Additional Financing of Social Protection Support Project (RRP PHI 43407-014) UPDATED FINANCIAL ANALYSIS 1. The financial analysis for the proposed additional financing of the Asian Development Bank (ADB) Social Protection Support Project uses updated information to examine the financial viability, the efficiency, and the sustainability of the Pantawid Pamilyang Pilipino conditional cash transfer (CCT) program (Pantawid, formerly known as the 4Ps). The analysis finds Pantawid to be financially viable, efficient, and sustainable for several reasons. First, the fiscal space for the program is likely to expand due to expected rapid economic growth in the Philippines. Second, the national household targeting system for poverty reduction (the Listahanan) makes Pantawid a more effective and efficient alternative to many existing social assistance programs. Effective targeting makes the financial burden of helping the poor less costly. Third, Pantawid improves human capital, which will have a positive impact on returns to labor and thus taxable income. This will raise government revenues and expand fiscal space for future social spending. Fourth, Pantawid can facilitate politically difficult but critical policy reforms needed for the development of a progressive and stable economy that has expanding fiscal space for social spending. 1 2. The baseline scenario treated spending on social assistance as a fixed proportion of gross domestic product (GDP). Under this scenario, potential efficiency gains from reallocation or systemic improvements were ignored. Even so, investment in Pantawid and the Listahanan were found to be financially sustainable. The potential efficiency gains in the overall social protection portfolio could create additional fiscal space by lowering the overall government budget burden associated with any given level of public services. Pantawid is therefore highly likely to be financially sustainable. A. Financial Analysis for the Baseline Scenario 3. The Philippine economy has performed well since the 2008 2009 global financial crisis. GDP grew by an annual average of more than 6.0% during 2010 2014, despite slower growth in 2011. ADB expects growth to continue at about the same pace. 2 The debt-to-gdp ratio has been falling, the country s international credit rating has risen, and investor and consumer confidence has improved. Together, these developments indicate that rapid, stable growth will be sustainable, which in turn will expand the fiscal space for financing additional investments in poverty reduction and social protection programs. Specifically, prospects are good for more government spending on Pantawid and other major social assistance programs for the poor and vulnerable. The national budget for those programs stood at 0.90% in 2009 but rose to 2.46% in 2012 (Table 1). 4. Using fixed sector shares, a yearly GDP growth of 6.0% 7.0% would add P242 billion P252 billion per year in constant 2006 prices to potential fiscal space for social assistance (Table 2). This additional fiscal space could support the Pantawid budget requirements (Table 3), based on the Department of Social Welfare and Development (DSWD) medium-term expenditure plan (MTEP) for 2016 2019. Pantawid and the Listahanan targeting system have become flagship social protection programs. Their remarkable performance and the positive 1 For example, Mexico s CCT Oportunidades facilitated the government s reduction of an untargeted tortilla subsidy and the reallocation of funds to poverty-targeted social assistance. Indonesia s targeted unconditional cash transfers allowed the government to save billions of dollars in oil subsidies. Through the Bolsa Familia, Brazil s government rationalized fragmented and overlapping programs and reduced administrative costs. 2 ADB. 2015. Asian Development Outlook 2015: Financing Asia s Future Growth. Manila.
