Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004781 IMPLEMENTATION COMPLETION AND RESULTS REPORT TF017877 ON A SMALL GRANT IN THE AMOUNT OF USD (4.55) MILLION TO THE Republic of India FOR Karnataka Multisectoral Nutrition Pilot (P149811) March 29, 2019 Health, Nutrition & Population Global Practice South Asia Region Regional Vice President: Hartwig Schafer Country Director: Junaid Kamal Ahmad Senior Global Practice Director: Timothy Grant Evans Practice Manager: Rekha Menon Task Team Leader(s): Abeyah A. Al-Omair ICR Main Contributor: Di Dong ABBREVIATIONS AND ACRONYMS ASHA Accredited Social Health Activist ANM Auxiliary Nurse Midwife BPL Below Poverty Line CPF Country Partnership Framework CPS Country Partnership Strategy EDF Energy Dense Foods ESMP Environment and Social Management Plan FM Financial Management FFSSAI Food Standards and Safety Authority of India GAIN Global Alliance for Improved Nutrition GMP Good Manufacturing Processes GDP Gross Domestic Product HCP Human Capital Project ICDS Integrated Child Development Services Scheme IEC Information, Education and Communication IUFR Interim Unaudited Financial Report IFA Iron and Folic Acid ISR Implementation Status and Results Report JSDF Japan Social Development Fund KCNM Karnataka Comprehensive Nutrition Mission KSRLPS Karnataka State Rural Livelihoods Promotion Society M&E Monitoring and Evaluation NFHS National Family Health Survey NGO Non-governmental organization NIN National Institute of Nutrition SC Scheduled Caste SOP Standard Operating Procedure ST Scheduled Tribe US$ United States Dollar VNV Village Nutrition Volunteers TABLE OF CONTENTS DATA SHEET ................................................................................. ERROR! BOOKMARK NOT DEFINED. I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 4 A. COUNTRY AND SECTOR CONTEXT ................................................................................................4 B. RATIONAL FOR BANK ENGAGEMENT.............................................................................................6 II. OUTCOME ...................................................................................................................... 8 A. RELEVANCE OF PDO ................................................................................................................8 B. ACHIEVEMENT OF PDO (EFFICACY) ..............................................................................................9 C. EFFICIENCY........................................................................................................................... 11 D. OTHER OUTCOMES AND IMPACTS.............................................................................................. 13 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................. 15 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME.... 16 A. BANK PERFORMANCE ............................................................................................................. 16 B. QUALITY OF MONITORING AND EVALUATION (M&E) .................................................................... 16 C. COMPLIANCE ........................................................................................................................ 18 D. RISKS TO DEVELOPMENT OUTCOME ............................................................................................ 19 V. LESSONS LEARNED AND RECOMMENDATIONS .............................................................. 19 A. KEY LESSONS LEARNED DURING GRANT IMPLEMENTATION ............................................................... 19 B. RECOMMENDATIONS .............................................................................................................. 20 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ............................................................ 21 ANNEX 2. PROJECT COST BY COMPONENT ............................................................................ 27 ANNEX 3. INNOVATIVE BENEFICIARY IDENTIFICATION METHODOLOGY................................. 28 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P149811 Karnataka Multisectoral Nutrition Pilot Country Financing Instrument India Investment Project Financing Original EA Category Revised EA Category Partial Assessment (B) Partial Assessment (B) Organizations Borrower Implementing Agency Department of Rural Development and Panchayati Raj, Karnataka State Rural Livelihoods Promotion Society Government of Karnataka (KSRLPS) Project Development Objective (PDO) Original PDO To increase utilization of nutrition-improving services by under-three children, adolescent girls and pregnant and nursing women from poor households in the target areas. Page 1 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) FINANCING FINANCE_TBL Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) Donor Financing TF-17363 4,550,000 3,480,558 3,480,558 Total 4,550,000 3,480,558 3,480,558 Total Project Cost 4,550,000 3,480,558 3,480,558 KEY DATES Approval Effectiveness Original Closing Actual Closing 25-Jun-2014 15-Jul-2014 31-Jul-2017 28-Sep-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 08-Jun-2017 1.35 