WORLD BANK SOCIAL PROTECTION POLICY NOTE September 2019 No. 18 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance Pablo Acosta, Jorge Avalos, and Arianna Zapanta This is the fourth benefit incidence analysis of the Philippines’ conditional cash transfer program that uses standard measures to assess the implementation performance of the Pantawid Pamilyang Pilipino Program. The analysis shows that despite the program’s rapid expansion since it was piloted in 2007, it maintains good targeting accuracy, progressivity, and cost efficiency in delivering assistance to the poor. However, the recent halt in program expansion and use of outdated targeting system have resulted in lower coverage levels and incidence rates among the poor. Also, the inability to adjust benefit levels with inflation has resulted in lower generosity of benefits. Still, using the latest nationwide household survey data for 2017, the analysis shows that Pantawid Pamilya helps reduce poverty incidence and income inequality by 1.3 percentage point and 0.6 percentage point, respectively. Adjustments in the benefit level and program coverage are recommended to maintain the program’s relevance, adequacy of assistance, sustained impact on beneficiary welfare. I. Introduction Inclusive growth and poverty reduction remain at the center of the Philippine Development Agenda. The Philippines Development Plan (2017–2022), launched in June 2017, reflects a heightened ambition to lift roughly six million Filipinos from poverty and to achieve upper-middle-income status by 2022. The Government is aiming to leverage the solid position of the Philippine economy (growing at more than 6 percent per year over the last decade) to scale up public investment for poverty reduction, job creation, and economic growth and to accelerate poverty reduction, which has had a less-impressive performance. Over 2006–2012, poverty in the Philippines remained stagnant at 25 percent, despite impressive economic growth in the period. However, between 2012 and 2015 it fell to 21.6 percent, and in addition to the improvements in labor incomes, an important factor contributing to this reduction was documented as obeying government transfers to the poor.1 The Philippines has also undertaken significant steps to build its social protection system and it currently features a much improved—better-designed, coordinated, and implemented and more efficient and effective—suite of social protection programs than it had a decade ago. Since 2010, The Philippine Social Protection Note series aims to disseminate the social protection system also became increasingly more sophisticated, with the country expanding experiences, good practices, and key its conditional cash transfer program, its social pension program, providing livelihood opportunities findings from the Philippines on the to the poor, as well as food and other subsidies to compensate for higher inflation. While effective topics related to social protection. It coverage gaps still exist in contributory and noncontributory programs and the expansion of the social also aims to broaden the dialogue on security system and labor market regulation and policies have been constrained by the structure of the social protection and stimulate public economy with predominantly informal jobs, most programs are at least designed to reach and prioritize engagement in moving forward the policy agenda. The views expressed the poorest and vulnerable populations. here are those of the authors and do not necessarily reflect those of the World Bank. 1 World Bank (2018), “Making Growth Work for the Poor: A Poverty Assessment for the Philippines.” www.worldbank.org 1 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance After 11 years of implementation, the Philippines’ national conditional cash transfer (CCT) program, called the Pantawid Pamilyang Pilipino Program (Pantawid Pamilya), has remained the cornerstone of the social protection policy in the country. Patterned after successful CCT schemes in Latin America, the program provides cash grants to poor households with children 0-18 years old that have limited capacity to invest in their children’s future and well-being. The objective of Pantawid Pamilya is two-fold. In the short- term, it provides income support to poor households helping them afford their basic needs. In the long-term, by conditioning the cash support to desired health and education behavior, it incentivizes investment in the health, nutrition and education of children can be more so they Conditional te of the Philippine Cashproductive citizens in the Transfer’s Implementation future and break away from the cycle of poverty. The program has been recently Performance1 institutionalized. Signed on April 2019, Republic Act 11310 or the “Pantawid Pamilyang Pilipino Program (4Ps) Act” provides stability and continuity to a program that plays a central role in the government’s strategy to accelerate investment in human capital and break the intergeneration cycle of poverty (see Box 1). Since its inception in 2007, the Pantawid Pamilya has expanded rapidly, but its registered beneficiaries has plateaued in recent years. From its pilot year in 2007 with 6,000 household beneficiaries, Pantawid Pamilya has expanded to its peak number of beneficiaries in 2014, reaching 4.4 million households (Figure 1). However, starting 2015, the annual increases in the total regular beneficiary households have been stagnant and, in 2018, the regular beneficiaries 2 fell to 3.9 million— the first time it reached below 4 million since 2013. 2018 2018 to reach 200,000 households . The overall declining Meanwhile, the to reach Modified 200,000 Conditional households Cash 2 Transfer . (MCCT)The overall households, declining number which arenumber ofselected bene not of benefithe ficiaries through ciaries has has been been Listahanan national household re flec targeting ted in database, the haveprogram’s marginally coverage increased of from the 2014total to popula 2018 to reflected in the program’s coverage of the total population that has remained at 21 percent tion reach that 200,000has remained households 2 . at The 21 percent overall declining number of beneficiaries during has period the been reflected in the program’s 2013-2017. As aof coverage share popula of tiof the total tion,population that has remained at 21with percent during during the period the period 2013-2017. 2013-2017. As a share As a share popula on, rela tiverela to tive other tocountries other 3 countries with large 3 of population, relative to other countries with large CCT programs , Pantawid Pamilya has the third large CCT largest CCT programs coverage programs in the 3 , Pantawid , Pantawid world, next Pamilya to Colombia Pamilya and thehas hasMexico, third the each third largest largest covering coverage 22 percentcoverage in of the world, its population. world, in the next 1). In to to next (Figure Colombia Colombia terms of absolute numberand and Mexico, of households, Mexico, each each Philippines covering covering ranks 22 fourth 22 percent of Brazil, behind percent of its popula Indonesia, tion.and its popula ti on. Mexico. (Figure (Figure 1). In terms of 1). In terms of absolute number absolute number of households, of households, Philippines Philippines ranks behind ranks fourth fourth behind Brazil, Indonesia, Brazil, Indonesia, and Mexico. and Mexico. Figure 1. Pantawid Pamilya Program Coverage: 2007-2018 Figure Figure 1. Pamilya Pamilya 1. Pantawid Pantawid Program Program Coverage: Coverage: 2007-2018 2007-2018 Cumulative Number of Registered Beneficiary Coverage (% of total population) Cumulative Number of Registered Beneficiary Coverage (% of total population) Households Households 22% 22% 21% 22% 22% 21% 21% 21% 21% 21% 18% 18% 18% 18% 4,236,739 4,236,739 9% 9% 3,949,855 3,949,855 4 ) 4 ) 7) 5) 4 ) 3) 6) 6) 4) 4) 01 7) 01 5) 01 4) 01 3) 01 6) 01 6) 01 01 6,000 01 01 (2 201o (2 201 s (2 201 il (2 201 a (2 201 a (2 201r (2 (2 6,000 (2 (2 bia s (exic zil ( ine la ( raz na ( mal or ( ntin sia (ado sia 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 bia i c o m e a p a B n e 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 m ex lo ppin M Br ilip tem nti ate uad rge oneEcu do lo M Co i h il ge u c A a In d Regular MCCT Co Ph P Gu Ar G E In Regular MCCT Source: Data Source: Data on beneficiary on bene households ficiary from DSWD households from DSWD Pantawid Pamilyang Pantawid Pilipino Program Pamilyang Pilipino Implementation Program Status Implementa Reports (2012-2018) tion World and Status Reports Bank Update of the Source: Data on beneficiary households from DSWD Pantawid Pamilyang Pilipino Program Implementation Status Reports Philippine Conditional Cash Transfer’s (2012-2018) Implementation and World Performance Bank Update (2007-2011); of the Philippine Coverage Condi data tional from Cash the APIS 2017, Transfer’s World Bank Implementa tionASPIRE and World Performance Bank’s “The State (2007-2011); (2012-2018) of Social Safety Nets and World Bank Update of the Philippine Conditional Cash Transfer’s Implementation Performance (2007-2011); (2018).” Coverage data from the APIS 2017, World Bank ASPIRE and World Bank’s “The State of Social Safety Nets (2018).” Coverage data from the APIS 2017, World Bank ASPIRE and World Bank’s “The State of Social Safety Nets (2018).” Box 1 – Pantawid Box 1 – Pantawid Pamilyang Pamilyang (4Ps) Act(4Ps) Act Pilipino Program Pilipino Program The 4Ps Act is a signi cant milestone for the program and The 4Ps Act is a signi cant milestone for the program and