita growth figures are from the January 2022 issue of the World Bank’s Global Economic Prospects database. Size of fiscal response figures are from the March 2022 issue of the IMF’s database of fiscal responses to COVID-19. Note: The blue line indicates a Locally Weighted Scatterplot Smoothing (LOWESS) curve. PHL=Philippines; IND=Indonesia; MYS=Malaysia; THA=Thailand; VNM=Vietnam; and CHN=China. Social Amelioration Program households were able to receive their SAP quickly The major component of the government’s through the digital channel by the first week of April COVID-19 SP response was SAP, a new and large 2020 in most cases (Cho et al. 2020a). unconditional cash transfer program. The initial intention was to deliver ₱5,000-8,000 (US$100-160) Identifying SAP beneficiaries beyond 4Ps and in cash per month for two months to 18 million transferring cash grants to them turned out to households across the country (DSWD 2020).22 The be far more challenging than initially envisaged. level of benefits was calculated to cover one month Given that there was no readily available, accurate, of household subsistence expenses based on the and recent list or registry of poor and vulnerable minimum wage of each region. The eligibility criteria persons,23 DSWD introduced a paper application of SAP included categorical conditions (e.g., solo form (i.e., Social Amelioration Card or SAC form) and parent, pregnant women, informal sector workers) manual application process, while LGUs prioritized and welfare conditions (e.g., indigent individuals, poor and vulnerable populations based on their local those unemployed and with income loss). 4Ps knowledge. Intake of beneficiaries based on self- beneficiaries were automatically included for top-up reported assessment forms and local verification benefits from SAP. As of 2019, nearly two-thirds of (e.g., visual inspection of housing conditions for 4Ps beneficiaries received their benefits digitally disaster relief) has been commonly used during through LBP cash cards (Acosta et al. 2019). When disaster relief and humanitarian assistance delivery.24 SAP was introduced, close to 90 percent of 4Ps However, given the nature of the shock brought beneficiaries already had the cash cards, and these about by the pandemic with mobility constraints, 22 The total number of households in the country is approximately 23-25 million based on PSA estimates. So the initial target was to provide 72 to 78 percent coverage of the support. 23 The National Household Targeting System (Listahanan) covers around 70 percent of population with socioeconomic status at the household level, but the information was outdated as the latest data (Listahanan 2) were collected in 2015. Enumeration for a new Listahanan 3 was ongoing until the activity was suspended in March 2020 due to the pandemic. The enumeration for Listahanan 3 resumed in September 2020, but due to large-scale grievances and validity challenges, the targeting system was not yet operational as of December 2021. 24 The Joint Memorandum Circular No. 1 Series of 2020 identified the SAC as the primary means of identifying SAP beneficiaries. The SAC form is distributed in the barangay; it captures the family profile that will be used by families to access the social amelioration programs of government, similar to the Disaster Assistance Family Access Card. However, given the situation related to COVID-19 with restrictions in mobility, implementation challenges were anticipated: SACs were physical forms that needed considerable effort and coordination to distribute to LGUs; geographically isolated areas had difficulties receiving SAC in a timely manner, raising the concern for delay and exclusion; LGU staff were identified to consolidate and encode the SAC, which added to the workload of already stretched workers engaging in other important COVID-19 activities; SAC required physical submission of forms which was challenging during community quarantines and restricted mobility. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 10 infection concerns, and absence of visual verification geographically, with each FSP delivering the mechanisms for the eligibility criteria, implementing assistance in its assigned locality (though in some the process was a huge challenge. Without a areas there were both a bank and an EMI assigned comprehensive digital ID that enables beneficiary for cash delivery to accommodate beneficiaries that verification and cross-checking with government did not have smartphones). databases, and with the time pressure for program implementation, beneficiary enrolment was done In addition to the front-end differences in the with little deduplication or eligibility verification.25 delivery of cash assistance, there were also Consequently, there were significant delays in SAP substantial back-end differences between the delivery and confusion among LGUs and non-4Ps three channels in handling beneficiary data. beneficiaries entitled to the SAP. DSWD’s record Little back-end processing was required for 4Ps shows that the distribution of the first tranche of households since all of them were included in SAP SAP (SAP 1) for non-4Ps households stretched and systems were in place to transfer cash to most beyond June 2020.26 Moreover, a greater number of 4Ps beneficiaries. For non-4Ps households receiving beneficiaries requested SAP benefits than initially SAP 1 payments, collection of application details and anticipated, in part due to the confusion between approval of applications were performed manually “family” as commonly understood and “household” at the local level.30 In SAP 2, beneficiary details from as a unit of SA support as well as few mechanisms to SAC were first digitized and then shared with the differentiate them.27 FSPs for digital payments. To facilitate data entry, the government, with the support of a group of The experience in the implementation of SAP 1 led volunteer coders, created a web-based application to several changes in the implementation of the called ReliefAgad for entry of SAP beneficiary details. second tranche or SAP 2.28 The government had to However, with various technical challenges, the accommodate many waitlisted families that applied application was unable to fulfill its intended purpose, for SAP and did not receive the first tranche benefits. with self-reported data of approximately 2 million To include more families within the allocated budget, households entered without any verification or LGUs still under ECQ in May 2020 were prioritized. deduplication. More common was that LGU officials DSWD’s records29 indicate that SAP 1 payments coded SAC information using Excel spreadsheets were delivered to 16.7 million households, among that they then shared with DSWD (Cho et al. 2021). whom close to 4.3 million were 4Ps beneficiary- households. SAP 2 benefitted a total of 12.6 million Localized Ayuda households, including 1.4 million 4Ps households, In addition to SAP, and as in previous crises, the among whom 7.6 million households received the government delivered large quantities of food aid. two transfers. In addition, the government scaled up While some of this food aid was centrally managed digital payments for the SAP 2 by partnering with six by DSWD, a large portion of the aid was managed financial service providers (FSPs)–three traditional locally by LGUs. From the onset of the pandemic, banks and three e-money issuers (EMIs). local ayuda had often been provided, but starting from the lockdown period of April 2021 in NCR and As a result, SAP payments ended up using three surrounding regions, LGU-led ayuda became the different channels for delivery. For 4Ps households, major source of support as centrally-managed large SAP payments were delivered through their existing scale cash assistance came to an end. The ayuda 4Ps payment channel, in most cases LBP cash cards; package was approximately ₱1,000 for an individual for non-4Ps households receiving SAP 1 payments, or ₱4,000 for a family, or in-kind assistance with an cash was delivered by LGU officials; and for non- equivalent value. A typical ayuda package included 4Ps households receiving SAP 2 payments, digital rice and noodles, canned goods (e.g., sardines, fruits, transfers through FSP partners in addition to physical and sauces), light snacks, and instant foods and cash delivery were used. Responsibility for digital drinks. In addition to the food package, some LGUs delivery of SAP 2 grants was assigned distributed a hygiene package including masks, hand 25 With the time pressure for assistance, a ‘pay first, verify later’ approach was generally condoned. See this article for instance. https://www.rappler.com/ voices/thought-leaders/258072-analysis-pay-now-verify-later-subsidy-program-coronavirus/ 26 See Cho et al. 2020a for discussions of the implications of delay in SAP on the well-being of low income households. 