TECHNICAL ASSESSMENT ADDENDUM 1. The Program supported by the Additional Financing (AF) addresses relevant and critical issues for economic development and poverty reduction in Pakistan. While poverty rates in Pakistan declined between 2001 and 2018, the COVID-19 pandemic pushed 2 million people below the poverty line and the floods in 2022 pushed an estimated 8.4 million to 9.1 million people into poverty. Although price pressures have moderated somewhat recently inflation remains high, affecting particularly the most vulnerable. Very low human capital levels remain a key constraint for economic growth and the gap in human capital outcomes between households in the richest two quintiles and households in the poorest quintiles has grown in the past 30 years. The AF focus areas are aligned with the priorities of the national poverty alleviation program pursued by the Ministry of Poverty Alleviation and Social Safety. Adaptive safety net delivery systems and human capital development remain strategic areas for the government. 2. The measures to be supported under the AF are technically sound. Given that the areas to be covered under the new DLIs closely align with those in the Parent program, in broad terms, the technical assessment for the original operation still holds. The AF Program aims to reinforce improvements to make the National Socio- Economic Registry (NSER) more dynamic, facilitate the use of the NSER, including through more ownership by the provinces, and consolidate its position as the preferred targeting instrument for poverty-focused programs in the country. The positive impacts of the Benazir Income Support Programme (BISP) will be ensured by adjusting the level and thereafter maintaining the value of the unconditional cash transfer (UCT) - Kafaalat - which makes a meaningful contribution to beneficiaries’ consumption. Agreements will be reached and initial actions taken, which clarify the respective roles of the federal and provincial governments on the Health and Nutrition Conditional Cash Transfer (CCT) programs, with the aim of opening more space for provincial actions in Punjab and Sindh. 3. Results Area 1 (Building Crisis-Resilient Delivery Systems): The UCT under the BISP provides consumption support to the poorest but its benefit value has tended to lose purchasing power over time. Between 2018 and 2022, the benefit level was increased a few times but despite this, the real value of the transfer is still not as high as in 2008. As a result, while evaluations indicated that BISP has had a significant positive impact on some measures of welfare such as consumption, savings, investments in health and education, and child nutrition, there was no significant impact on headcount poverty (Cheema et al, 2016; Nayab and Farooq 2020). The declining value of the benefit over time is one of the reasons suggested for this limited impact on poverty. Thus, an increase in its value is needed to reach the 15 percent of household consumption that evaluations1 suggest is necessary for BISP to have an impact in reducing poverty. Thereafter, the value of the transfer needs to be adjusted regularly to ensure an impact on poverty levels. Updating seeks to adjust for changes in prices, so that households can maintain their purchasing power. Many countries have instituted regular updates of benefit levels based on indicators such as CPI or minimum wages. Pakistan must follow suit to ensure that BISP can help households break the cycle of poverty. In addition, the enrolment of additional households currently in the NSER and eligible for the UCT will help to reduce exclusion errors. 4. Other DLIs under Results Area (RA) 1 reinforce measures in the original Program related to promoting the availability of up-to-date data in the NSER (through the expansion of mobile touchpoints) and its use in BISP through regular recertification of beneficiaries according to standard procedures. These measures will help sustain the good targeting performance of BISP and the availability of the benefit to those who need it most. Accessibility of the dynamic registration centres will be improved by the expansion of touchpoints to the 1 Pakistan Institute of Development Economics. 2020. Unconditional Cash Transfer and Poverty Alleviation in Pakistan: BISP’s Impact on Households’ Socioeconomic Wellbeing. PIDE Policy Viewpoint No. 18:2020 union council level as people living in remote areas as well as the poorest of the poor have significant mobility constraints2. International evidence suggests that the positive impacts of cash transfers are higher for those at the bottom of the income distribution. Wider use of the NSER as a common entry point for poverty-focused interventions in Pakistan, through its adoption as a national targeting mechanism, more provincial ownership through their participation in the BISP board, and the use of technology to facilitate its role in the delivery of the benefits of other programs, should lead to efficiency gains and leverage the investments made in it. The use of technology to support the exchange of NSER data with other programs will reduce the current use of semi-automated systems because most organizations lack the technological capacity to participate in bi- directional data sharing with BISP. The DLIs build on strong performance under the original Program. More than 650 dynamic registration centres were established in 157 districts equipping each tehsil with at least one dedicated centre, supplemented by 25 mobile registration vehicles to reach remote areas in Balochsitan and Sindh. As of now, these centres have processed the data of 8 million households. Interoperability of NSER with NADRA was enhanced resulting in increased accuracy and reliability of collected data. Two meetings of the NSER Steering Committee with provincial representation have been held so far to further improve coordination. 5. Results Area 3 (Protecting human capital in the early and primary school years): There have been substantial achievements under the original Program, compared to the baseline. The newly established health and nutrition CCT with a design based on global evidence3 was rolled out across all provinces while covering all districts in the federal domain. Currently over 1.8 million beneficiaries are enrolled in the program. The baseline for an impact evaluation has been carried out, with the midline measurement expected by the end of June 2024. Impact evaluations for the education CCT show positive results on enrolment raging between 8-14 percentage points, depending on the methodology, gender and age group, as well as a reduction in grade repetition4. Based on these positive results, the program has expanded from 1.1 million to 11.87 million beneficiaries ever paid against compliance. The cumulative number of ever enrolled children rose from 3.2 million to 14 million. 