for the social protectionfor the social protection system of system of the Philippines. the Philippines. The The Act is in Act synciswith the with in sync the country’s country’s social pro tectiopro social tection framework n framework and strategy and strategy to protect to Filipinos protect Filipinos from povertyfrom and poverty and reduce vulnerability to risks. Signed on April 17, 2019, Republic Act 11310, an Act reduce vulnerability to risks. Signed on April 17, 2019, Republic Act 11310, an Act institutionalizing the Pantawidins ti tu ti onalizing the Pantawid Pamilyang Pamilyang Pilipino Pro Pilipino Pro gram (4Ps) , guarantees sustainability and con ti nuity of the program, gram (4Ps), guarantees sustainability and continuity of the program, along with its budget, budget, along with its and follows good and follows good global prac global prac tice for tice for social prosocial tectiopro tection ti n interven interven tions to a ons to provide provide a solid solid legal legal founda tiofounda n and tio n and reduce reduce the risk of policy reversals. The law makes Pantawid an added function of the DSWD and provides regular funding funding the risk of policy reversals. The law makes Pantawid an added fun cti on of the DSWD and provides regular from its from its annual annual appropria appropria tion. tion. The provisions The provisions the law of keep of the law thekeep the focus of focus of the program the program onnutrition on health, health, nutrition and education, and education, and that and ensures ensures that 2 The MCCT program poor familiespoor families is designed get toget support reachsupport to to develop out to families develop who are the the human human vulnerable capital andcapital their children. ofassistance in need of of their comprise children. The but are not The provisions provisions covered the ofare in the regular of the law CCTlaw clear are because and clear and they were not captured by the enumeration are based on Listahanan, ofthe lessons the targetingby learned database. The beneficiaries implemen ti ng the program of Homeless for the lastStreet 11 Families years. (HSF), The Indigenous Act covers Peoples all (IPs), and aspects of are Families based in Need on the of Special lessons Protection learned by implementing the program for the last 11 years. The Act covers all aspects of (FNSP). 3 The the the basis program program thoroughly, for determining thoroughly, from the size of the from CCT program program program is theobjec number objec tives, ti ves, targeting of individual targe ti methods,ng beneficiaries. methods, eligibility Source: eligibility criteria, World Bank. criteria, 2018,to “The to convergence convergence State of Social Safety and and Nets 2018”. Only the top 8 coordina largest CCT tio n programs with with other coverageprograms data in the for last gradua 5 years tio are n. The included clarity as and comparison. details coordination with other programs for graduation. The clarity and details of the law reduce the risk of rules being of the law reduce the risk of rules being interpreted interpreted and in and applied applied in an arbitrary an arbitrary manner, manner, increases increases predictability predictability in its in its applica applica tion, on, likelihood andtithe and the likelihood the the 2 program will be able to ful fil l its objectives and achieve desired outcomes. WORLD BANK SOCIAL PROTECTION POLICY NOTE program will be able to fulfill its objectives and achieve desired outcomes. Box 1 – Pantawid Pamilyang Pilipino Program (4Ps) Act The 4Ps Act is a significant milestone for the program and for the social protection system of the Philippines. The Act is in sync with the country’s social protection framework and strategy to protect Filipinos from poverty and reduce vulnerability to risks. Signed on April 17, 2019, Republic Act 11310, an Act institutionalizing the Pantawid Pamilyang Pilipino Program (4Ps), guarantees sustainability and continuity of the program, along with its budget, and follows good global practice for social protection interventions to provide a solid legal foundation and reduce the risk of policy reversals. The law makes Pantawid an added function of the DSWD and provides regular funding from its annual appropriation. The provisions of the law keep the focus of the program on health, nutrition and education, and ensures that poor families get support to develop the human capital of their children. The provisions of the law are clear and are based on the lessons learned by implementing the program for the last 11 years. The Act covers all aspects of the program thoroughly, from program objectives, targeting methods, eligibility criteria, to convergence and coordination with other programs for graduation. The clarity and details of the law reduce the risk of rules being interpreted and applied in an arbitrary manner, increases predictability in its application, and the likelihood the program will be able to fulfill its objectives and achieve desired outcomes. The provisions of the 4Ps Act follow the current design and implementation of the program and provide guidance on key aspects of the program. • Advisory councils and institutional partnerships: A National Advisory Council (NAC) will oversee the implementation of the program together with similar councils recreated at the regional level. The law identifies other government agencies involved in the supervision of the program and provides coordination and supervisory functions. • Targeting and beneficiary selection: The 4Ps Act requires DSWD to continue selecting beneficiaries through a national standardize targeting system like Listahanan and to conduct eligibility reassessment or recertification of beneficiary targeting every three years. The law expands the eligibility criteria to include “near-poor” households with pregnant women and children 0-18 years old at the time of registration. • Program conditions: The program’s conditions that help beneficiaries improve their standard of living (e.g. check-ups at health centers, school attendance) remain. The law enables suspension of conditions during times of calamities, war and armed conflicts. • Benefits: The law increases health benefits by 50 percent and the senior high school benefit by 40 percent (health benefits increased from 500 to 750 pesos and senior high school benefit from 500 to 700 pesos) which will help the program reduce malnutrition, dropouts and increase completion of secondary education. The benefit levels are protected against the risk of inflation as the Act mandates the NAC to review the real benefit of the value over time. The Act specifies clearly the benefit levels, the frequency and duration of benefit payment and the exit/suspension rules. • Program duration and exit rules: The Act limits the duration of beneficiaries in the program to seven years, unless the NAC recommends a longer period under exceptional circumstances. The Act also provides clear guidelines for suspension and/or removal from the program for non-compliance of conditions. • Compliance monitoring and evaluation: The Act requires the program to continue monitoring pregnant and children and describes consequences of non-compliance. The law mandates program assessments every three years conducted by the Philippine Institute for Development Studies (PIDS), the government’s think tank, to evaluate the effectiveness of the program, the veracity of the list of household beneficiaries, and program implementation. • Convergence with other programs: The Act reinforces convergence and coordination with other programs within DSWD and across other government agencies. It specially focuses on the complementarities and linkages with productive economic activity and employment. The Sustainable Livelihood Program (SLP) and similar programs for employment facilitation and small business development will give priority to 4Ps beneficiaries. All of the provisions are in line with good practice in social protection programs around the world, except for the limitation of a maximum of seven years for beneficiaries. The exact procedures will be defined in the law’s Implementing Rules and Regulation, currently under elaboration. www.worldbank.org 3 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance Previous benefit incidence analyses (BIAs) confirm that the Pantawid Pamilya contributes to reducing national poverty, though its targeting performance has declined over time. BIA has been a standard approach in evaluating the targeting performance and progressivity of a CCT program by looking at how beneficiaries and benefits (cash grants) are distributed between poor and non-poor households or across income groups. BIA also assesses the poverty effect of the program by comparing standard poverty and inequality indicators with and without a program, assuming everything else remains unchanged in the living conditions of both beneficiaries and non-beneficiaries4. Fernandez and Velarde (2012), Acosta and Velarde (2015) and World Bank (2018) conducted the BIAs on Pantawid using three Pamilya Conditional te of the Philippine rounds Cash (2009, Transfer’s 2012, and 2015) Implementation of the Family Income and Expenditure Survey (FIES) and one round (2013) of the Performance1 Annual Poverty and Indicator Survey (APIS). World Bank (2018) provided a useful compilation of these estimates across all survey rounds, allowing a snapshot of the program’s progress over the years. While the program has continued to be progressive, its targeting performance—measured by the share of beneficiaries