Change in Loan Closing Date(s) Change in Implementation Schedule 26-Jul-2018 3.51 Change in Loan Closing Date(s) Change in Implementation Schedule KEY RATINGS Outcome Bank Performance M&E Quality Satisfactory Satisfactory Substantial RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 02-Sep-2015 Unsatisfactory Unsatisfactory 0.55 02 27-Jun-2016 Moderately Satisfactory Moderately Unsatisfactory 1.35 03 06-Dec-2016 Moderately Satisfactory Moderately Satisfactory 1.35 Page 2 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) 04 01-Nov-2017 Moderately Satisfactory Moderately Satisfactory 3.36 05 23-Jul-2018 Satisfactory Satisfactory 3.51 ADM STAFF Role At Approval At ICR Regional Vice President: Philippe H. Le Houerou Hartwig Schafer Country Director: Onno Ruhl Junaid Kamal Ahmad Senior Global Practice Director: Jesko S. Hentschel Timothy Grant Evans Practice Manager: Julie McLaughlin Rekha Menon Task Team Leader(s): Abeyah A. Al-Omair Abeyah A. Al-Omair ICR Contributing Author: Di Dong Page 3 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES Context A. Country and Sector Context 1. At appraisal, malnutrition remained a critical social concern in India, leading to high current and future economic losses. Despite high economic growth and a decrease in the proportion of population living below the official poverty line, the prevalence of malnutrition in children under three years of age was one of the highest in the world. Approximately one fourth of newborns had low birthweight, 45% of children under three were stunted and 40% were underweight in 2005-06 1 . Micronutrient deficiencies were pervasive, with approximately 70% and 57% of children 6-59 months of age being anemic and Vitamin A deficient, respectively. Undernutrition in the first two years of a child’s life (i.e. in the first 1,000 days after conception) results in potentially irreversible deficits in physical and cognitive development, affecting future health status, educational achievement as well as productivity and income in a life time 2 . On the macroeconomic level, undernutrition negatively affected India’s economic growth. Productivity loss associated with malnutrition in India (specifically, protein energy malnutrition, iodine and iron deficiency) had been estimated at US$114 billion between 2003 and 2012 3 . Studies also suggested that the productivity loss from stunting, iodine deficiency and iron deficiency together was almost 3% of gross domestic product (GDP) annually2,4. 2. While Karnataka had relatively high Gross Domestic Product (GDP) and high GDP growth among all Indian states, it had a large population that suffered from deep poverty and high food insecurity, particularly in the northern districts and in the more isolated non-irrigated and tribal areas. Household surveys found both persistently high overall malnutrition levels and significant inequalities. In 2005-06, 42% of children under- three years of age were chronically malnourished (stunted), unchanged from 1998-99. Anemia among women aged 15-49 years who were ever married was higher in 2005-06 at 52% compared to 42% prevalence in 1998- 99, and 58% of children under-three years of age in the poorest wealth quintile were chronically malnourished, much higher than the 25% prevalence in the highest quintile, raising alarms regarding inequalities in terms of access and utilization of nutritious foods, as well as services and practices known to reduce malnutrition. 3. The high inflation (8-9%) during the project preparation stage, mostly driven by rising food prices, worsened the access to nutritious foods by children under two years as well as pregnant women and adolescent girls, particularly among the vulnerable poor households. In fact, household surveys indicated that average caloric consumption had declined, and that in 2004-05 as many as 80% of rural households in India were considered to be "calorie poor"5. Spending on food represented 66% of total household spending among the poorest third of households in rural areas6. Rising food prices had likely reduced the ability of poor households to purchase foods or to access other goods and services that have an impact on nutritional status, 1National Family Health Survey (NFHS-3) (2016) https://dhsprogram.com/pubs/pdf/frind3/frind3-vol1andvol2.pdf 2GragnolatiM. et al. (2005) “India’s Undernourished Children: A call for reform and action,� World Bank. 3CARE India and Linkages India (2003)“PROFILES for India 2003.� 4Horton S et al. (2010) “Scaling Up Nutrition: What will it cost?� World Bank. 5Deaton A and Dreze J (2009)"Food and Nutrition in India: Facts and Interpretations," Economic and Political Weekly 44(7):42- 65. 6Dev SM (2011)"Rising Food Crisis and Financial Crisis in India: Impact on Women and Children and Ways of Tackling the Problem," Indira Gandhi Institute of Development Research, Mumbai. Page 4 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) notably clean water and health care. The need for investing in nutrition was therefore apparent. 