27 The government intended to provide support for each household. Few mechanisms existed to prevent multiple family members within the same household from receiving benefits. 28 See the case study on SAP 2 (Cho et al. 2021) for details. 29 As of February 2021. 30 DSWD originally intended to conduct a centralized review of applications but due to the urgency of the situation, it delegated the responsibility to LGUs. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 11 sanitizers, and alcohol. While anecdotal evidence the COVID-19 Bayanihan I and about 1 million workers suggests that the ayuda is predominantly in-kind, few crisis (supporting about 400,000 workers under data on the share of in-kind versus cash exist. Bayanihan II). In addition, the government launched two major wage subsidy programs, namely, the Worker protection programs COVID-19 Adjustment Measures Program (CAMP) The government launched several new programs and the Small Business Wage Subsidy (SBWS) to support formal and informal sector workers. Program.32 For CAMP, the DOLE provided a one- Under the Abot Kamay ang Pagtulong (AKAP) time financial assistance of ₱5,000 (US$100) to about Program, the Department of Labor and Employment 660,000 workers (under Bayanihan I) and 1.5 million (DOLE) provided US$200 to stranded migrants.31 As workers (under Bayanihan II) in the formal sector. of December 2021, financial assistance totalling ₱5.5 On the other hand, the SBWS provided ₱5,000-8,000 billion was extended under AKAP, benefitting 540,876 (US$100-160) twice to workers in small businesses displaced overseas Filipino workers (OFWs). DOLE’s affected by lockdown measures. The Financial Subsidy TUPAD programme provided temporary employment for Rice Farmers Program provided ₱5,000 to about to informal sector workers who lost their jobs due to 600,000 smallholder rice farmers (World Bank, 2021). Photo by: junpinzon 31 In addition, the Philippine Overseas Labor Offices (POLOs) provided an extra US$200 to migrants who tested positive for COVID-19. Migrants also received support such as hygiene kits and food packs from the POLOs. 32 In the guidelines for the SBWS Program, small businesses are sole proprietorships, corporations, or partnerships not under the jurisdiction of the Bureau of Internal Revenue (BIR) Large Taxpayers Service. The SBWS is a joint undertaking of the Department of Finance, Social Security System, and BIR. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 12 3. Assessment of SA Performance This note uses data from a variety of surveys experience of using digital payment mechanisms for and draws from the findings of multiple studies SAP 2, the World Bank and Innovations for Poverty (Table 2). In the section describing the impact of Action (WB-IPA) survey was used. COVID-19 on the welfare of low-income households, the primary data sources are the Household Panel 3.1 Coverage and Adequacy and Economic (HOPE) survey of poor and near-poor Nearly all households received some form of households from across the country tracing the social assistance in the first half of 2020 based same households over time, and the High Frequency on the APIS 2020.34 The most common forms of Monitoring (HFM) survey of a representative sample social assistance were non-DSWD government relief of households in the Philippines. The HOPE survey (mainly food aid distributed by LGUs) at 85 percent was conducted over five rounds between December of households, SAP (excluding 4Ps beneficiaries) at 2019 and October 2020 (see Cho et al 2020b for a 42 percent of households, non-government aid at detailed description of the sample). The findings 24 percent of households, and 4Ps at 15 percent of were corroborated by the nationwide HFM survey households (Table 3). There is little variation across conducted by the World Bank over three rounds regions in the coverage of LGU food aid, indicating in August 2020, December 2020, and May 2021. In a similar level of support provided by LGUs. The the sections on coverage and targeting accuracy of average amount of support was about ₱1,700, which social assistance programs, we use the World Bank’s was far below the monthly subsistence level of HFM survey as well as data from the Annual Poverty households. Indicator Survey (APIS) 2020.33 Finally, for beneficiary Table 2. Data sources of COVID-19 impacts on households and SP program availment Data Source Sample Coverage Data during the pandemic Description Philippine Statistical Nationwide/households • APIS 2020 • Household welfare measures and Authority (APIS) SP programs • Quarterly LFS (monthly from Nationwide/households/adult January 2021) • Individual/household labor market individuals (LFS) activities Social Weather Stations Nationwide/adult individuals Quarterly SWS pulse survey Public sentiment on various topics (SWS) such as food security, vaccine hesitancy, and government response World Bank Nationwide/households High Frequency Monitoring Complementary to APIS and LFS, Survey with more detailed information • R1: July 2020 on COVID impact and coping • R2: Dec 2020 mechanisms, SP support, and health • R3: May 2021 and education situation World Bank Nationwide/low income Household Panel and Economic Focusing on low income households households Surveys with a sample of 4Ps beneficiaries • R0: Dec 2019 and non-4Ps comparable/low income households; capturing the initial • R1: April 2020 two phases of the pandemic; and • R2: June 2020 comparing with the pre-pandemic • R3: August 2020 period of late 2019 • R4: October 2020 World Bank and Four regions/recipients of WB-IPA survey in April 2021 Focusing on experience of Innovations for Poverty digital payments of SAP 2 beneficiaries who received SAP 2 Actions (WB-IPA) digitally 33 APIS and Family Income and Expenditure Survey (FIES) are both official, regular large household surveys conducted by the Philippine Statistics Authority (PSA). FIES collects detailed data on consumption, income, and expenditure. It is conducted every three years while the APIS survey collects self-reported and simplified data annually except the years of FIES. 34 Surveyors explicitly asked about regular social assistance and Bayanihan emergency programs by name and implementing agency. In cases where a household received relief from an LGU, the surveyor probed to determine whether the relief was distributed by DSWD. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 13 SAP was provided to close to 60 percent of total means a very small amount, is pursued in the face of a households (4Ps 15 percent and non-4Ps SAP 42 deep crisis. Compared with other countries, the share percent). The overall coverage of 4Ps is 21 percent of Filipinos who received a COVID-19 cash transfer is when translated to the share of population, which is relatively high based on the SAP1 administrative data. similar to the share found in the 2018 FIES (about 23 Even if we consider the lower estimate from APIS percent) and 2017 APIS (about 21 percent).35 DSWD at about 57 percent and exclude all other COVID-19 official records show that 4Ps covers about 4.2 assistance, the proportion of the Filipino population million households representing about 18 percent of that received SAP is higher than the share of people household coverage. Receipt of SAP, which reflects receiving cash transfers in many other countries. the coverage of the first tranche, is substantially lower than the official figure of 16.7 million households (as Both TRAIN-UCT and SocPen show very low of February 2021 record), equivalent to 76 percent coverage in the APIS (Table 3 rows b and c). The coverage. The World Bank HFM survey conducted administrative data suggest that these programs in July 2020 shows about 75 percent coverage of should cover 2 million households and 3.4 million cash transfer (reflecting the coverage of SAP 1). Low individuals, respectively. The low level of reporting reporting of SAP in the APIS 2020 may be due to the the receipt of the benefits may have to do with the timing of the survey, which was conducted in early delivery challenges. Both programs provide cash July, when a small share of households were still assistance by delivering physical cash, and thus receiving SAP 1 (Cho, Johnson, et al. 2020). Confusion the payments have been infrequent. Unlike 