from the bottom quintile and the share of benefits they receive— has declined because it is still using an old targeting database. At the national level, the cash grants were able to reduce poverty and income inequality by around 1.2-1.5 percentage points and 0.5-0.6 percentage point, respectively, from 2012 to 2015. This note aims to update the BIA of Pantawid Pamilya using APIS 2017 and compare results with the second BIA that used APIS 2013 and the third BIA that used FIES 2015 to analyze trends. The release of the 2017 round of the APIS by the Philippine Statistics Authority (PSA) presents a timely opportunity to update Pantawid Pamilya’s targeting performance and progressivity as well as its impact on poverty to provide relevant recommendations to decision makers especially at the onset of the elaboration of the Implementing Rules and Regulations (IRR) of the 4Ps Act. Like FIES, the value in the APIS nationwide survey is that it asks whether a respondent is a Pantawid Pamilya beneficiary or not, as well as how much was received from the program5. The APIS also contained non-income indicators to describe the socioeconomic profile of Filipino families, making it useful for analyzing the poverty situation and various programs designed for the poor. Complementing this BIA analysis, the note presents an updated discussion of the program’s impact on specific program outcomes using the latest Third Wave Impact Evaluation (IE 3)6. Unlike BIA, a rigorous impact evaluation can account for behavioral changes induced by a program and thus could assess impact on specific development outcomes such as health, education, household welfare and other socio-behavioral indicators. Conducted by the Philippine Institute for Development Studies (PIDS), the evaluation examined program impacts following key design modifications that the program has undergone after the IE done in 2014. II. Overview of the Program Significant investments in Pantawid Pamilya have created an efficient means of delivering assistance to poor and vulnerable households. From a meager budget of Php 4 million (USD 78,0007) in 2007, the program grew to Php 89.4 billion (USD 1.7 billion) in 2018 to cover 4.2 million regular CCT and MCCT households, representing 0.5% of the country’s Gross Domestic Product (GDP) and 2.6% of total government spending in 2018 (Figure 2). At this size, the program is mostly financed out of the Government’s own funds, with less than a quarter (22 percent) coming from external financing of development partners (World Bank and the Asian Development Bank). By major item, the majority or about 80-90% of the budget is appropriated for cash grants, while the remainder of the budget goes to administrative expenses, comprising staff and personnel costs, monitoring and evaluation, bank fees, cost of equipment, trainings and printing of manuals and booklets. The CCT budget increased in 2017 to make provision for a cash-based rice allowance per household, which started being distributed through Pantawid Pamilya’s payment mechanisms and also became a component of the 4Ps budget. Based on 2018 budget data, the administrative costs of delivering the cash grants (inclusive of rice allowance) and monitoring of conditionalites amounted to 8.4% in 2018, a share that has been declining over time. Pantawid Pamilya remains an efficient means of delivering assistance to households8. 4 The assumption implies that the BIA does not account for behavioral changes (i.e. potential reduction in labor income for families in a program). For more information on BIA, see Acosta and Velarde (2015). 5 FIES is not directly comparable with APIS due to differences in survey questionnaire and length. However, the greater aim is seeing how the results change over time rather than performing a rigorous comparison. Thus, results of the latest BIA (2018) using FIES 2015 will be included in the discussion of trends, as they are also the latest estimates. For APIS, the questions on Pantawid Pamilya started appearing in the 2011 round. Of the four rounds of APIS prior 2017, only the 2013 round was used for a BIA. 6 Forthcoming. The findings discussed in this note are preliminary. They were presented during the Pantawid Program Review and Evaluation Workshop on March 13, 2019. 7 Php 51= USD 1 for simplicity. 8 The cost-benefit ratio for Pantawid Pamilya was previously compared with another large scale program of the government in 2009, the rice subsidy program. The latter was found to cost the government PHP 68.40 to deliver a PHP 100 direct assistance to its beneficiaries (Fernandez and Velarde, 2012). 4 WORLD BANK SOCIAL PROTECTION POLICY NOTE budget 2018 2018 data, data, budget the administra the administra tive costs tive costs of delivering of delivering the cash grants the cash (inclusive grants of rice (inclusive of rice allowance) allowance) and monitoring and monitoring of condi of condi tionalites tionalites amounted amounted to 8.4% to 8.4% in 2018, in 2018, a share a share that that has has been been declining declining over ti over me. ti me. Pantawid Pantawid Pamilya remains Pamilya an effi remains an cient means efficient means of delivering assistance of delivering to to assistance 8 8 households households . . FigureFigure 2. 2. 2. Budget Figure Budget Budgetonon Pantawid on Pantawid Pamilya: Pantawid Pamilya: 2007-2017 2007-2017 Pamilya: 2007-2017 appropriated Budget Budget appropriated Budget Budget appropriated appropriated byitem by major major item 0.60% 0.60% 100,000 100,000 Php 89.4 Php 89.4 12% 12% 10% 10% 9% 9% 9% 12% 9% 11%12% 8% 11% 8% 8% 8% 90,000 90,000 18% 18% 19% 19% 0.50% 0.50% 80,000 80,000 0.40% 0.40% 70,000 70,000 60,000 60,000 0.30% 0.30% 50,000 50,000 88% 90% 90% 91% 91% 91% 88%91% 89%88% 93%89% 92%93% 92% 40,000 40,000 82% 88% 82% 81% 81% 0.20% 0.20% 30,000 30,000 0.10% 0.10% 20,000 20,000 0.00% 0.00% Php 4 Php 4 10,000 10,000 0.00% 0.00% 0 0 2009 2011 2009 2010 2011 2013 2010 2012 2012 2014 2013 2015 2014 2016 2016 2018 2015 2017 2017 2018 07 08 07 9 08 0 09 1 10 2 113 124 135 146 15 7 168 17 18 00 01 01 01 01 01 201 20 20 20 20 20 20 20 20 20 20 20 20 20 20 2 2 2 2 2 2 Cash grants Cash grants Administra Administrative cost tive cost Actual Spending Actual Program (in Php millions) Program Spending (in Php millions) % to GDP% to GDP Source: Data Source: Source: on budget Data on appropriations Data budget on budget and appropria cost from appropria tions DBM ti and cost ons National and cost from Expenditure DBM fromNa tional DBM Program Na and tional General Program DBM Program Expenditure Expenditure Appropriations and DBM and Act; GDP General DBM figures Appropria Appropria General from PSA tions tions National Accounts. Act; Note: GDP figures Act; Starting GDP 2017, the fi gures from budgetPSA for from Nati the onal PSA rice Na tional is Accounts. allowance Accounts. included under the component of cash grants. Note: Note: Star tingStar 2017, ting 2017, the the budget budget for the for ricethe allowance is included rice allowance under the is included under component the component grants. of cash of cash grants. III. Benefit III. Bene III. Bene Incidence t Incidence Analysis Analysis t Incidence Analysis Data Description Data Description Data Description APIS 2017 survey captured a nationally representative sample of CCT beneficiaries. The APIS 2017 captured 1,839 Pantawid Pamilya household beneficiaries in the survey or about 18 percent of the national survey sample, which represented 3.9 million of the 4.4 APIS million APIS 2017 2017 household survey captured survey beneficiaries incaptured 2017. nationally a Apart nationally a from representative representative the regular sample CCT beneficiaries, sample ofisCCT it ofbene CCTthat presumed ciaries bene .APIS The APIS . The captured this ciaries sample MCCT 2017 2017 beneficiaries captured as captured well, as 1,8391,839 it is also associated Pantawid with Pantawid the Pamilya household overall Pamilya 4Ps 9 .household The benefi number of ciaries bene ficiaries beneficiary in the in survey the survey households with or about 0-18 or about years 18 18 — old children the specific target of the CCT program— are slightly lower covering 1,782 households, which represented 3.8 million in 201710. In terms percent of the percent of na tional the nationalsurvey sample, survey sample, which represented which represented 3.9 million 3.9 million of the of grant value, the APIS asks for the amount received from the Pantawid Pamilya for two durations: last month and last six months. Like of 4.4 4.4 million the million household previous household BIAs, thebene amount fibene ciaries in 2017. ficiaries for the longer in 2017. Apart six-month Apart from from duration, the regular estimatedtheat CCT regular Php bene 5,500 CCT (USDfibene ciaries, 108)fi ciaries, in it isispresumed 2017, it is presumed chosen in this note that that as it has 9 morethis this sample sample observations. captured Lastly,capturedMCCT average benehousehold MCCT semestral bene fi ciaries as ficiaries as well, income well, as it is is estimated is associated it Php asalso at also associated 135,933 (USDwith with the 2,665) the by in overall 2017; overall CCT . 