4. Undernutrition has multiple determinants and therefore requires a multi-sectoral approach. At the most proximate level, undernutrition is determined by food intake, nutrition and health care behaviors and environmental health, which primarily includes access to safe water, sanitation and hygiene practices. Underlying factors such as income poverty, gender, education, poor access to basic amenities and services act as barriers, hindering improvements in nutritional status. Furthermore, it is critical to ensure good nutrition in the first 1,000 days of a child’s life since conception as this is the period when the child’s brain and body develop rapidly. Any damage or growth faltering during this period is largely irreversible. In the same vein, it is important for adolescent girls (who are likely to bear children in the future) and pregnant and lactating women to be well nourished to prevent intra-uterine growth restriction and low birth weight as well as to be able to practice appropriate child care and stimulation behaviors. 5. While the national and state governments had significantly increased support to social sector programs including nutrition program7, institutional and other bottlenecks continued to restrict access by many of the poorest families8. In 1993, the National Nutrition Policy was developed. Many effective interventions had been prescribed, such as popularization of low-cost nutritious foods, reaching the adolescent girls, fortification of essential foods and control of micronutrient deficiencies. However, these had not been translated to national programs that were backed by budgets, and there was inadequate awareness regarding appropriate nutrition practices among the public9. Some subsequent programs were not able to achieve large coverage. In Karnataka in 2005-06, the Integrated Child Development Services Scheme (ICDS), the government’s primary vehicle to address the problem of malnutrition, was only able to reach approximately 34% of children under-six years of age and 34% of pregnant women in the poorest quintile, leaving most potential beneficiaries uncovered. Therefore, more comprehensive nutrition programs and better beneficiary coverage were needed. 6. With high-level political commitment and support, the Karnataka Comprehensive Nutrition Mission (KCNM) had developed a comprehensive nutrition pilot program adopting a life-cycle approach. The Japan Social Development Fund (JSDF) grant supported this project (Karnataka Multisectoral Nutrition Pilot) to implement the above mentioned multi-sectoral nutrition program in two blocks in Karnataka state, and generated lessons to inform future scale up to other poor areas in the state and other low-income settings across the country. The project interventions were designed to meet three main objectives (Figure 1) : (1) addressing the proximate determinants of nutrition through the daily provision of nutritious food supplements to under-nourished children, adolescent girls and pregnant and lactating women from households below the poverty line (BPL); (2) addressing the intermediate factors that affect food intake and health status, such as facilitating and promoting access to nutritious foods, motivating adoption of appropriate health and nutrition household behaviors through intensive behavior change communication, and facilitating access to services in the areas of water and sanitation and livelihoods, in addition to health and nutrition; and (3) providing a unique learning opportunity for the government on effective multi-sectoral strategies at both government and household levels for improving nutrition outcomes. The project adopted an innovative life cycle approach to nutrition by focusing on three beneficiary groups: adolescent girls, pregnant women and children, while most nutrition projects focused only on pregnant women and children but missed out on adolescent girls. 7 ICDS Restructuring, Multi-sectoral nutrition program, NRLM, NREGA etc. 8World Bank (2011) "Social Protection for a Changing India," Washington. 9 Rao VS (2016) “�Under-nutrition in India – A Forgotten National Nutrition Policy without a National Programme�. Page 5 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) Figure 1. Project interventions targeted at different levels of the nutrition causal chain Adapted from World Bank (2011) “South Asia Regional Assistance Strategy for Nutrition� B. Rational for Bank Engagement 7. The Bank had supported several prior nutrition programs in India, notably ICDS and the Tamil Nadu Integrated Nutrition Program, as well as nutrition-related activities in other sectoral operations on livelihoods (in Andhra Pradesh) and local government development (in Bihar). There was a need to continue the support to public programs, as malnutrition remained a critical challenge. Importantly, the previous programs supported by the Bank had generated important lessons and knowledge that needed to be incorporated in future nutrition program designs. In addition, the World Bank was uniquely placed to support this Project for several other reasons: (i) comparative advantage in procurement, fiduciary, and environmental social oversight, (ii) complementarity with existing engagement at country level and synergies with ongoing similar initiatives carried out by other