4Ps on households versus family, and receipt of benefits that transfers grants bi-monthly (once every other by multiple members in the same households may month), these programs transfer their respective also contribute to lower coverage in survey relative to grants of ₱300 and ₱500 per month once or twice the administrative data. The average reported amount a year, varying largely by region. Recognizing the of SAP is about ₱6,100, closely reflecting the official challenge of physical cash delivery and infrequent amount set between ₱5,000 and 8,000. payment, DSWD started distributing the LBP cash cards to these beneficiaries similar to 4Ps. However, Countries that experienced a greater shock little is known about the status of digital payments (measured by growth rate of per capita GDP of TRAIN-UCT and SocPen through LandBank cash in 2020) tend to have a greater coverage of cards. Some other programs surveyed include the COVID-19 cash transfers as a share of population individual medical assistance, student financial (Figure 3). This indicates the political economy of assistance, and emergency shelter assistance, all of social assistance that a large coverage, even if this which have very limited coverage. Figure 3. Coverage of COVID-19 cash transfers and level of development Source: Log GDP per capita growth figures are from the January 2022 issue of the World Bank’s Global Economic Prospects database. Data on the share of population receiving COVID-19 cash transfers in 2020 are from Johnson and Palacios (2022). Note: PHL=Philippines; IND=Indonesia; MYS=Malaysia; THA=Thailand; VNM=Vietnam; and CHN=China 35 See Acosta, Avalos, and Zapanta (2019). COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 14 Table 3. Social assistance support to households between January and June 2020 based on APIS 2020 Recipient Recipient Average amount per recipient Program individuals households (%) household (Jan-Jun 2020, ₱) (%) Regular social assistance (selected major programs) a. 4Ps 15.2% 20.8% 9,129 b. UCT (under tax reform) 2.1% 2.5% 3,017 c. Indigent Senior Citizen’s Social Pension (SocPen) 9.7% 8.4% 4,018 Social assistance under Bayanihan Act d. SAP (excluding 4Ps) 41.9% 41.8% 6,176 e. DOLE TUPAD 0.7% 0.8% 3,987 f. DOLE COVID-19 (CAMP) 0.6% 0.7% 5,383 g. DOLE AKAP 0.1% 0.1% 10,441 h. DTI Livelihood Seeding Program 0% 0% 4,293 i. DA Rice Farmers Assistance 2.9% 3.1% 4,967 j. DSWD Relief Assistance 22.0% 22.7% 940 k. Relief Assistance from Government (other than DSWD) 85.5% 86.2% 1,712 l. Relief Assistance other than from Government 24.3% 24.9% 1,070 m. Any social assistance 96.0% 97.0% N/A Source: APIS 2020 Note: DTI=Department of Trade and Industry; DA=Department of Agriculture Despite the reported high coverage of social households reporting a decline in per capita income assistance through regular programs and varied significantly by region (Figure 4). Households Bayanihan laws, many households experienced in Luzon that were hit hard by the pandemic (e.g., large declines in total income. Two-thirds of NCR or regions IVA and V) and already vulnerable households experienced a decline in total per capita regions (e.g., BARMM) reported large income income despite social assistance between the declines. The share of households experiencing first and second quarters of 2020, and a tenth of income decline was also higher among lower income households experienced a decline in total per capital and female-headed households, based on round 1 of income of 50 percent or more. The share of the World Bank HFM survey in July 2020. Photo by: rene_salta COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 15 Figure 4. Share of households reporting income decline by region CALABARZON - IVA BARMM NCR Bicol - V Central Luzon - III Ilocos Region - I Western Visayas - VI Cordillera Zamboanga Penin sula - IX Cagayan Valley - II MIMAROPA - IVB Eastern Visayas - VIII Davao - XI SOCCSKSARGEN - XII Central Visayas - VII Northern Mindanao - X Caraga - XIII 0 10 20 30 40 50 60 70 % households experiencing income decline Source: APIS 2020 Figure 5. Mental health and food insecurity 35% 90.0 32% 32% 80.0 30% 70.0 25% Mental Health Index 60.0 19% 50.0 20% 17% 40.0 15% 30.0 10% 20.0 10.0 5% 0.0 1 2 3 4 5 0% Severe depression Moderate depression Mild depression Not depressed (food secure) (food insecure) (0 to 52) (0 to 52) (0 to 52) (above 68) Food Insecurity Index Source: HOPE survey in August 2020 adopted from Cho et al. 2020b The large income declines were accompanied by economic hardship and food insecurity, a less visible a sudden and severe rise in hunger, along with stress that the households suffered included mental mental health issues. Data from the Social Weather health issues. The HOPE survey in August 2020 Stations show that the share of households reporting shows that almost a third of surveyed low income hunger came down to under 10 percent prior to the households reported severe depression and mental crisis but soared to 31 percent in September 2020. health declining with food insecurity (Figure 5). Data from the HOPE and HFM surveys also reveal high rates of hunger and food insecurity throughout A straightforward assessment of how SA 2020. The vast majority (>80 percent) of HFM survey assistance helped cash-strapped households respondents in July 2020 who reported that they cope with the pandemic’s impact is difficult to were unable to purchase food cited lack of money undertake. An analysis based on the HOPE survey in as the main reason, rather than mobility restrictions, April 2020, when most 4Ps beneficiaries had received increases in prices, or closed markets. Along with their top-ups but non-4Ps beneficiaries had not COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 16 Photo by: Rey Borlaza received their SAP 1 benefits, suggests that the Despite SA support, the poverty rate is expected timely provision of SA support helped beneficiaries to rise significantly due to the pandemic. Prior to with food security during the lockdown. The results COVID-19, poverty was on a downward trend from suggest that 4Ps and non-4Ps experienced similar 23.5 percent in 2015 to 16.7 percent in 2018. As the labor market shocks, but 4Ps households were less crisis and economic disruptions continued, poverty likely to report food insecurity: 53 percent of 4Ps deepened despite the modest level of mitigation households reported food insecurity compared resulting from large-scale social assistance. The PSA with 60 percent non-4Ps. The gap between the two estimated that the poverty rate of the first semester groups widens to almost 10 percentage points when of 2021 was about 23.7 percent. This means that pre-pandemic covariates are added in the regression. almost 26.7 million individuals from 4.74 households When disaggregated by pre-pandemic per capita were living in poverty and approximately 7 million household earning quintile, the results indicate that more individuals living in poverty compared with that SAP emergency top-ups helped the 4Ps households of 2019.36 A World Bank estimate suggests that the cope better with the shock, especially those in the poverty rate would have been almost 2 percentage lower end of income distribution. This underscores points higher in 2020 if it were not for government the importance of having a functioning SP program assistance (Belghith 2021). Nonetheless, given the already in place before the shock arrives and sheer magnitude of the shock and the prolonged providing timely support when the households are in pandemic, SA support appears to fall short of the greatest need. daunting needs of households. 