4Ps9 4Psstatus, . with an average of Php 150,000 (USD 2,941) non-beneficiaries have nearly twice the 6-month income of CCT beneficiaries which reported only8 an average 8 of Php 83,000 (USD 1,628). The cost-benefit The cost-benefit ratio for Pantawid ratio Pamilya for Pantawid Pamilya was previously was previously compared with another compared large scale with another scale program large program of the of the government Coverage in government 2009,in 2009, the rice subsidy the rice program. subsidy The program. The latter latter was found was found to cost the to government cost the PHP government PHP 68.40 68.40 to to deliver deliver a a PHP 100 PHP 100 assistance direct direct assistance to its beneficiaries to its beneficiaries (Fernandez and Velarde, (Fernandez 2012).2012). and Velarde, 9 In 2017, 9 Pantawid Pamilya covered The The questionnaire APIS APIS does less questionnaire not does than nothalf make a make of the targeted distinction between a distinction population, regularregular between representing CCT and CCT MCCT anda significant MCCT decline beneficiaries. beneficiaries.in coverage May May need need since 2015. Programto explain brieflybriefly coverage refers what is to explain to MCCT what the portion is MCCT for theforof population uninitiated in the uninitiated each The questiongroup that The question receives refers torefers the to the the 4Ps benefit. 4Ps program program In 2017, in general the CCT in general Program : “During: “During targeted the the poor households period period with January-June January-June pregnant 2017, women 2017, has any and children any member hasmember 0-18. Among of your all these of family your family poor received households, received the coverage rate was 49 percent benefits/grants/assistance/payment benefits/grants/assistance/payment from from(Figure 3). This is Pantawidlower than Pamilyang Pantawid the 58 Pamilyangpercent Pilipino coverage Pilipino Program rate (4Ps)?” Program in 2013, (4Ps)?” wherein the target are poor households with pregnant women and children 0-14, and the 59 percent coverage in 2015, wherein the target are also poor households with pregnant women and children 0-18. 5 5 9 The APIS questionnaire does not make a distinction between regular CCT and MCCT beneficiaries. The question refers to the 4Ps program in general: “During the period January-June 2017, has any member of your family received benefits/grants/assistance/payment from Pantawid Pamilyang Pilipino Program (4Ps)?” 10 Not all households claiming to belong to 4Ps have 0-18 year old children in the APIS dataset. Children in these households may study and live outside the household and thus not being captured by the APIS survey. www.worldbank.org 5 2013, wherein the target are poor households with pregnant women and children 0-14, and the 59 percent coverage in 2015, wherein the target are also poor households with pregnant and Pamilya Pantawid women 2017 children Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance 0-18. Figure 3. Program coverage (% of the population within a referenced group) Figure 3. Program coverage (% of the population within a referenced group) 62 61 58 59 te of the Philippine Conditional Cash Transfer’s Implementation 49 Performance1 50 38 41 33 19 22 17 6 8 6 2 1 1 Poor Q1 Q2 Q3 Q4 Q5 (Poorest) (Richest) 2013 2015 2017 Source: Acosta and Velarde (2015); World Bank (2018); WBG staff calculations based on APIS 2017 using World Bank Social Protection ADePT Software Source: Acosta and Velarde (2015); World Bank (2018); WBG staff calculations based on APIS 2017 using World Bank Social Protection ADePT Software The declining program coverage reflects a halt in program enrollment and lack of a replacement policy. The Pantawid Pamilya stopped enrollment of new households in 2016 even though the 2015 Listahanan targeting database identified a large number of declining Thehouseholds eligible who program coverage were in need a halt re ects (see of social protection in Inprogram Box 2). enrollment addition, until and 2017, Pantawid lack did Pamilya of not a have any replacement provision policy. for replacement The Pantawid of households that exited from thestopped Pamilya enrollment program either of new through natural households attrition in 2016 or other forms of exit. Natural attrition occurs when the last monitored child in the household graduates from high school. Latest data in 2019 show that the natural attrition rate is 2.2 percent (93,000 households), slightly higher than the 2.0 percent (88,000 households) recorded in 2015. The households 10 that naturally exit are growing over time, and, without a replacement policy, are contributing to the declining program Not all households claiming to belong to 4Ps have 0-18 year old children in the APIS dataset. Children in these coverage. More poor households could have been served to reach the multi-year household target of 4.4 million households covered households under may the regular . and live outside the household and thus not being captured by the APIS survey. study CCT 11 6 11 The targets are from the National Economic and Development Authority’s Philippine Development Plan 2017-2022. In response to the challenge of a lack of replacement policy, National Advisory Council (NAC) Resolution 43 or the “Pantawid Pamilya Household Replacement Policy to Reach the Annual Household Coverage” was adopted on 13 December 2018. Approval of replacement households is scheduled on December 2018. 6 WORLD BANK SOCIAL PROTECTION POLICY NOTE Box 2 - Listahanan: The Philippines Targeting System Listahanan is the gateway for individuals and households to a menu of social protection programs in the Philippines. The Listahanan allows households to register and be considered for potential inclusion in social programs based on a transparent assessment of those households’ needs and conditions. It establishes a social registry (i.e. a socio-economic database of potential applicants for social programs) and uses a proxy-means test (PMT)a as the methodology for estimating per capita income of households for ranking and identifying poor households. Registration of households into the social registry follows a “census sweep” approach where all or most households in specific areas in the entire country are registered en masse. The DSWD, the lead agency designing, implementing and managing Listahanan, completed the first Listahanan registration in 2011 and a second one in 2015. Listahanan 2011 and 2015 registered and collected data for 10.9 and 15.1 million households respectively (around 50 and 70 million people). Listahanan 2011 became one of the Philippines’ largest household databases when it was completed in 2011. It covered 58 percent of the 18.5 million households in the Philippines in 2010. Out of the 10.9 million households enumerated, 5.2 million were classified by the first PMT model as poor. Listahanan 2015 covers three quarters of the 20.3 million Filipino households in 2015. Out of the 15.1 million households enumerated, 5.1 million were classified by the second PMT model as poor (Velarde, 2018). Listahanan functions as the single common and authoritative source to facilitate determination of eligibility of potential beneficiaries for multiple programs. Initially designed to serve the Pantawid Pamilya program, the Listahanan was quickly adopted by the Philippines government as the primary mechanism to target poor households for social programs. Listahanan provides registration support and eligibility determination for key social programs from DSWD (such as Pantawid Pamilya, Sustainable Livelihood Program, Social Pension for Indigent Senior Citizens and KALAHI-CIDSS), and other national government agencies such as the subsidized health insurance program (PhilHealth), Department of Labor and Employment (DOLE), Commission on Higher Education (CHED), among othersb. In 2018, the Listahanan was used to identify the 10 million households eligible to receive the unconditional cash transfer (UCT) program of the government that helps mitigate the increase of prices brought about by the implementation of the Tax Reform Law (TRAIN). The use of Listahanan for UCT is a clear reflection of its credibility and performance. While Listahanan has become the primary social registry in the country and achieved significant gains in targeting efficacyc, it still has areas for improvement in the frequency and scope of updates and flow of information from user programs to Listahanan. The current static “census sweep” approach used by Listahanan conducted every four years presents risks of rising errors of exclusion and inclusion with the passage of time, as the information becomes out of date. The current Management Information System (MIS) supporting the registry does not allow interoperability with user programs (even within DSWD programs) and limits flow of information and updates coming from beneficiary registries. In addition, high vulnerability to poverty in Philippines requires a social registry that supports dynamic inclusion, meaning access to registration is open and continuous, usually via a combination of on-demand applications plus active outreach to vulnerable populations. The Government may consider developing an integrated information system with a social registry which supports dynamic inclusion and is interoperable with beneficiary registries. An integrated information system, with a common data exchange platform to facilitate links to user programs, the PhilSys (unique national ID), and other information systems, will enhance the provision of social programs and services. A dynamic and interoperable Listahanan can prevent undesirable duplication of benefits, track beneficiaries to monitor “who receives what” across programs to improve coordination of social policy and programming, and in the longer term, allow for cross-checking different databases. a. Proxy Means Test (PMT) is a statistical model that aggregates multiple proxy indicators