2 results from the DSWD s rigorous impact evaluations make them attractive for government funding vis-à-vis less effective public programs. Table 1: Comparison of Social Protection and Social Spending (%) Share of GDP Social Protection as Portion of Social Spending a Latest Social Social Social Social Region/ Country Assistance Insurance Protection Spending (a) (b) (c ) = a +b (d) = c + education + health Brazil 2010 15.5 27.1 b 57.2 East Asia and Pacific 0.9 2.6 3.5 9.5 36.8 Latin America Caribbean 1.3 3.8 5.1 13.4 38.1 Mexico 2011 5.0 12.8 b 39.1 Philippines 2005 0.2 1.4 1.6 5.7 28.1 2009 0.9 1.4 2.3 6.9 33.3 2012 2.5 1.6 4.1 7.1 57.7 South Asia 0.9 1.4 2.3 8.2 28.0 Thailand 2011 5.0 15.0 b 33.3 Viet Nam 2010 3.7 12.6 b 29.4 a Figures for the Philippines, Brazil, and Mexico are greater than those for Viet Nam and Thailand due to the inclusion of health and education grants in conditional cash transfer programs (counted under social protection) in the former group of countries. b Expenditure on social protection and health from International Labor Organization. 2015. World Social Protection Report 2014/15: Building economic recovery, inclusive development and social justice. Geneva; World Bank expenditure on education as percentage of government expenditure available at http://data.worldbank.org/indicator/se.xpd.totl.gb.zs/countries Sources: ADB. 2010. Report and Recommendation of the President to the Board of Directors: Proposed Loan, Technical Assistance Grant, and Administration of Technical Assistance Grant to the Republic of the Philippines for the Social Protection Support Project. Manila. Appendix 6: Financial Analysis; Government of the Philippines, Department of Budget and Management data. a b c d Table 2: Estimated Potentially Available Additional Budget for Social Assistance and Social Protection (P billion) Min. GDP and Min. SA/SP a Min. GDP and Max SA/SP b Max GDP and Min. SA/SP c Max GDP and Max SA/SP d 2016 Social assistance 90.7 248.0 92.5 252.7 2016 Social protection 161.3 413.4 164.4 421.2 2017 Social assistance 96.2 262.9 98.9 270.4 2017 Social protection 171.0 438.2 175.9 450.7 2018 Social assistance 102.0 278.7 105.9 289.3 2018 Social protection 181.3 464.5 188.2 482.2 2019 Social assistance 108.1 295.4 113.3 309.6 2019 Social protection 192.1 492.3 201.4 516.0 2020 Social assistance 114.6 313.1 121.2 331.3 2020 Social protection 203.7 521.9 215.5 552.1 GDP growth is projected at 6%. As a share of GDP at 2006 prices, social assistance is 0.9%, social protection is 1.6%. GDP growth is projected at 6%. As a share of GDP at 2006 prices, social assistance is 2.46%, social protection is 4.1 %. GDP growth is projected at 7%. As a share of GDP at 2006 prices, social assistance is 0.9%, social protection is 1.6% of GDP. GDP growth is projected at 7%. As a share of GDP at 2006 prices, social assistance is 2.46%, social protection is 4.1 % of GDP.
3 GDP = gross domestic product, SA = social assistance, SP = social protection. Source: DSWD Pantawid Pamilya medium-term expenditure plan. 5. The Pantawid conditional cash transfer program design has strong economies of scale, which indicates that the DSWD systems will have adequate absorptive capacity. This absorptive capacity, the government's strong commitment, and the institutionalization of Pantawid and the Listahanan within the MTEP all reinforce the finding of this analysis that the magnitude of ADB s proposed additional financing (Table 3) and of a proposed new World Bank loan is reasonable. Table 3: Summary of Financial Requirements for the Pantawid Pamilyang Pilipino Program, 2016 2019 (P billion) Total GOP ADB amount Pantawid amount (share) budget (share) 2016 P62.01 2017 P71.71 P232.51 P17.60 2018 P69.62 2019 P66.48 (86.2%) (6.5%) Total P269.91 ADB = Asian Development Bank, GOP = Government of the Philippines. Source: DSWD Pantawid Pamilya medium-term expenditure plan. World Bank amount (share) P19.8 (7.3%) 6. The analysis reassessed the long-term sustainability of Pantawid and the Listahanan targeting system following the completion of catalytic ADB and World Bank support, using recent data. Table 3 shows the resources that will be required to sustain planned expenditures. Ignoring possible savings generated under other programs and assuming zero real government budget growth, funding for the MTEP would correspond to investing about 2.27% of the government s total budget or about 0.5% of GDP in subsequent years. 