36 https://psa.gov.ph/infographics/subject-area/Poverty COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 17 3.2 Targeting The most common indicators used in assessing (Figure 6; see also Box 1 for discussion on the current targeting performance are benefit incidence, targeting systems in the Philippines). During the beneficiary incidence, and coverage.37 Benefit initial period of 4Ps and before the nationwide scale- incidence indicates the percentage of benefits up, targeting performance was excellent based on reaching each quintile of the welfare distribution (e.g., the recently completed Listahanan 1 with over 90 poorest quintile). Similarly, beneficiary incidence percent of beneficiaries belonging to the bottom two indicates the percentage of beneficiaries in each quintiles. In 2013 the share of beneficiaries belonging welfare quintile. Coverage indicates the share of to the bottom two quintiles was 83 percent, but it households (individuals) in each welfare quintile who gradually decreased to 71 percent in 2018. According are receiving benefits. If households were all the same to APIS Q1 and Q2 of 2020, only 65 and 62 percent size and benefit amounts did not vary by household, of 4Ps beneficiaries belong to the lowest two income benefit and beneficiary incidence should be identical. quintiles. On the flip side, an increasing share of Well-targeted programs expect benefit and beneficiary 4Ps beneficiaries now belong to higher income incidences to be progressive and concentrated around groups and they may no longer need government the poor. Coverage, along with the generosity of the assistance. While the improving welfare status of 4Ps benefits, depends on the available fiscal space for SA beneficiaries is encouraging, this highlights the need support and captures more than targeting accuracy. for the recertification of existing beneficiaries and However, if targeting accuracy improves and finite enrolment of the new poor. As these processes are not resources are concentrated on the poor, the coverage taking place, the ability of 4Ps to provide assistance of lower quintile households would rise. to low income households has also been diminishing over time (Figure 6). For instance, since 4Ps became The targeting performance of 4Ps, measured as the a nationwide program, it was able to provide benefits share of beneficiaries belonging to the bottom two to over 60 percent of the lowest income households quintiles (lowest 40 percent), has been decreasing in 2013, but only 35 percent of the lowest income and the trend continued during the pandemic households benefitted from 4Ps in Q2 2020. Figure 6: Distribution of 4Ps beneficiaries and 4Ps coverage over time Distribution of 4Ps beneficiaries by income quintile over time 100% 90% 80% 17% 70% 29% 60% 31% 31% 30% 50% 30% 27% 40% 75% 30% 54% 20% 45% 46% 41% 39% 35% 10% 0% 2009 2013 2015 2017 2018 2020 _Q1 2020 _Q2 Q1 Q2 Q3 Q4 Q5 Share of households in each income quintile supported by 4Ps 70% 60% 50% 40% 30% 20% 10% 0% Q1 Q2 Q3 Q4 Q5 2013 2015 2017 2018 2020 _Q1 2020 _Q2 Source: 2009, 2015, 2018 FIES and 2013, 2017, 2020 APIS from Acosta et al. (2019) and Acosta and Velarde (2015) for years up to 2018, authors’ calculation for 2020 37 We also considered other indicators as specified in Annex A. Key findings are similar to the outcomes presented here. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 18 Box 1. Targeting Systems in the Philippines The National Household Targeting System for A new standardized targeting system envisaged in Poverty Reduction (NHTS-PR or Listahanan) is an the 4Ps Act has not been developed. The Act’s information management system used to identify Implementing Rules and Regulations (IRR) specify who and where the poor are (DSWD, 2019). It was the adoption of a standardized targeting system designed to identify poor households that will be (STS) for beneficiary identification with a regular prioritized for government assistance programs, revalidation of beneficiary targeting every three especially 4Ps. The list of the poor is used by 59 years. Until the STS is set up, the Listahanan national government agencies including several large continues to be the main targeting mechanism. programs, such as the Social Pension, TRAIN-UCT, However, with the delay of L3, the establishment of and the PhilHealth subsidized health insurance an STS or a validation and update of records were program. It is also used by local government units also delayed. At the same time, the Community- (LGUs) and other institutions. In 2020, the Philippine Based Monitoring System (CBMS) Act under Statistics Authority (PSA) used Listahanan to Republic Act No. 11315 of 2019 and its IRR adopted in prioritize the first registrants for the Philippine 2020 requires national government agencies to use Identification System (PhilSys). CBMS data (collected by LGUs and managed by PSA) in prioritizing the beneficiaries of social Listahanan follows a census sweep approach protection programs with updated data every three where all or most households in specific localities years. Although data generated through CBMS could are surveyed en masse. Listahanan uses three provide more up-to-date socio-economic information steps to identify poor households: 1) geographical about Filipinos and those data can be used in prioritization; 2) welfare ranking using a Proxy Means Listahanan to identify the beneficiaries of 4Ps and Test (PMT) method to predict the household’s level of other SP programs, the coordination mechanism income and classify them into poor, near-poor, and between CBMS and Listahanan/STS is not defined. non-poor by applying the official provincial poverty Compounding the policy disconnect as to which threshold; and 3) community validation by DSWD government agency has the mandate in identifying and LGUs, during which households and the poor was the passage of the Magna Carta for the communities can submit grievances (e.g., for Poor (Republic Act 11291 of 2019). The law provides misclassification, not surveyed) to be addressed (e.g., for the development of a single system of re-survey, re-application of PMT). By incorporating classification to be used for targeting the households’ key characteristics – demographic beneficiaries of the government’s poverty alleviation information, living conditions, and assets, Listahanan programs and projects. The responsibility to identify enables geographic, demographic, and welfare target beneficiaries was given to DSWD, National targeting based on the relevant information captured Economic and Development Authority, and National at the time of the census sweep. Anti-Poverty Commission. Listahanan census sweeps were conducted in The STS will need to incorporate various 2010-11 (L1), 2014-15 (L2), and 2019-2021 (L3). institutional developments for an objective Listahanan was originally designed to support intake targeting based on dynamic information updates. and registration for 4Ps through a pilot registration in The objectivity of the targeting is achieved by a 2007 but was expanded to a nationwide census scientific and data driven approach in assessing sweep in L1. L1 and L2 collected data for 10.9 and 15.2 poverty and vulnerability and identifying potential million households (around 50 and 70 million people) beneficiaries for social assistance. At the same time, and identified 5.2 and 5.1 million poor households, flexibility and dynamism can be promoted by frequent respectively (Fernandez 2012; Velarde 2018). As updating of information, especially those related to census sweeps were planned every four years (based changes in demographic and socioeconomic status. on Executive Order 867 Series of 2010), the L3 census The former (i.e. objectivity of the targeting) can be sweep was expected to be completed around 2019 done at the central level including through Listahanan with the aim of covering 70 percent of the total while the latter (i.e., flexibility and dynamism of the population. However, implementation experienced targeting) can also be done at the local level including delays and was further disrupted by the COVID-19 through CBMS. Progress in digital technology and pandemic. L3 data collection was completed in data governance, along with the use of PhilSys, are November 2021, and it was at its final stage of key to balancing these two. As more up-to-date community validation by the year end. DSWD records information and data become available, the census show that L3 surveyed over 14 million households, sweep and PMT-based targeting method that uses but many grievances about exclusion complaints a statistical model of a limited set of observable were received. variables should be revisited. COVID-19 and COVID-19 and Social Social Assistance Assistance inPhilippines: the Philippines: in the Lessons Lessons for for Future Future Resilience Resilience 19 Nonetheless, 4Ps demonstrates the best targeting The high frequency monitoring survey conducted accuracy among various programs in the country in July and December 2020 captures the progres- (Figure 7).38 Other regular SA programs including sivity and coverage of government support. The SocPen and TRAIN-UCT show similar progressive round 1 survey covers the assistance from govern- targeting. Indeed, beneficiaries from these programs ment between March and July 2020, and round 2 were identified through L2. While the SAP and other covers August to December 2020. Assistance was pandemic relief had a long list of eligibility criteria more progressive during the earlier period of the combining categorical conditions