into a single score that represents the household’s welfare level. b. The database is used by 59 national programs and has been shared with 1,095 local government units, and 56 non-government organizations. c. See Fernandez and Velarde (2012), Acosta and Velarde (2015) and World Bank (2018). www.worldbank.org 7 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance Beneficiary incidence and Targeting accuracy Mainly due to the outdated targeting database, the program has seen falling levels of targeting accuracy since 2013, in terms of both exclusion and inclusion errors, and when compared to CCT programs of other countries. Beneficiaries’ incidence is a measure of targeting accuracy that refers to the proportion of beneficiaries in each group. In terms of income quintiles, 46 percent of all Pantawid Pamilya beneficiaries belonged to the poorest 20 percent of the population (Quintile 1 or Q1) and the next 31 percent belonged to the second lowest income group (Q2) (Figure 4). In other words, 77 percent were from the bottom 40 percent of the population, which in te of the Philippine Conditional the Philippines Cash Transfer’s comprises Implementation poor households Performance1 as well as households vulnerable to poverty, e.g. likely to experience episodes of poverty during the year. These results are the same as the third BIA (based on FIES 2015) but lower relative to the second BIA (based on APIS 2013), which found that 82 percent of all beneficiaries were poor, and 53 percent and 29 percent of beneficiaries were from Q1 and Q1, respectively. The targeting accuracy was expected to decline over time, given that the data collection for the first Listahanan took place in 2009 and 2010. Additionally, these targeting registries require continuous update in its survey methodology and formula for predicting poverty status (“proxy-means test”). The household poverty status would then be most accurate in 2010, but between 2013 and 2017 the former would be less outdated. The issue of static massive data bases and dynamic changes in household income is faced by other countries with similar programs as well. That is why, in the 4Ps Act, the Philippines government has stipulated that the beneficiary targeting system undergo revalidation every three (3) years to keep the data accurate. In the world-wide context, the targeting of many other CCT programs perform better than the Philippines currently. Whereas in 2013 it performed better than most countries, in 2017 the Philippines’ targeting performance occupies the lower range of percentages, just above Colombia and Guatemala. Figure 4. Distribution of Beneficiaries Figureti Figure 4. Distribu 4.on Distribu of Beneficia ries tion of Beneficiaries Distribu Distribution of tion Beneficia ries of by Bene ficia incomeries quin by lecome quintile tiin ficia Share of BeneSh are ries of belonging Beneficiaries belonging to bo to bottom 20%: ttom 20%: Compa International Intern rison ati onal Comparison 55% 53% 55% 13% 13% 53% 52%53% 53% 52% 16% 16% 16% 16% 47% 47% 46%47% 46%47% 46% 46% 41% 41% 29% 29% 31% 31% 31% 31% 29% 29% 53% 53% 46% 46% 46% 46% ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) ) 4) ) ) 15 13 15 13 14 14 13 16 13 15 14 17 16 14 15 14 17 01 14 14 ( 20 2 20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 (20 20 20 2013 2013 2015 2015 2017 2017 zil si a( s ( l i a a o s r a s o s r a s a s a ( ala ( a ne az n si ic ne do n ne ic ne do bi ne al ne bi Br ne pi Br enti one ex ppicua entippi exppi cua lom ppi tem ppi lo m te m n do i lip r g nd M ili E rg ili M ili E o ili a ili C u C o u a Q1 Q2 Q3 Q1 Q5 Q3 Q4 Q2 Q4 Q5 I Ph A I Ph A Ph Ph Ph G Ph G Sourcand Source: Acosta e: Ac Sou osta Velardeand Vel rce: Aa (2015); os crdeta(2015); Worldand Vel Bank World arde (2015); (2018); B ank (2018); WBG World staff WBG Bank s(2018); taff calculations calc based ul WBG onati staff ons APIS ased bcalc 2017 ulon APIS ations using b2017 Worldased Bankon using APIS World Social ankADePT 2017Busing Protection World So Bank So cial Software; cial Bank World Protecti ASPIRE database Prote on other for ADePT So on ft cti ware countries.ADePT SoftB ; World waare ; Worldd nk ASPIRE Bata ankbaASPIRE se for other databcaountries. se for other countries. Bene t incidence Bene t incidence and Progressivity and Progressivity Benefit incidence and Progressivity While While it it has has declined declined While over ittime, over has Pantawid time, declined over Pantawid Pamilya time, remains Pamilya Pantawid Pamilya remains progressive, progressive, with theremains greatest progressive, with ofthe share with greatest benefits the to going greatest the poorest share households. of distribution share ts bene The bene ofgoing of ts the to benefits goingpoorest mirrorsto howthe households. poorest well the program The households. distribu reached The poortiondistribu of bene households tionints fi ofmirrors the bene fits bottom how 20mirrors how As or 40 percent. and 2015, in 2013well the the prowellbiggest gramthe reshare pro acgram hed of benefits poor reached went to the households poor poorest in the households households bott inom the20 (43 bo or ttpercent om 40 per 20to or c Q1) 40 ent. and As perthe in share cent. 2013 Asdrops and to nil in in 2013 andhigher groupsthe income 2015, (Figure 5). biggest 2015, The the Lorenz sha re of bene biggest curve of share the fitsof Pantawid bene went fits to thePamilya poorest went shows to the that the households program poorest households is progressive, (43 percent (43 ercea topQ1) that nd nt is, to the the poorest Q1) and the receive a higher share of program benefits than their actual share in the national income distribution. However, it must be noted that the level sh ar e drops sh to ar e drops nil in higher to nil in in come higher in groupsc ome groups (Figure 5). The (Figure Lorenz 5). c Theurve Lorenz of c the urve of Pantawid the of progressivity has declined over time. The share of benefits that the poorest quintile received fell from 54 percent in 2013 and 45 Pamilya Pantawid Pamilya shows th a t the shows pro th gra a t m the is pro gra m is progressive, th progressive, a t is, theth a poorest t is, there ceive poorest a re ceive higher a sh ar higher e percent in 2015 to 43 percent in 2017, at the same time as benefits to those in the third quintile increased notably (Figure 5).of sh pro gra ar e mof pro gram benefits th an their bene ac fits th atu na l shar their e tu ac l shar inathe nati e in onthe al incati n omeonaldistribu income n. However, tiodistribu it must beit tion. However, must be noted noted that of that the level the level of progressivity progressivity has declined declitime hasover ned .over The time shar.eThe sharfi of bene ofth ets at the bene that the poorest fitspoorest ceived quintile requin recfrom tilefell eived54 perc fell from ent inp 54 ercent 2013 in 45 and 2013perce and 45 nt 2015 inpe rcent 43 toin 2015 percto ent43inp ercent 2017, in 2017, at at theesam the same tim e tim as bene fiets asto those bene fitsinto the those third the til inquin e inc third reased quin tile in not ably cre ased notably (Figure 5).(Figure 5). Figureti Figure 5. Distribu 5. Distribu on tion of Benefits of Benefits Distribution Distribution of income benefits byof benefits group by income group of the Pantawid Progressivity Progressivity the Pantawid Pamilya of Pamilya 8 100.0 100.0PROTECTION 98.6 SOCIAL WORLD BANK 98.6 POLICY NOTE 13% 13% 100 100 91.9 91.9 17% 17% 18% 18% ived ived 74.3 74.3 shar share drops to ein nil drops higher nil to in in higher come groupsinc(Figure ome groups 5). The Lorenz (Figure 5).curve The Lorenz of the c urve of the Pantawid Pantawid Pamilya Pamilya shows shows that the tham progra is progressive, t the that is, the poorest program is progressive, that is, the receive poorest receive a higher shar of progra aehigher sharme of program benefits than their bene acth fits tuaan l sh are ac their the in tu al sh nati aron e in income al the nation al income distribu tion. distribu However, it must tion. be noted However, it must be noted thof that the level progressivity at the level of progressivity has declinedhover time as decli ned over . The shtime are of bene . The shfi ar tseth of the fi atbene ts that the poorest poorest quintile quintile received from fell re 54 p ceived fell erc ent in from 54 perca 2013 nd in ent 452013 perce a ndin nt 45 perce 2015 tont 43 ercent inp2015 toin 2017, 43 at in 2017, at percent the same tim e as the bene sam e tim fitse toas benein those fitsthe to third thosequin in the incre tilethird ased quin til not e inac rea(Figure bly sed not5).ably (Figure 5). Figure 5. Distribution of Benefits Figure Figure 5. Distribu tion of 5. Benefits Distribution of Benefits Distribu Distribution of benefits by incof tion omebene group fits by income group Progressivity Progressivity of the Pantawid P amily of Pantawid Pamilya the a 98.6 100.0 98.6 100.0 13% 13% 100 100 91.9 91.9 17% 17% 18% 18% Cumulative % of benefits received Cumulative % of benefits received 74.3 74.3 29% 29% 80 80 31% 31% 32% 32% 60 60 42.8 42.8 40 40 54% 54% 45% 45% 43% 43% Source: Source: Acosta Acosta and and Velarde Velarde (2015); (2015); World Bank World Bank (2018); WBG (2018); 20WBG staff calcula staff calcula 20 tions based on APIS 2013 and APIS 2017 using World tions based on APIS 2013 and APIS 2017 using World Bank Social Protec tion ADePT So ft ware; World Bank ASPIRE database for Bank Social Protection ADePT Software; World Bank ASPIRE database for other countries. other countries. 