6. The new DLIs under RA3, seek to promote an agreement on the respective role of the federal and provincial governments for CCTs. While up to now, the federal government has played the lead on BISP, both on the UCT and CCTs, according to the 18th amendment of the Constitution5, social protection is a responsibility devolved to the provinces. To address the current ambiguity on roles and in some cases overlapping provincial and national CCTs, the DLIs include an agreement not to expand further any vertical federal CCTs. Specifically relating to the health and nutrition CCT, an action plan to remove current overlaps for the provinces of Punjab and Sindh, followed by a gradual transition of districts from the national to the provincial program is envisioned. These issues are not unique to Pakistan. A large share of safety net programs in the world involves multiple tiers of government. While the division of responsibilities for social assistance between the sub- national and national levels varies among federal countries, the arrangements supported by the AF, are aligned with international experience and public finance literature.6 First, the proposed arrangements for the health and nutrition CCT follow a typical pattern whereby sub-national governments supplement existing federal programs, in this case for a specific sub-group of UCT beneficiaries by adding a benefit linked to participation in additional activities. Second, recognizing the appropriate role for redistribution and emergency response, and 2 Guven, Melis, Zaineb Majoka, and Gul Najam Jamy. 2024. The Evolution of Benazir Income Support Programme’s Delivery Systems: Leveraging Digital Technology for Adaptive Social Protection in Pakistan Washington, D.C. World Bank (2024) 3 A large body of evidence on the efficacy of nutrition specific interventions exists and was reviewed in the 2013 Lancet series on Maternal and Child Nutrition. 4 Cheema, Iftikhar, Simon Hunt, Sarah Javeed, Tanya Lone, and Sean O’Leary. 2019. “Benazir Income Support Programme: Evaluation Report.� Oxford Policy Management, October 2019. 5 Passed by the Parliament in 2010 6 For a summary of the principles as applied to safety nets, see Ch.3 of For Promotion and Protection: The Design and Implementation of Effective Safety Nets, Margaret Grosh et al, World Bank (2008). to treat citizens in different localities alike, the central government will continue to finance the UCT, which accounts for the largest share of BISP expenditures. Third, functions such as the NSER and UCT and education CCTs, that are subject to significant economies of scale will be kept at the central level. Fourth, out of the provincial governments, Punjab and Sindh, the two provinces which would take on more responsibility for the health and nutrition CCT have higher financial and technical capacity, leaving the lead role of the central government for the rest. Finally, the social protection framework for cash assistance, agreement on which by national and provincial Cabinets would be supported by the AF would clearly specify roles, resolving current ambiguities. 7. BISP will continue to be the implementing agency for the expanded Program including AF requirements and has demonstrated its capacity under the parent Program. BISP will take on responsibility for implementing additional measures under RA1. The DLIs under RA3 do not imply additional tasks for BISP. BISP has established a strong foundation for an adaptive safety net system that has met expectations in all recent shocks/crisis. For example, in the face of devastating floods in 2022, BISP reached 2.8 million food-affected families within ten days, showing the efficiency of program delivery mechanisms. Furthermore, the parent Program has realized considerable achievements in all the RAs with some surpassing expectations. Specifically, in RA1, the dynamic update mechanism (MELA) is now operational in all Tehsils, updating NSER data for 8 million households. The development of an API has enabled the online and secure two-way exchange of data with 7 institutions, while staff of another two is being trained before they also receive data. Prior to using APIs, BISP had shared NSER data with more than 150 organizations and programs for various purposes. In RA3, the Taleemi Wazaif (TW) education CCT has expanded to all 160 districts across Pakistan, overachieving the expected results. Updated Economic Analysis 8. Raising the value for the base cash transfer and institutionalizing an indexation mechanism for the benefit can provide effective cushioning to poor households against the cost-of-living hikes from impending economic reforms and improve the program’s poverty impact in the long term. BISP’s poverty impact has seen a reduction over time due to the benefit amount not being more regularly adjusted to inflation. A study on BISP’s cash transfers’ impact on households’ socioeconomic well-being found that although the nominal value of cash transfers increased by 67%, the real value of transfers declined by 9% during the 2011-2019 period. Further, the study found that BISP beneficiaries continue to face high levels of food insecurity and that no significant impact is observed on the beneficiary households despite the consistent reduction in the Multidimensional Poverty Index over time. The analysis concludes that the reason for the limited impact on poverty levels is due to the declining value of cash transfers. Thus, while BISP has increased its benefit value over the years, it has not been sufficient to preserve its purchasing power. An increase in the benefit value followed by an indexation mechanism that ensures that such decisions are not ad hoc and if an appropriate index is selected, it will likely fill the gap between the nominal and real value of the benefits. This will maintain the purchasing power of beneficiaries and thereby improve the poverty impact of the UCT program - Kafaalat. 