and welfare pandemic than in the later period (Figure 9). As the conditions, there were few methods to verify such pandemic continued and the government support information. Share of households in each income for the poorest quintile diminished, nongovernment quintile covered by these programs (Figure 8) shows organizations and religious organizations appeared that there is a wide variation in the progressivity and to have stepped in (Figure 10). targeting performance of SA programs. Figure 7: Benefit incidence of various programs in Jan-Jun 2020 Share (%) of benefits going to each income quintile by program in 2020 100% 90% 80% 70% 60% 50% 27 40% 30% 21 20 21 19 20% 16 33 10% 20 17 22 16 12 0% 4Ps Other regular SAP DSWD Relief Non-DSWD Non-Gov Q1 Q2 Q3 Q4 Q5 Figure 8: Coverage of various programs in Jan-Jun 2020 Share of households in each income quintile supported by program in 2020 100 90 80 70 60 50 40 30 20 10 0 4Ps Other regular SAP DSWD Relief Non-DSWD Non-Gov Q1 Q2 Q3 Q4 Q5 Source: Figures 7 & 8 from APIS 2020 Note: Programs examined here include 4Ps, other regular programs (e.g., social pension, TRAIN-UCT), SAP excluding 4Ps, other DSWD pandemic relief, non-DSWD programs (e.g., LGU relief, DOLE programs, farmers’ subsidy, and other government’s programs), and non- government support. 38 Typically, analysts subtract the value of a transfer when calculating household welfare for the purpose of assessing targeting accuracy. As APIS income data is very noisy we instead use the raw income measure. All of these figures use unadjusted reported income from January to June 2020 and thus may understate targeting accuracy especially for programs with large transfer amounts. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 20 Figure 9: Coverage of government assistance by Figure 10: Coverage for the poorest households by income quintile type of assistance Received government assistance (%) Received assistance from selected coping mechanisms: Poorest Income Quintile 87.3 90 81.5 82.9 80.0 78.2 78.3 77.4 73.3 80 69.1 70 60.9 60 50 40 30 20 10 Q1-poorest Q2 Q3 Q4 Q5-richest 0 Government Friends and family NGO/Church Wave 1 (since March) Wave 2 (since August) Wave 1 (since March) Wave 2 (since August) Source: WB HFM survey in July and December 2020 Given the large coverage of SAP, one may think the question of feasibility of a Universal Basic Income that a universal rather than targeted program (UBI) in the Philippines (Arriola 2021). However, would have made sense. In such case, the program providing support to 23 million instead of 18 million could have avoided all errors of exclusion as well households would have required an additional as the cost and issues associated with processing budget of around US$1-1.6 billion (US$100-150 program applications. Given the urgent need for given twice to 5 million additional households). a fiscal stimulus, it is not uncommon for countries Moreover, in an environment where information to adopt (near) universal subsidies. In response to about individuals, families, and households is difficult the COVID-19 crisis, governments of 12 countries to verify and digital payments are limited, universal provided one-off universal cash transfers to citizens assistance could have brought significant risks of (Gentilini et al. 2022), and some have raised leakages with duplicative receipts of benefits. Photo by: Nomad1988 COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 21 3.3 Financing and Delivery The country’s COVID-19 response, while The Philippines has a few financing options being swift in the budget processing, brought that can be used in case of disasters, but they to light significant gaps in the SP delivery have not been scalable or quick to provide cash systems. Most processes including application assistance to affected families.39 The General for programs, beneficiary enrolment, and cash Appropriations Act in 2020 provides a legal basis transfer payments were carried out manually. The through which regular social protection programs Philippines lacked updated social and beneficiary such as 4Ps and social pension find the budget by registries and a national ID system, both of which the annual congressional approval. The same Act are crucial for countries to expand their support includes a provision on calamity funds—National quickly (Palacios, 2020). When there is accurate Disaster Risk Reduction and Management Fund and up-to-date information relevant in determining (NDRRMF) so the fund can be used for various the basic socioeconomic status of households, the purposes including relief, recovery, reconstruction government can more readily identify potential and other work or services in connection with beneficiaries. An inclusive and trusted national natural or human induced calamities and disasters. digital ID system40 can help by facilitating secure About 30 percent of NDRRMF is allocated as identity verification and deduplication of potential Quick Response Funds, a standby financing beneficiaries. However, the Philippines had neither mechanism that agencies such as DSWD can use the registries nor the ID system when the pandemic for their disaster responses. Previous experience happened. however suggests that the administrative process of triggering and appropriating the funds can be Lack of digitized information of beneficiaries lengthy, with delays of up to one year or more limited the use of digital G2P payment reported against the prescribed timeline for national mechanisms and delayed digital payments. government agencies (NGAs) of 15 to 30 days (Qian Learning from the experience of manual and et al. 2020). physical delivery of SAP 1, the shift to digital payments through partnerships with multiple FSPs Local governments also maintain about 5 percent was a major milestone. The country’s financial of their revenue as contingency funds, which sector ecosystem, which has been developing can be supplemented by NGAs’ funds in case and evolving rapidly, could provide a conducive needed. These funds are often used to finance environment for digital G2P payments. For instance, food/in-kind assistance, but are not sufficient to inter-bank payment systems (e.g., PhilPaSS) and finance large scale cash-based interventions. It was national retail payment systems (e.g., InstaPay and only in 2019 that emergency cash transfers (ECT) PESONet) 41 enable account holders to receive, pay in the case of disasters were formally established and transfer funds securely and promptly across and LGUs and NGAs were authorized to use different financial institutions and affiliated agents.42 funds from the Quick Response Funds and other Despite this delivery option, there were significant sources, for cash assistance purpose. Although the delays and challenges in the digital payment of ECT guidelines include a range of funding options SAP 2, with the delays reaching up to six months in including contingent financing, these options have some cases (Australian Aid et al. 2021). The problem not been widely used to date and DSWD has yet was attributed to the use of paper-based forms in to mainstream ECT. Meanwhile, local governments gathering beneficiary data in the first place as the continue to stock up and rely mostly on food information had to be encoded first to enable digital and in-kind support rather than cash assistance. transactions. However, when the digitalized COVID-19 response resorted to a new series of laws data became available, data quality issues (e.g., (Bayanihan I and II) to allow the reallocation of the duplicate mobile phone numbers and beneficiary budget to finance SAP cash transfers instead records) were raised, which prevented seamless of activating ECT or using existing financing financial transactions (WB-IPA 2021). Verification of mechanisms. information and deduplication took some time, so 39 Financing options include national and local contingency funds for more frequent but smaller scale disasters and insurance and risk transfer mechanisms for less frequent but greater scale disasters. 40 If not a digital and foundational ID, a functional ID that is widely used could also facilitate identity verification and deduplication of potential beneficiaries. For instance, Togo used the country’s voter ID for identifying and deduplicating beneficiaries of COVID-19 assistance. See Aiken et al. (2022). 41 See Acosta et al. (2019). 42 PESONet and InstaPay were used to transfer grants to beneficiaries who registered their account via an online portal, while PhilPass was used to transfer batch grants from LBP to partner FSPs that did not have accounts with LBP. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 22 that unresolved cases had to rely on physical cash trip travel time was on average less than 30 delivery. minutes, but the wait time was quite long. The SAP 2 beneficiaries using digital payment channels had Notwithstanding the challenges associated to cash out their benefits at FSPs’ contact points with data governance, once digital payments including bank branches and ATMs (for brick-and- were made, delivery efficiency and beneficiary mortar banks) and local agents (for EMIs). There is a satisfaction improved. The WB-IPA (2021) shows large variation by FSP (Figure 11, right). Beneficiaries a high degree of beneficiary satisfaction and ease who received their SAP 2 payments through two of access among SAP 2 beneficiaries who received of the EMIs spent significantly less time accessing payments digitally. The study also found that their payments. This suggests that if beneficiaries average total time costs for traveling and waiting could choose an FSP that provides the nearest and for services were lower in SAP 2 than in SAP 1. most convenient cash-out point as well as ensures This was despite the fact that many beneficiaries that the cash-out point fulfills the required service had to travel beyond their barangays to get to the standards, the beneficiaries’ travel and wait time designated FSP’s contact points to receive SAP 2 costs can be significantly reduced (WB-IPA 2021). payments and the traveling time was indeed longer for SAP 2 than SAP1. Note that beneficiaries did not The financial account through EMIs or banks have the option to choose their FSPs and instead for SAP 2 could have provided a way for greater had to use the FSPs assigned to serve their locality. financial inclusion. A large number of beneficiaries opened the restricted purpose financial account The comparison of three payment channels in that could be used beyond receiving SA benefits terms of beneficiary travel and wait time provides once converted into a regular account. However, an important insight for G2P payments policy as the large scale cash assistance at the central direction (Figure 11, left). 4Ps beneficiaries rely level phased out, such an opportunity was not almost entirely on Land Bank ATMs, but these ATMs utilized. Relying mostly on the local ayuda meant are quite limited—on average there are about five an increase in food-based aid whose delivery Land Bank ATMs for 10,000 beneficiaries (Cho et al. has implications on the efficiency and impact of 2020). This could explain the significant amount of assistance, as discussed in the following section travel and wait time. SAP 1 physical cash delivery (see Box 2). was mostly done in barangay offices, so the round Figure 11. Mean travel and wait time by payment channel 300 300 250 250 200 200 Minutes Minutes 150 150 100 100 50 50 0 0 4Ps pre-COVID SAP R1 (4Ps) SAP R1 (non-4Ps) SAP R2(FSPs) EMI 1 EMI 2 Bank 1 Bank 2 EMI 3 Bank 3 Travel time (both ways) Wait time + payment processing time Travel time (both ways) Wait time + payment processing time Source: Data on travel and wait time for all channels except for SAP 2 come from the HOPE survey. Data on the travel and wait time for SAP 2 are from the WB-IPA survey. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 23 Box 2. Efficiency of Cash versus In-Kind Transfers One of the most basic decisions when designing criteria, thus the administrative costs of identifying a social assistance program is the type of benefit the beneficiaries and verifying the eligibility is small. to provide, often either cash or in-kind. In the Meanwhile, social protection delivery systems, past, most social assistance programs in both which determine the ability and ease of identifying developed and developing countries provided beneficiaries, verifying eligibility, and transferring in-kind benefits (Cunha 2014). Recently, more cash assistance, have been weak in the Philippines countries including the Philippines are choosing to and this state could explain the delays in cash disburse cash rather than in-kind benefits in social delivery. assistance programs. Evidence suggests that, in most contexts, cash is faster and less expensive While data on the cost of delivering cash and to disburse (Gentilini 2016). Further, defeating a food during the COVID-19 crisis are limited, common myth that cash transfers may lead to data from 4Ps prior to COVID-19 as well as increased spending on “temptation goods” such procurement data suggest that in the Philippines, as alcohol and tobacco, evidence suggests the similar to other countries, cash is much less opposite, indicating that the inflexibility of food aid expensive to deliver than food. Researchers is a weakness rather than a strength (Gentilini 2016; estimated the cost of delivering cash to 4Ps Evans and Popova 2017). Nevertheless, the relative households at 1.2 percent of the total value of the advantages and disadvantages of each form of transfer—one of the lowest delivery costs of any benefit depend greatly on context. major cash transfer program (Bowen 2015). In contrast, procurement estimates indicate that the During the COVID-19 crisis in the Philippines, cost of packaging and transporting food aid to unlike in other countries, food aid reached LGUs is approximately 13 percent of the total cost households significantly more quickly than of the aid (DSWD 2021). These estimates do not cash payments. The first round of the HOPE include the cost of local officials’ time incurred in survey conducted between the end of March and delivering food aid, which will further raise the cost early April found that 90.5 percent of respondent of delivering aid (though the estimates also do not households received food aid from the government. take into account potential differences in local retail A survey by the Social Weather Stations around and national wholesale prices of food items, which the same time (May 4 and May 10) found that 99 might slightly reduce the cost of food aid). These percent of all households received government food data, while incomplete, suggest that the cost of aid. 43 Yet another household survey conducted by delivering food aid is likely substantially higher than IPA in late June found that 86 percent received food the delivery of cash. In particular, the cost of digital aid from the government. These data suggest that payments of cash should be quite small. food aid reached a majority of households within two weeks of the first lockdowns in late March. By Food aid is preferred in the context of high contrast, it took a few weeks to several months for food insecurity and local market disruptions SAP beneficiaries to receive their cash. with limited food supply. The flip side is that if food insecurity concern is low or local markets The quick delivery of food aid is likely attributed are functioning well, food aid is generally less to the readily available stock and well- efficient than cash. During the peak of pandemic established delivery mechanisms in response lockdowns in Phase 1, there seemed to be some to frequent natural disasters in the country. For modest disruptions to local food markets. According instance, after Typhoon Yolanda (Haiyan), which to the IPA RECOVR survey carried out in June was one of the largest natural disasters in the 2020, about 60 percent of respondents reported country destroying about 2.7 percent of GDP and difficulty in accessing local markets due to mobility affecting 5.1 million households, family food packs constraints, and 50 percent reported that local were distributed within the first six weeks of the markets were closed (Warren, Parkerson, and response while the first cash transfers were only Collins 2020). However, the HFM survey shows delivered a month after the disaster (Bowen 2015). that as community quarantines eased by August Also, once the locality of distribution is determined, and December 2020, only 15 percent and 6 percent food aid tends to be provided to all households in of households respectively reported mobility the selected locations without individual eligibility restrictions or markets being closed as the main 43 https://www.pna.gov.ph/articles/1103616 COVID-19 COVID-19 and and Social in the Social Assistance Assistance in the Philippines: Philippines: Lessons Lessons for Future Resilience for Future Resilience 24 reasons for their difficulty in buying rice. This needs. Similarly, Gomez and Ignacio (2020) suggests that food aid was appropriate at the start reached a similar conclusion that the packs during of the pandemic, but cash assistance was more the pandemic fell short of meeting nutritional effective as local markets resumed operation. requirements. Some may argue that food aid offers nutritional Taken together, these data suggest that there is value that can contribute to food security. a role for food aid in the initial response to crises However, independent studies found that the to provide support when the market is disrupted family food packs in the Philippines do not meet and cash transfers are not ready. However, when nutritional standards. An analysis (UNICEF 2018) the market is functional and cash transfers can be of the contents of family food packs, typically delivered, cash is superior. To make cash assistance consisting of rice and canned goods, found that more efficient, timely, and responsive, overall the packs were not sufficient to meet nutritional delivery systems should be enhanced. 