2013 2015 2013 2017 2015 2017 0 0 0 20 40 0 60 20 40 80 100 60 80 100 Q1 Q2 Q3 Q4Q1 Q5Q2 Q3 Q4 Q5 Generosity Generosity Cumul Cumulative % of population ative % of population Source: Acosta and Velarde (2015); World Bank (2018); WBG staff calculations based on APIS 2013 and APIS 2017 using World Bank Social Protection ADePT Software; World Bank ASPIRE The The for database amount amount ofother grant of countries. fromgrant from Pantawid Pantawid Pamilya Pamilya continues continues to be low to. be low. In In 2017, 2017, bene beneficiary ficiary households received received households an average average 6-month an 6-month grant9 of Phpgrant of 9 Php 1,092 (USD1,092 21) (USD 21) perThis per person. person. means This means that for an Generosity that for an bene average average ficiary bene ficiary household, household, with six members, with six members, a Pantawid a Pantawid Pamilya household Pamilya household received Php5,458 (USD 107) in a semester or about received Php5,458 (USD 107) in a semester or about Php 10,916 (USD 214) for the full Php 10,916 (USD 214) the of foryear full year of The amount of grant 2017. fromThePantawid Pamilya rice allowance continues may to be not included low . In be fully included2017, beneficiary in the households total bene received fit as Pantawid an average Pamilya6-month grant of 2017. The Php 1,092 rice21) (USD allowance per person. mayThisnot means bethat fullyfor an average inbeneficiary the total bene household,fit as Pantawid with Pamilya six members, a Pantawid Pamilya householdstarted received started distributing distributing Php5,458 (USD rice 107) rice allowance allowance progressively progressively in a semester or about Php 10,916 This in 2017. in 2017. This corresponds (USD corresponds 214) for the full to only year only to percent 6.6 of 2017. 6.6 percent The rice allowance may of bene not be fully fiof ciary included inbene the ficiary households’ total households’ benefit pre-transfer as Pantawid pre-transfer income Pamilya income in started in 2017 2017 (10.2 distributing (10.2 for percent rice allowance percent for households households progressively in the in 2017. This in the corresponds to only 6.6 bottpercentbo om 20), tt om of which beneficiary 20), which is households’ is lower lower than the than the pre-transfer 7.0 percent 7.0 incomepercent recorded recorded in 2017in (10.2 2013 in percent 2013 (11.4for (11.4 households percent percent for in the bottom 20), which is for households households lower than the 7.0in percent the borecorded tt om in 2013 20) and (11.4 percent slightly higher households for than the in the 5.4 bottom percent 20) and slightly recorded in higher 2015 (8.8than the 5.4for percent percent in the bottom 20) and slightly higher than the 5.4 percent recorded in 2015 (8.8 percent for recorded in 2015 (8.8 percent households for households in the 20), in bottom the bottom 20),on 20), mainly mainly on of on account account of the of tional the addi additional tional rice subsidy rice (Figure (Figure subsidy 6). 6). Comparing (Figure 6). households internationally, in the of the generosity bo ttom Pantawid mainly Pamilya continuesaccount to lag behindthe addi other programs rice from subsidy other countries. Comparing Comparing internathe internationally, tionally, the generosity generosity of Pantawid of Pantawid Pamilya con Pamilya tinues contotilag nues to lagother behind behind other programs from programs from other countries. other countries. Figure 6. Generosity of Pantawid Pamilya Figure 6. Figure 6. Generosity Generosity of Pantawid Pamilya Pamilya of Pantawid Generosity Generosity by by income income quin tile quintile Generosity Generosity in in bottom 20%: bottom 20%: InternationalInternational Comparison Comparison 11% 11% 10% 27% 10% 27% 23% 22% 9% 23% 22% 9% 20% 20% 7% 7% 7% 7% 6% 7% 6% 7% 6% 12% 11% 11% 11% 12% 11% 11% 11% 10% 5% 6% 10% 9% 5% 5% 9% 5% 4% 4% 4% 4% 4% 4% 2%2% 2%2% 2%2% 2%2% 1% ) 6) )14) ) 3) ) 15) ) 14) 3) )4 ) 14 ) 17 ) 15 ) 1% 16 3)01 514 4)1 3 4 4)1 71 ) 15 (20 0 0r1(2 20(1 0 20 0(120 01 ( 20 01 ( 20 0(1 20 0 ( 10 2 ( 2 0 0 ( 20 (2 2 (2 2 (a 2 l 2 (zi a ( ru c (es 2 a2 ic( 2 s s or co a( do ae (o zixlic er utin es ra i e bp liin esbi( bls ne ne Philippines Q1 Q2 Q3 Q4 Q5 uad exi ticnua BM r P en pinB mb P plu loinm ne ppi piu pi Philippines Q1 Q2 Q3 Q4 Q5 c e nE r g p l o i ip o p p R ep p i l i i l ip M g A hil i Reh i ilC nil i E Ar Co nP Ph Ph P Ph ah icP 2013 2015 2017 ica in 2013 2015 2017 in m m Do Do Source: Acosta Source: and Velarde Source: (2015); World Bank (2018); WBG staff calculations based on APIS 2013 and APIS 2017 using World Bank Social Protection ADePT Software; World Bank ASPIRE database for other Bank Social countries. Protec tion ADePT Software; Bank Social Protection ADePT Software; 2007 2007 to maintain to maintain a reasonable a reasonable p p www.worldbank.org 9 12 29 29 percent percent 12 (Figure 7) (Figure 7) Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance The low generosity may be attributed to the benefit schedule that has remained fixed since 2007 and resulted in a steady decline in the real value of transfers. Over the years, the Government has prioritized covering all poor households with children over updating benefits to maintain a reasonable program budget. However, because the nominal value of the grant has not been adjusted since it was piloted in 2007, at an annual inflation rate of 3.5 percent from 2007 to 2017, the maximum entitlement of Php 15,000 (USD 294) per household in 2007 was is valued only introduced inat about the 2014, Php 10,600 (USD 208) generosity in 2017, equivalent of Pantawid Pamilyato a cumulative con tinues toreduction lag behindof 29 percent12 (Figure other 7). Even after the additional countries. Php 500a Assuming (USD 10) for similar high school composi tion children that was of 3 children, introduced the 4Ps Act in 2014, the provides a generosity of Pantawid Pamilya higher maximum continues to lag behind te of the Philippine Conditional Cash other countries. Transfer’s Assuming a similar composition of 3 children, the 4Ps Act provides a higher maximum nominal nominal value forImplementation the grant amoun Performance1 ting to Php 30,000 (USD 588), valued at about Php 21,300 value for the grant amounting to Php 30,000 (USD 588), valued at about Php 21,300 (USD 418) in 2007, equivalent to a cumulative increase of 42(USD 418)The percent. in generosity 2007, equivalent to a cumula is thus expected tive increase to rise when ofis42 the 4Ps Act percent. The implemented. generosity (Figure 7). is thus expected to rise when the 4Ps Act is implemented. (Figure 7). Figure 7. Real Value of Maximum Entitlements Figure 7. Real Value of Maximum Entitlements 9.0 15,000 16,000 8.0 14,000 7.0 12,000 10645 6.0 10,000 5.0 8,000 4.0 6,000 3.0 4,000 2.0 1.0 2,000 0.0 - 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Real value of Php 15,000 Year-on-year inflation rate Source: WBG staff WBG sbased calculations Source: on PSA taff calcula price tions data.on PSA price data. based with the Note: In 2014, Note: Inprogram expansion 2014, with to 15-18 the program year old children, expansion to 15-18the maximum year entitlement old children, per household the maximum is Php 21,000 entitlement a year. Tois compute per household for the real value Php 21,000 using 2007 terms, Php 15,000 is multiplied to the quotient of the CPI in 2007 over the CPI in a given year. Same procedure applies for Php 21,000 and Php 30,000. a year. To compute for the real value using 2007 terms, Php 15,000 is multiplied to the quotient of the CPI in 2007 over the CPI in a given year. Same procedure applies for Php 21,000 and Php 30,000. Apart from the non-adjustment Apart of the benefit schedule, from the non-adjustment of theother bene factors could explain t schedule, otherthe low generosity factors could .explain An observation the low from national surveys and confirmed with administrative data is that some Pantawid Pamilya households do not maximize the 3-children cap of the programgenerosity . An and only have, observa on average,tio 2n fromchildren eligible national surveys between and 3-14 con years firmed old. Basedwith administra on APIS 2017, thetiv e data average is household size that some of CCT beneficiaries Pantawid is 6 persons, Pamilya translating tohouseholds an average of do not maximize 4 children the 3-children per household. cap of Another factor thedelays is the program in payment and and only thus underreporting ofhave, onin benefits average, 2 eligible surveys, which children could arise frombetween 3-14 unreported years in changes old. Based on household APIS 2017, information suchthe as transfer of school ofhousehold residence oraverage children. According to DSWD, size of CCT ficiaries of benemajority 6 persons, related is transactions translato beneficiary ting information to an average received by program of 4 children implementers are requests to update schools where children’s attendance should be monitored. per household. Another factor is the delays in payment and thus underreporting of benefits in surveys, which could arise from unreported changes in household information such as transfer of residence Impact on Poverty or school of children. According to DSWD, majority of transactions related to and Inequality beneficiary information received by program implementers are requests to update schools declining Despite the where real value children’s and att progressivity endance should ofbe the benefit, Pantawid Pamilya continues to make an important contribution monitored. to reducing poverty. The program is estimated to have reduced total poverty in 2017 by 1.3 percentage points: from an estimated pre-Pantawid rate of 19.8 percent to post-Pantawid rate of 18.5 percent (Figure 8). This is slightly lower than the estimated poverty reduction ofImpact point and on Poverty 1.4 percentage Inequality in 2013 and 1.5 percentage point in 2015. In addition, the program’s poverty focus helped reduce national income inequality by 0.6 percentage point from 46.8 percent pre-Pantawid to 46.2 percent post-Pantawid. Despite the declining real value and progressivity of the bene t, Pantawid Pamilya continues to make an important contribution to reducing poverty. The program is estimated to have reduced total poverty in 2017 by 1.3 percentage points: from an estimated pre-Pantawid rate of 19.8 percent to post-Pantawid rate of 18.5 percent (Figure 8). This is slightly lower than the estimated poverty reduction of 1.4 percentage point in 2013 and 1.5 percentage point in 2015. 