3 Maintaining coverage for more than 4 million beneficiary households beyond 2019 could be more than fully accommodated by maintaining the 4.1% share of GDP already allocated to social protection expenditures in the 2012 budget. This level is within standard international practices of social spending. Complete program funding will be more likely since the number of beneficiaries is expected to fall over time as families exit from the program for reasons of non-eligibility. This suggests that Pantawid and the Listahanan can be well sustained beyond the end of ADB and World Bank support. Adding potential efficiency gains from rationalization and/or improved targeting in other social protection programs further strengthens this conclusion. B. Improved Efficiency of Government Programs 7. Continuing Pantawid will allow more time for the program to improve and to demonstrate that targeted cash transfers are an effective and efficient vehicle for helping the poor and the vulnerable. First, the second round of the Listahanan in 2015 and the recertification of beneficiaries this involves will provide the DSWD with the tools to reduce inclusion errors. Second, as Pantawid s potential is realized, and the resulting gains become more visible, the program will likely attract more public funds from other public programs that are less efficient and less pro-poor. Table 5 shows the distribution of beneficiaries of selected government programs by economic status and the fact that the Pantawid is the most strongly weighted toward helping the poor. Families below the poverty line made up 74.4% of total Pantawid beneficiaries in 2011 and 58.8% in 2013. However, this assessment is not based on households pre-transfer income, and the amount of the transfer moves some households 3 The time frame is 4 years: 2016 2019. The analysis used the national budget for 2015 and assumed it would be the same for the next 4 years. Annual GDP growth over the 4 years was assumed to be 6%.
4 above the poverty line. According to benefit incidence analysis, 82% of Pantawid beneficiaries belonged to the poorest 40% of the population, which compares very favorably to similar largescale CCT programs in Latin America. 4 The estimated inclusion error under the new Listahanan proxy means test model has been reduced to 10.6% for the National Capital Region and 13.8% for the rest of the country. With these improvements, the cost to the government of delivering a fixed amount of support to poor households will be reduced. 8. Efficiency can be improved not only by reducing the inclusion error, but also by lowering administrative costs and increasing the program s impact. The updated economic analysis calculated the administrative cost of Pantawid to be about 8% 10% of its overall cost. 5 Estimates for similar national programs put administrative costs at about 12.3% 13.9% for Brazil s Bolsa Familia and about 5.0% for Mexico s Oportunidades. 6 The Philippines Senate has recommended that the Pantawid administrative costs not exceed 10% of its total budget, and this target has been reflected in the Pantawid MTEP. The figure for conventional in-kind social assistance programs is about 15% or higher, taking into account wastage due to pilferage, expired medicines, and food rotting in warehouses. One review of social assistance estimates the median administrative cost of food-related programs at 25.4% of total cost, while for fee waivers the estimate is 16.0%. 7 DSWD s rigorous impact evaluations have shown that Pantawid has significant positive impacts on the well-being of the poor and vulnerable. 8 Table 5: Social Assistance Program Beneficiaries by Economic Status (%), 2011 and 2013 Poor Near-Poor Non-Poor Program 2011 2013 2011 2013 2011 2013 Pantawid (4Ps) program 74.41 58.81 13.84 17.78 11.75 23.41 Training for work program 10 20 20 0 70 80 CHED scholarship for Pantawid - 0-37.5-62.5 Community-based employment program 50 0 23.33 0 26.67 100 Scholarship benefits and students financial assistance (CHED) 19.84 9.09 10.91 15.15 69.25 75.76 Scholarship benefits and students financial assistance - 13.42-12.75-73.83 Supplemental feeding program 56.86 48.59 13.87 16.2 29.27 35.22 Cash and/or food for work programs 55.9 50 15.38 21.67 28.72 28.33 Disaster relief 42.93 35.54 14.75 12.40 42.32 52.07 Agrarian reform community development program 50-50 - 0 - Disability benefit 11.11-6.67-82.22 - Note: Program names are as they appear in the Annual Poverty Indicators Survey questionnaire. - = Program not included in the survey for given year, 4Ps = Pantawid Pamilyang Pilipino Program, CHED = Commission on Higher Education. Source: ADB estimates based on Annual Poverty Indicators Survey (APIS) 2011, 2013. 9. Savings from improved targeting. The government s adoption of the Listahanan targeting system will generate significant savings by lowering the cost to government of delivering any given level of benefits to the poor and thus greatly enhance the sustainability of 4 World Bank. Benefit Incidence Analysis of the Pantawid Program. Unpublished. 5 Economic Analysis (accessible from the list of linked documents in Appendix 2). 6 Lindert, K. et al. 2007. The Nuts and Bolts of Brazil s Bolsa Familia Program: Implementing Conditional Cash Transfers in a Decentralized Context. Social Protection Discussion Paper No. 0709: World Bank; Hodges, A. et al. 2011. Guidance for DFID country offices on measuring and maximizing value for money in cash transfers programmes. London. 7 Grosh, M. et al. 2008. For protection and promotion: the design and implementation of effective safety nets. Washington, DC: World Bank. 8 Department of Social Welfare and Development. Pantawid Pamilya Impact Evaluation Report. http://www.dswd.gov.ph/downloads-2/reports/pantawid-pamilya-impact-evaluation-report/ (accessed 15 April 2015).