4. Policy Recommendations The Philippine experience in SA program system could have enabled a relatively quick implementation during the pandemic presents a expansion or introduction of new SP programs to new useful learning opportunity for delivery of future beneficiaries during the pandemic (Palacios, 2020). It social assistance. The country will likely face more is encouraging that the government doubled down crises given its exposure to natural disasters and the commitment and efforts for PhilSys rollout during man-made risks (Behlert et al. 2020). The large and the pandemic. Together with the PhilSys, efforts to varied social assistance response to the COVID-19 open Land Bank accounts for the unbanked crisis allows us to compare different forms of social populations were also made. For these measures to assistance and different methods of targeting and be an effective enabler for SP delivery, there is a need delivering assistance. Using data from a variety of for SP programs to actively adopt and use the sources, we find that the government was able to PhilSys for identification and authentication and to quickly introduce emergency social assistance to a promote digital payments. large proportion of households but that the social assistance measures varied greatly in terms of speed 2. Enhance beneficiary targeting by developing a and targeting accuracy. dynamic social registry and fully adopting PhilSys. Providing the right services and benefits to the right Lessons of experience highlight the need for beneficiaries can maximize the impact of assistance. strong policies, systems and tools. First, it is Even before the pandemic, 4Ps targeting performance critical to have well-established social safety policies, has been declining due to the lack of active programs, and systems upon which emergency recertification and new enrolment. Other SA programs operations can be built. Second, SP delivery were not any better in targeting performance as they systems require regular updates and modernization faced the same challenges as 4Ps in identifying investment and efforts. Third, digital tools and vulnerable populations for assistance. Thus, there is a technological advances should be actively used strong need to improve the targeting mechanism, for SP delivery. Fourth, timely and agile policy making it more dynamic and accessible to citizens. It adaptations building on the lessons learned from can be achieved by creating a social registry that will each shock can help strengthen overall SP systems receive dynamic updates from various databases (e.g., and delivery for future resilience. tax registry, PhilHealth, civil registry). The role of existing systems such as Listahanan and the CBMS Social assistance should be enhanced for better and their synergies should be clearly defined for coverage and adequacy, targeting, and delivery. objective and transparent targeting. Our findings suggest specific policy options to further improve the Philippine social protection 3. Make common and interoperable digital tools systems and help the country be better prepared to and infrastructure available to national and local rapidly provide aid to households that need it most in governments and accessible to citizens. During any future crisis: the pandemic, many local governments delivered their own assistance with limited digitized records. 1. Prioritize the rollout of PhilSys and its adoption Some LGUs introduced their own digital initiatives for SP delivery. A pre-existing foundational ID (e.g., tracing application, digital vaccine certificate, COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 25 or social assistance), but despite good intentions, release order (SARO). The activation of institutional they did not optimize the economies of scale and mechanisms for disaster response including cash some reinvented the wheel. Meanwhile, citizens have assistance should be processed in a timely manner limited access to systems where they can report if they are to be useful. Multiple scenarios of budget their needs and request for government assistance. requirements can be developed ex-ante, using With the introduction of PhilSys and development of historical data to assess the size and type of a shock data ecosystems as well as interoperable information and scope of assistance. Further, the activation systems, there is an opportunity for the central process and mechanisms for financing for central government to lead the establishment of shared and local governments, depending on the scenarios, infrastructure (e.g., shared software, standards, and should be developed. interfaces), which the national and local governments can use to design and deliver national and local SP 6. During a crisis, consider using food aid as a programs. short-term measure but replace with cash as soon as the local market becomes operational. 4. Continue to improve digital G2P payments Delivering food aid locally without much targeting through multiple FSPs that beneficiaries can during a crisis has been so common and widely access. Building on 4Ps and SAP 2 experience, used that the government was able to deliver food the Philippines can leapfrog to the next generation aid more quickly than cash assistance during the G2P framework involving multiple FSPs (banks COVID-19 response. Nonetheless international and e-money issuers) from which customers and evidence shows that cash is far less expensive, beneficiaries can select based on convenience of easily delivered, more transparent and better access. This framework should also enable people to targeted, especially if used with enhanced targeting use the accounts and resources for their transactions and payment mechanisms. Important factors for beyond G2P payments (e.g., purchase, payments, governments to monitor in the food to cash transition savings, remittances). in disaster response include the functionality and prices of the local food market, especially in remote 5. Streamline the administrative process of areas. As long as the markets are able to supply social protection contingency financing for sufficient food and other necessities, and inflation better disaster readiness. The mechanisms for is not of major concern, cash assistance should be contingency financing include NDRRMF and QRF as prioritized over food aid.44 well as budget allocation through a special allocation Source: https://www.facebook.com/dswdserves 44 See Filmer et al. (2021) on the caution against providing cash assistance when the local market function is limited. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 26 References Acosta, Pablo, Jorge Avalos, and Arianna Zapanta. 2019. Pantawid Pamilya 2017 Assessment. World Bank, Washington, DC. https://doi.org/10.1596/32510. Acosta, Pablo, Isaku Endo, Jose Antonio Garcia Garcia Luna, Aisha de Guzman, and Yuko Okamura. 2019. Making Payments More Efficient for the Philippines Conditional Cash Transfer Program. World Bank. https://doi.org/10.1596/32540. Aiken, Emily, Suzanne Bellue, Dean Karlan, Chris Udry, and Joshua E. Blumenstock. 2022. “Machine Learning and Phone Data Can Improve Targeting of Humanitarian Aid.” Nature, March. https://doi.org/10.1038/ s41586-022-04484-9. Arriola, Clara. 2021. “The Basics of Universal Basic Income | Inquirer Opinion.” 2021. https://opinion.inquirer.net/145265/the-basics-of-universal-basic-income. Australian Aid, DSWD, EPRI, and UNICEF. 2021. “Effects of COVID-19 on Child Poverty and Efficacy of the Philippine’s Social Protection Response. Component 2: Rapid Qualitative Assessment of Social Amelioration Programme.” https://www.unicef.org/philippines/media/2701/file/UNIPH-2021- RapidAssessmentSAP.pdf. Behlert, Benedikt, Rouven Diekjobst, Carsten Felgentreff, Timeela Manandhar, Peter Mucke, Ludger Pries, Katrin Radtke, and Daniel Weller. 