12 The maximum entitlement is for illustration purposes only. Note that this is uncommon as the average household has 2 children monitored. 11 10 WORLD BANK SOCIAL PROTECTION POLICY NOTE In the In addition, program’s addi poverty focus tion, the program’s helped poverty focus reduce helped reduce nati national onal income income inequality inequality by 0.6 by 0.6 percentage point frompoint percentage 46.8 percent from 46.8 pre-Pantawid percent pre-Pantawid to 46.2 to 46.2 percent percent post-Pantawid. post-Pantawid. Figure 8. Impact of Pantawid on Poverty and Inequality Measures Figure Figure 8. Impact of8. on Impact of Pantawid Poverty and Pantawid on Poverty Inequality and Inequality Measures Measures onal poverty Impact on natiImpact on nati incidence onal poverty incidence onal income Impact on natiImpact tional income inequality on nainequality 0.6% 0.6% 1.4% 1.4% 0.6% 0.6% 0.6% 0.6% 1.5% 1.5% 1.3% 1.3% 49.2% 49.2% 45.9% 46.8% 45.9% 46.8% 26.4% 26.4% 23.1% 23.1% 19.8% 19.8% 2013 20132015 20152017 2017 2013 20132015 20152017 2017 Post-transfer Post-transfer Pre-transfer Pre-transfer Post-transfer Post-transfer Pre-transfer Pre-transfer Source: Source: Acosta Acosta and and Velarde Source: Velarde Acosta (2015); (2015); and World Velarde Bank World (2015); Bank (2018); World WBG(2018); staff Banks(2018); WBG taff calculations onti calcula basedWBG ons staff APIS calcula based onti 2013 and ons2017 APIS APIS based 2013 on APIS and APIS2013 using World2017 Bankand using World APIS Social 2017 using Protection World ADePT Software Bank Bank Social Protec tion Social ADePT Protec tiware Soft on ADePT Software IV. Current IV. Past and Past and Impact Current Evaluation ndings Impact Evaluation ndings IV. Past and Current Impact Evaluation findings Previous Previous impact impact evaluation evaluation the impact studies onstudies on the of impact the program program of the show Pantawid thatshow that Pantawid Previous been studies impact evaluation Pamilya has Pamilya has been on the successful in impact keeping successful ofinthe program children keeping show that healthy children and Pantawid in school healthy andPamilya (DSWD hasand been in school successful World (DSWD in keeping and World children healthy and in school (DSWD and World Bank, 2014; DSWD, 2014). The first-round impact evaluation was conducted in Bank, 2011 and 2014; a second Bank, DSWD, study 2014) 2014; was DSWD, conducted . The first 2014) in 2013. In The rounds -round . both first impact -round impact evalua tion of evaluation, evalua was conducted findings tion was showed that in 2011 conducted the programand in 2011 a had beenand a successful second in achieving study was second its primary conducted study objective ofwas in children keeping 2013. In conducted in in both 2013. schoolrounds and bothof evalua In keeping themti rounds of on,evalua findings healthy tion, through findings showed increased showed that the of utilization that thecare health services among both children and pregnant women. In contrast, while evaluation methods used in the studies does not allow rigorous program had been successful program in achieving had been successful achieving in its primaryits objec primary tive of keeping objec tive of keeping children comparison across findings, the two impact evaluations have shown mixed results for some outcomes while no program impact has school in school inchildren and keeping been observed and in some keeping them healthy crucial them through indicators healthy such increased as totalthrough u household tili zation u increased consumption tili of health za tion and care of health infant services care immunization. among servicesbothamong both children children and pregnant andwomen. pregnant Inwomen. Inwhile contrast, contrast, evalua while tion evalua methods methods tionused in the used studies does in the studies does Generally, results of the IE 3 indicate that the program shows desirable impacts on most of the target education and health outcomes not allow of children not allow rigorous and pregnant rigorous comparison women. acrossno comparison Nevertheless, findings, across fiimpact the discernible ndings, two impact wasthe evalua two found impact amongtions evalua have ti children shown ons have below mixed six shown years mixed old on nutrition outcome some for results forresults indicators (i.e. wasting, some outcomes while no program outcomes underweight, and while no impact stunting), in program contrast has to impact been has earlier observed been in evaluation some crucial observed findings. The in some crucial disappointing results on the lack of impact on nutrition, particularly on the incidence of stunting, provide strong motivation to refocus health interventions and indicators indicators such such as total as total household household consumption consump and infant immuniza tion and on. infanttiimmuniza tion. compliance monitoring on pregnant mothers and young children during critical growth periods such as the first 1000 days’ window. results of the Generally, Generally, IE 3of results the IE 3 indicate indicate that the program that theshows program shows impacts desirable on desirable most of impacts on most of Educational Outcomes the target education the target education and healthand outcomes and health outcomes of childrenof pregnant children andwomen. Nevertheless, pregnant no women. Nevertheless, no Despitediscernible impact discernible the low generosity, IE3wasimpact found results was confirmamongfound that the amongbelow children Pantawid children Pamilya six below years to continues old six years on nutri improve old and ti on on outcome maintainnutri tion outcome educational outcomes. The program indicatorsincreased (i.e. indicators wastienrollment gross (i.e. ng, wastirates underweight, for and children ng, underweight, stun andyears 12-17 ting), stun in old by in ting), contrast 5topercentage contrast points to earlier evalua tifrom earlier fia baseline ndings. on evalua tion of 80 percent findings. (Table 1). The Further to disappoin this, The theresults tidisappoin ng program on tingis raising results the lackthe on age-appropriate ofthe impact lack on enrollment nutri of impact tion, of the on par nutri 12–15 tion, ticularly parage on group, ticularly the which incidence on the an is of indicator of incidence that is positively correlated with high school completion, by 5.8 percentage points, from a baseline of 66.9 percent. As more high school-age stun children are ti stunon ti ng, provide enrolled ng, strong time, strong motivati provide the high school on to tivati mo refocus completion on to health rate refocus is int likely erventi to health increaseonsaserventi int and well.compliance ons and Among monitoring the compliance 6–11 years age monitoring group, the lack impacton on pregnant of observed pregnant asmothers compared mothers and with young and previouschildren young rounds children during can critiduring cal growth be rationalized critical periods there such growth because as periods is little the room such firsthe tas for the firstto 1000 Program 1000 increase days’ the already days’ window. window. high enrollment rate of 89.4 percent in this age group. Equally important, the cash transfer is shown to influence related behaviors positively. Results show that Pantawid households have higher expenditures on education and clothing and footwear, as well as expenses associated with school attendance. 12 12 www.worldbank.org 11 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance Table 1. Summary of IE 3 Results for Educational Outcomes Indicator Baseline Impact13 School enrollment 3–5 years 35.4 None 6–11 years 89.4 None te of the Philippine Conditional Cash Transfer’s Implementation Performance1 12–17 years 80.4 +4.9 pp 16–17 years 60.8 +10 pp School attendance 3–5 years 65.0 None 6–11 years 88.4 None 12–17 years 85.2 None Age appropriateness enrollment 12–15 years (high school) 66.9 +5.8 pp Dropout 6-14 years 1.7 -1 pp Child labor 5.7 None Source: PIDS (forthcoming), Third Impact Evaluation of the Pantawid Pamilya CCT Program. However, the Program has not been able to replicate earlier positive findings on enrollment among the youngest cohort of beneficiaries (3–5 years). Using APIS 2017, only about 40 percent of poor children ages 3–5 were attending school, in comparison to about 47 percent of the non-poor children in the same cohort (for the baseline impact evaluation, it is 35.4 percent) (Figure 9). The Kindergarten Education Act, enacted in 2011, mandates kindergarten education for children at least 5 years old as compulsory for entrance to Grade 1. This means that the incentives for school enrollment among the 3–5 age group are low, especially for poor households who face budget constraints. Lack of kindergarten facilities in many rural areas, which has not been addressed yet since program started, and limited awareness by parents on the importance of sending children to school are also critical factors. On the other end of the spectrum, there was a positive impact on enrollment for children 16-17 years old. The program has a strong impact on increasing enrollment among students 16–17 years, by 10 percentage points, from a baseline of 60.8 percent. Indeed, nationally, significantly higher gaps in school enrollment between poor and non-poor children are observed for students in senior high school, where almost 84 percent of non-poor children were enrolled in school and only about 72 percent of poor children continued their education for the 15–18 year group. 