9. Regular NSER updates through the Dynamic Registration centers in each Tehsil supplemented by additional touchpoints at the union council level can significantly reduce the cost of data collection and update. The registration desks at the tehsil and union council level, where households can register to be covered by the program and existing beneficiary households can update their information, can help expand the coverage of households in the NSER database. This removes the need for complicated and costly door-to-door surveys in the future. The dynamic registration approach has the advantage of smoothing the costs of intake, registration, and updating over time. The experience of Columbia found a 43.5% decrease in cost per family by switching from door-to-door household registration to regular updates through local government offices. This transition to a dynamic registration model requires an initial investment, which includes physical and technological infrastructure and human resources. However, the significant reduction in per-household survey cost offsets this initial investment. In Pakistan, the cost of dynamic registration per household is estimated at US$0.53 (PKR 149), which is significantly less than the per household cost in the 2011 door-to-door survey, estimated at US$2.7 per household. However, it is expected that the cost borne by households will increase since they now need to travel to a dynamic center to update data. While it is hard to estimate, a recent study’s findings indicate that a BISP beneficiary, on average, spends PKR 371 on two-way transport to pick up the benefit, while beneficiaries who have to travel longer distances spend over PKR 1000. Tehsil offices, on average, tend to be located considerably farther away from households, leading to beneficiaries having to bear a substantial transport cost. Additional touchpoints at the union council level would help to improve accessibility, thereby reducing costs for beneficiaries. 10. Furthermore, regular updates will improve the targeting efficiency and shock responsiveness of NSER. After the first update of NSER in 2021, almost 60 percent of then-BISP beneficiaries were no longer eligible, i.e., their household welfare improved so much between 2011 and 2021 that they crossed the PMT threshold for eligibility. With Dynamic Registration Centers in place, data of each household will be updated at least once every four years. This will help further minimize inclusion and exclusion errors. At the same time, an updated NSER will also reduce the time taken to respond to crises. Timeliness of support in a crisis is essential to mitigate negative coping strategies, such as taking children out of school or skipping meals, the adverse effects of which can be intergenerational. In the aftermath of the devastating floods in Pakistan in 2022, NSER set up kiosks where households could register or update their information. This was done to ensure that the poor and vulnerable were not left out of the flood response. This causes delays in responding to such households, which can aggravate the welfare losses due to the time lag. 11. Furthermore, better targeting helps improve the effectiveness of social spending, improving the fiscal sustainability of cash transfer programs. Adoption of recertification protocols to ensure predictability in program exit and entry for BISP programs during future recertifications can reduce exclusion errors and improve the coverage of social assistance programs, promoting the optimal application of the targeting mechanism, thereby improving the programs’ poverty impact. 12. Adopting NSER as the exclusive welfare-based targeting mechanism will optimize the targeting of cash transfer programs in the country and their results. Better targeting would reduce costs incurred due to inclusion errors. The implementation of smartcards for pensions and employment programs in Andra Pradesh found 11 percent of existing beneficiaries to be ghost beneficiaries, causing significant welfare losses. Well-targeted cash transfers have considerable multiplier effects. Cunha et al. (2022) estimate a GDP multiplier effect of Brazil’s’ 2020 federal cash transfers to vulnerable households in the range of 0.5- 1.5. The most immediate multiplier effect is the economic gains that are realized through spending and consumption. Beneficiaries spending their cash may lead to an increase in demand for goods and services, injecting liquidity and stimulating economic activity. Cash transfers—especially CCTs such as Nashonuma and Taleemi Wazaif—can help in developing human capital. While in the cases of the CCTs, the impact may be direct, UCTs such as the Kafaalat could also trigger an uptake of services in education, health, nutrition, and early childhood development and reduce intimate violence, improving psychological well-being. 13. Furthermore, using an up-to-date unified registry would strengthen partnerships between BISP and the provincial governments leading to better understanding of local needs. This will not only make it possible for the government to respond rapidly after a shock hits, but will also enable policymakers to focus on building resilience among vulnerable populations even prior to the onset of shocks. Evidence suggests that people covered by cash transfer programs are 22 percent less likely to be food insecure in the event of a shock. Cash transfers contribute to the development of coping strategies for climate shocks, making beneficiaries more likely to reduce or mitigate their negative impacts. 14. Measures to prevent an overlap of federal and provincial CCT programs may aid in shaping a common fiscal framework and resolve constraints arising from systematic resource misallocation. Weaknesses in revenue and expenditure policies have contributed to economic imbalances in Pakistan, distorted resource allocation, and constrained productivity growth. Continuing to support agreements on an appropriate alignment of national and provincial social protection roles and and intergovernmental coordination can support fiscal consolidation, improve macroeconomic sustainability, address systematic resource misallocation, and increase the quality of spending with reduced allocations to administrative costs. These reforms would contribute to improving Pakistan’s fiscal space for pro-poor spending on basic services, social protection, and inclusive growth.