5 an improved package of social protection programs. An executive order by the President in 2010 requires all household-targeted social protection programs to adopt the Listahanan, and 25 national programs had done so as of the third quarter of 2015. Even without other systemic improvements or realignment across programs, the savings this could generate by lowering losses from poor targeting in existing programs could create substantial fiscal space for the Pantawid conditional cash grants. The updated simulation undertaken for this analysis found that reducing the number of non-poor beneficiaries of selected social programs could bring savings of P11.6 billion, 9 which would cover about 966,400 Pantawid households a year. 10 These potential savings (in 2014 prices) amount to about 22% of the projected Pantawid funding requirements for 2016. Table 6: Estimated Potential Savings from Efficiency Gains by Adopting Listahanan under Selected Social Protection Programs (in P millions) 11 NFA Rice School-Based Training for Work TESDA Subsidy Feeding Program Scholarship Program Program budget 35,841.49 1,134.33 735.52 3,121.87 Potential savings 10,394.03 164.48 257.43 780.47 Total for selected programs Program budget 40,833.21 Potential savings 11,596.41 Potential Pantawid households covered by savings Assuming Php 12,000 per HH 966,367.00 Assuming Php 14,400 per HH a 799,752.00 Note: All figures are in 2014 prices. Program budgets are from 2013 except for SBFP, which is from 2014. NFA = National food Authority, TESDA = Technical Education and Skills Development Authority. a Proposed 20% increase under consideration by DSWD. Source: ADB estimates based on data from DSWD and Department of Budget Management. C. Broader potential savings 10. Pantawid could result in substantial efficiency gains in education and health. In education, high numbers of dropouts, low completion rates, and grade repetition result in major wastage. This is particularly true given that the returns to education exhibit nonlinearity i.e., they rise sharply as each higher level of education is completed. Data for 2011 2013 show that completion rates for both primary and secondary schooling have hovered at or below 70%, meaning that about one-quarter of all entrants will leave the system without completing their education. Both initial enrollments and completion rates tend to be substantially lower in poor areas and for children from poor households. Thus, by encouraging and helping children to enter and complete their schooling, the Pantawid program can promote the improved individual and social returns of education at the same time that it reduces a substantial waste of public resources. Similar efficiency gains may be expected for health interventions, in addition to the longer-term benefits of improved nutrition. 9 The analysis was based on the assumption of the following leakage rates for 4 selected programs: (i) 58% for the National Food Authority subsidy (Fernandez, et.al., 2012); (ii) 29% for School-Based Feeding Program based on the Food for School Program leakage rate estimated by Manasan (2009); (iii) 70% for the Training for Work Scholarship Program (APIS, 2013); and (iv) 50% for TESDA. PhilHealth was not included, since it now uses the Listahanan to identify its poor beneficiaries 10 Per-household cost was based on the DSWD s projected estimates. It is P12,000 if the status quo is maintained and P14,400 if a proposed 20% increase is implemented. 11 Note: The scenario assumes 0% real growth in benefit levels.