2020. “World Risk Report 2020.” https://reliefweb.int/sites/reliefweb. int/files/resources/WorldRiskReport-2020.pdf. Belghith, Nadia. 2021. “Poverty and Distributional Impacts of COVID-19 Crisis,” mimeo, World Bank. Bowen, Thomas. 2015. “Social Protection and Disaster Risk Management in the Philippines: The Case of Typhoon Yolanda.” https://openknowledge.worldbank.org/bitstream/handle/10986/23448/ Social0protect0oon0Yolanda00Haiyan0.pdf;sequence=1. Calleja, Joseph Peter. 2020. “Domestic Violence All the Rage in Philippine Lockdown - UCA News.” Ucanews. Com. 2020. https://www.ucanews.com/news/domestic-violence-all-the-rage-in-philippine- lockdown/89821. Cho, Yoonyoung, Jorge Avalos, Yasuhiro Kawasoe, Doug Johnson, and Ruth Rodriguez. 2020a. “Mitigating the Impact of COVID-19 on the Welfare of Low Income Households in the Philippines: The Role of Social Protection.” https://documents.worldbank.org/en/publication/documents-reports/ documentdetail/698921611118950758/mitigating-the-impact-of-covid-19-on-the-welfare-of-low- income-households-in-the-philippines-the-role-of-social-protection. Cho, Yoonyoung, Doug Johnson, Yasuhiro Kawasoe, Jorge Avalos, and Ruth Rodriguez. 2020b. “The Impact of the COVID-19 Crisis on Low Income Households in the Philippines: Deepening Distress despite Rebounding Economy.” https://documents.worldbank.org/en/publication/documents-reports/ documentdetail/768871611898266913/the-impact-of-the-covid-19-crisis-on-low-income-households- in-the-philippines-deepening-distress-despite-rebounding-economy. Cho, Yoonyoung, Yasuhiro Kawasoe, Ruth Rodriguez, and Myra Valenzuela. 2021. “COVID-19 G2P Cash-Transfer Payments Case Study: Philippines.” https://documents1.worldbank.org/curated/ en/368351625163110870/pdf/COVID-19-G2P-Cash-Transfer-Payments-Case-Study-Philippines.pdf. Cunha, Jesse M. 2014. “Testing Paternalism: Cash versus In-Kind Transfers.” American Economic Journal: Applied Economics 6 (2): 195–230. https://doi.org/10.1257/app.6.2.195. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 27 DSWD, 2019. “Listahanan Info Kit,” https://fo1.dswd.gov.ph/wp-content/uploads/2021/01/Listahanan-3-Info- Kit.pdf DSWD. 2020. “DSWD COVID-19 Response and Recovery Plan.” https://www.dswd.gov.ph/issuances/ AOs/AO_2020-003.pdf. Evans, David K., and Anna Popova. 2017. “Cash Transfers and Temptation Goods.” Economic Development and Cultural Change 65 (2): 189–221. https://doi.org/10.1086/689575. Filmer, Deon, Jed Friedman, Eeshani Kandpal, and Junko Onishi. 2021. “Cash Transfers, Food Prices, and Nutrition Impacts on Ineligible Children.” The Review of Economics and Statistics, May, 1–45. https:// 10.1162/rest_a_01061. Gayares, Rhea Molato. 2021. “Philippine Basic Education System: Strengthening Effective Learning During the COVID-19 Pandemic and Beyond.” https://documents.worldbank.org/en/publication/documents- reports/documentdetail/198561622119902647/philippines-covid-19-monitoring-survey-policy-notes. Gentilini, Ugo. 2016. “The Other Side of the Coin: The Comparative Evidence of Cash and In-Kind Transfers in Humanitarian Situations.” 2016. https://documents1.worldbank.org/curated/ en/950781467995054556/pdf/106284-PUB-PUBLIC-ADD-ISBN-Other-side-of-the-coin-cash-inkind- Gentilini-PUBLIC.pdf. Gentilini, Ugo, Mohamed Almenfi, Hrishikesh TMM Iyengar, Yuko Okamura, John Austin Downes, Pamela Dale, Michael Weber, et al. 2022. “Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country Measures,” 907. Gomez, Katrina G., and Ma. Socorro E. Ignacio. 2020. “Adequacy of Food Aid Packs Provided during Natural Disasters: A Provincial Case Study.” Acta Medica Philippina 54 (5). https://doi.org/10.47895/amp. v54i5.2218. Gozum, Ivan, Harvey Capulong, Jose Gopez, and Joseph Galang, 2021. “Philippine Community Pantries as a Way of Helping the Marginalized during the COVID-19 Pandemic,” Journal of Public Health, May. Johnson, Doug and Robert Palacios 2022. “Scaling up cash transfers during the COVID-19 pandemic: what might explain differences in coverage,” mimeo, World Bank Orbeta, Aniceto, Arnelyn Abdon, Michael del Mundo, Melba Tutor, Marian Theresia Valera, and Damaris Yarcia. 2014. “Keeping Children Healthy and in School. Evaluating the Pantawid Pamilya Using Regression Discontinuity Design.” Palacios, Robert. 2020. “Scaling up Social Assistance as Part of the COVID-19 Pandemic Response.” http:// pubdocs.worldbank.org/en/655201595885830480/WB-G2Px-Scaling-up-Social-Assistance- Payments-as-Part-of-the-Covid-19-Pandemic-Response.pdf. Qian, Rong, Benedikt Lukas Signer, Tatiana Skalon, and Zidni Marohombsar. 2020. Public Expenditure Review: Disaster Response and Rehabilitation in the Philippines. World Bank. https://doi.org/10.1596/35064. Rivas, Ralf. 2021. “4.2 Million Filipino Families Stay Hungry in May 2021 – SWS.” 2021. https://www. rappler.com/nation/millions-filipino-families-stay-hungry-sws-survey-may-2021. Tee, Michael L., Cherica A. Tee, Joseph P. Anlacan, Katrina Joy G. Aligam, Patrick Wincy C. Reyes, Vipat Kuruchittham, and Roger C. Ho. 2020. “Psychological Impact of COVID-19 Pandemic in the Philippines.” Journal of Affective Disorders 277: 379–91. UNICEF. 2018. “Filling the Nutrient Gap: Philippines Summary Report.” https://reliefweb.int/sites/ reliefweb.int/files/resources/WFP-0000103298.pdf. COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 28 Warren, Shana, Doug Parkerson, and Elliott Collins. 2020. “RECOVR Philippines: Tracking the Effects of the COVID-19 Pandemic.” https://www.poverty-action.org/recovr-study/recovr-philippines-tracking- effects-covid-19-pandemic. World Bank – Innovations for Poverty Action, 2021. “Monitoring Digital Financial Payments of Cash Transfers in the Philippines,” Washington, D.C. : World Bank Group. World Bank. 2018. “Making Growth Work for the Poor: A Poverty Assessment for the Philippines.” World Bank, 2021. Philippine Economic Update: Regaining Lost Ground, Revitalizing the Filipino Workforce. https://www.worldbank.org/en/country/philippines/publication/philippine-economic-updates COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 29 Annex A. Additional indicators reflecting targeting accuracy The following table provides more detailed leakage • Benefits-cost-ratio – the share of program benefits and performance metrics for each of the major social that goes towards reducing the poverty gap. For protection programs. Definitions of each metric is example, if a household whose income is ₱10 below also included. For metrics that require the poverty the poverty line receives a transfer of ₱15, only ₱10 of line, regional poverty lines from PSA are used and the transfer is included in the numerator of the income is adjusted slightly to ensure that the overall benefits-cost-ratio. share of households that fall below the poverty line • Coady-Grosh-Hodinott – for a given value x matches official statistics. between 0 and 1, the Coady-Grosh-Hodinott metric is the share of households below the x percentile that • Leakage (benefits) – the share of program benefits received the program divided by x. The table uses .2 received by the non-poor for x. • Leakage (beneficiaries) – the share of program • F-score – the harmonic mean of beneficiaries who are not poor 1-leakage(beneficiaries) and 1-undercoverage • Under-coverage – the share of poor households that are not enrolled in the program Leakage Leakage Coady-Grosh- Program Under-coverage Benfits-Cost-Ratio F Score (benefits) (beneficiaries) Hodinott (20%) Regular social assistance (a few selected major programs) 4Ps 52% 52% 70% 30% 1.97 37% UCT under tax reform 43% 45% 95% 52% 2.46 9% Indigent Senior Citizen’s 74% 73% 89% 21% 0.99 15% Social Pension Social assistance under Bayanihan Act SAP (excluding 4Ps) 73% 72% 52% 20% 0.99 35% DOLE TUPAD 75% 74% 99% 20% 1.03 1% DOLE COVID-19 (CAMP) 91% 90% 100% 8% 0.38 0% DOLE AKAP 92% 91% 100% 4% 0.19 0% DTI Livelihood Seeding 83% 75% 100% 11% 0.42 0% Program DA Rice Farmers Assistance 68% 68% 71% 30% 1.28 30% DSWD Relief Assistance 68% 68% 71% 30% 1.28 30% Relief Assistance from Gov 78% 76% 15% 20% 0.80 38% (Other than DSWD) Relief Assistance other than 80% 74% 74% 18% 0.73 26% from Gov COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 30 The World Bank PHILIPPINES 26th Floor, One Global Place 5th Ave. corner 25th St. Bonifacio Global City, Taguig City Philippines 1634 T: +63 2-465-2500 F: +63 2-465-2505 W: www.worldbank.org/en/country/philippines COVID-19 and Social Assistance in the Philippines: Lessons for Future Resilience 31