13 Impact refers to statistically significant impact. 12 WORLD BANK SOCIAL PROTECTION POLICY NOTE Figure 9. School Attendance by Poverty Status and Grade level (%) Figure 9. School Attendance by Poverty Status and Grade level (%) 100 80 60 40 20 0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Non-poor Poor Source: Bank staff calculations based on APIS 2017. Source: Bank staff calculations based on APIS 2017. On the other end of the spectrum, there was a positive impact on enrollment for children 16 - Health Outcomes 17 years old. The program has a strong impact on increasing enrollment among students 16– As in past studies, IE 3 indicates that Pantawid Pamilya promotes the utilization of prenatal care services and skilled birth attendance, as well as17 years, care postnatal by and percentage 10 weight points, monitoring, which from baseline a positive all have of 60.8 effects on the long-term Indeed, percent. welfare ofna tionally, mothers beneficiary signiPantawid and children. ficantly is higher in women gapspoor encouraging schooltoenrollment between use maternal and poor child health and such services non-poor children as antenatal are 8 in 10 care: nearly pregnantobserved for students women of Pantawid in senior households avail of high school, where the recommended almost number 84 percent of prenatal of which checkups, non-poor children is slightly higher than the were enrolled 7 in 10 pregnant mothers of innon- school only about and households Pantawid 72 percent (Table of poor 2). The program, children however, has con tinued no impact ontheir education the utilization of postnatal care services within 72 hours, suggesting for the 15–18 year group. that beneficiary women do not give equal importance to both prenatal and postnatal care. Meanwhile, Pantawid pregnant women have higher incidence of giving birth assisted by a doctor by up to 10 percentage points compared Pantawid to non-and Health women. Nutrition Similarly, visiting of health facilities for weight monitoring as well as provision of deworming pill Outcomes at least twice are higher among beneficiary children by 9.1 percentage points and 8.5 percentage points, respectively. Consistent with previous IEs, the program has also showed positive impact on the intake of Vitamin A, sustaining its impact since the early stages of As in past studies, IE 3 indicates that Pantawid Pamilya promotes the utilization of prenatal the program implementation. care services and skilled birth attendance, which all have positive e ects on the long-term welfare of bene ciary households . Pantawid is encouraging poor women to use maternal and Table child health services 2. Summary such of IE 3 Results as antenatal care: for Health and nearly 8 inNutrition Outcomes 10 pregnant women of Pantawid Indicator number of prenatal households avail of the recommended Baselinecheckups, which is slightly higher Impact than the 7 in 10 pregnant mothers Maternal health of non- Pantawid households (Table 2). The program, however, has no impact on the utilization of postnatal care services within 72 hours, suggesting Prenatal checkup (4 times) 69.0 +7 pp 14 Skilled birth attendance (by doctor) 37.7 +10 pp Facility-based delivery 64.0 None Post-natal checkup within 72 hours 66.6 None Children’s health utilization Weight monitoring (0–2 years) 11.5 None Weight monitoring (2–5 years) 19.9 +9.1 pp Vitamin A supplementation* 69.0 +5.6 pp Immunization (complete) 18.0 None Deworming at least twice (6-14 years) 18.6 +8.5 pp Source: PIDS (forthcoming), Third Impact Evaluation of the Pantawid Pamilya CCT Program. www.worldbank.org 13 Pantawid Pamilya 2017 Assessment: An Update of the Philippine Conditional Cash Transfer’s Implementation Performance V. Conclusion and Recommendations This update has shown that Pantawid Pamilya maintains good impact on beneficiaries over the years since implementation, in terms of poverty reduction incidence, as well as main education and health outcomes. Still, there are several lessons and recommendations that need to be considered going forward, in particular considering the upcoming institutionalization of the Program through the CCT Bill implementing rules and regulations. The analysis presented suggest the following areas of attention: te of the Philippine Conditional Cash Transfer’s Implementation Performance1 Continue using an objective and regularly updated targeting database such as the Listahanan to select beneficiaries, and take opportunity of the rollout of the PhilSys (new unique ID system) to minimize exclusion errors. The DSWD updated its national household targeting system in 2015. Progress in underway in updating Listahanan in 2019-2020, and it’s important that when it is concluded Pantawid Pamilya conducts a full assessment and recertification of its beneficiaries. The most updated Listahanan database should also identify new poor deserving assistance and incorporate them in the program, resuming enrollment halted in the last four years. Using an updated targeting database will improve targeting performance of the Pantawid Pamilya and its effectiveness in reducing poverty. In the near future, Philippines should consider moving away from updates every three-four years and more into an on-demand approach with more frequent updating, and take advantage of the rollover of the new national ID system to ensure no one deserving to be in the program fails to be enrollment for lacking adequate identification. Regularly adjust the grant schedule. While the Pantawid Pamilya cash grant still retains its relevance as an incentive for poor beneficiaries to access health and education services, it has been perennially losing its real value. Beneficiaries are able to buy less and less with their cash grants. The additional cash rice allowance that was extended to Pantawid beneficiaries in 2017 is a welcome increase. However, it is a temporary adjustment. The provisions of the new CCT Bill enable regular adjustments in the grant structure and protect the benefits against the risk of inflation. What is needed is for these regular adjustments to be established in the annual budget/appropriation cycle of the Government. The Government can also start exploring customizing the grant structure by age/grade level, gender, location (e.g. geographically isolated areas) to affect specific outcomes as in other CCT programs (e.g., Mexico’s Prospera). Continue to keep abreast of the evolving needs of poor and vulnerable households that the program serves. After nearly a decade of implementation, the three rounds of impact evaluations show evidence that the program has already successfully met its objectives of bringing poor children to school and keeping them healthy. At the same time, it also augmented household resources so they can afford to buy their basic needs and alleviate the effects of sudden income shocks. The program is now mature enough to turn the attention to second-generation agenda on implementing CCTs. For instance, it can consider shifting the focus to influence necessary improvements in quality dimensions of education and health outcomes. This would entail adjusting program conditions and possibly the corresponding grant structure to provide enough incentives to both households and service providers. For example, attention can now be shifted towards ensuring children increase their grade transition rates, monitoring each health condition and providing a corresponding specific grant to ensure children receive medical/nutritional attention (complete age-appropriate immunizations, micronutrient supplementation, deworming pills), and that the Family Development Sessions (FDS) continues improving delivery of topics in an effective manner. Finally, complementation of the program with additional opportunities for livelihood support (such as training, microbusiness support, etc.) can also maximize the overall poverty impact and in the future reduce pressure of its continuous expansion. Other programs in the Philippines must learn from the experience of Pantawid Pamilya in terms of design, implementation, evaluation, and impact. For the Government of the Philippines to continue enhancing its capability in implementing programs that have proven ability to work, it must invest in putting in place systems for objective monitoring and evaluation, as well as accountability mechanisms as it has done with the CCT program. The lessons from Pantawid Pamilya is that impactful programs evolve to maintain relevance and generate consensus for continuity, even in time of political transitions. 14 WORLD BANK SOCIAL PROTECTION POLICY NOTE www.worldbank.org 15