partners, (iii) ability to act as a catalyst to support piloting of innovative programs with potential for future scaling up, and (iv) extensive multi-sectoral experiences to help build capacity and knowledge to improve efficiency and synergistic delivery of public services across various social and development sector related to nutrition. Project Development Objectives (PDOs) 8. The Project Development Objective (PDO) was to increase utilization of nutrition-improving services by children under three years of age, adolescent girls and pregnant and nursing women from poor households in the target areas in Karnataka (Chincholi Block in Gulbarga District and Deodurga Block in Raichur District). Page 6 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) Key Expected Outcomes and Outcome Indicators Results Indicators 9. The following three PDO indicators were to measure progress against the PDO: (i) Percentage of targeted children under-three, adolescent girls and pregnant and nursing mothers who receive nutritious supplementary foods produced and supplied by the project; (ii) Percentage of targeted households who utilize other social sector programs with a potential impact on nutrition (specifically ICDS, health services, and water and sanitation services); and (iii) Percentage of targeted pregnant and lactating women who practice core child nutrition and health care behaviors (specifically initiation of breastfeeding within an hour of birth, exclusive breastfeeding, immunization, timely and adequate complementary feeding after 6 months which includes breastfeeding and feeding with 3+ food groups a minimum number of times per day, diarrhea management and handwashing). 10. In addition, there were four intermediate outcome (IO) indicators measuring critical inputs and processes that were necessary to achieve the PDO: (i) Percentage of targeted households receiving counselling on improved child care and feeding behaviors; (ii) Percentage of target beneficiaries weighed monthly by nutrition volunteers; (iii) Percentage of targeted households who receive information about available social programs from nutrition volunteers; and (iv) Percentage of planned nutrition volunteers in place. Components 11. To achieve the PDO, the Project included the following three components: Component 1: Increase consumption of nutritious foods and improve household nutrition-related knowledge and behaviors (planned: US$3.51 million, actual: US$3.16 million) . This component delivered direct support to children under-three, adolescent girls and pregnant/lactating women from poor and vulnerable households in the form of locally-sourced nutrition supplements coupled with support to encourage good household behaviors that would have a large impact on nutrition, notably breastfeeding, complementary feeding and hygiene practices. The high-energy nutrition supplements were locally produced using local farm produce such as millet (ragi), chickpeas (gram), cane sugar (jaggery) and groundnuts. Village Nutrition Volunteers (VNV) engaged under the project in each village implemented the program at the village level with the support of grassroots groups, including women's self-help groups (SHG) and village health and sanitation committees. These groups helped the nutrition volunteers identify and provide support to women and children facing food insecurity and malnutrition. Capacity building support were also be provided to women’s self-help groups. The implementation of this component was carried out by a contracted non-governmental organization (NGO). The contracted NGO set up production units and trained women’s self-help groups in the production of the high-energy supplements in accordance with state food safety regulations. The Page 7 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) contracted NGO also distributed the food supplements through the VNVs and SHGs to targeted beneficiaries. Component 2: Improve access to multi-sectoral interventions with an impact on nutrition (planned: US$0.15 million, actual: US$0.12 million). This component aimed to leverage interventions and services in several other sectors that had an impact on the nutritional status of poor families in the target areas. At the policy and administrative levels, coordination was strengthened between key programs. On the ground, contracted NGOs, community-based organizations, and village nutrition workers facilitated access by poor families to programs in various sectors, so that integrated support is offered to the targeted poor families. In addition, demand generation activities were carried out to empower vulnerable households and communities to demand services and benefits to which they were entitled. This included programs and services with an impact on nutrition, such as ICDS, health services (including treatment of severe acute malnutrition, immunization, diarrhea treatment, de-worming, micro-nutrient supplementation, and antenatal care), social safety nets (such as the national rural employment guarantee scheme), agricultural and livelihoods programs, and water and sanitation schemes. Innovative ways of engaging other sectors were also be explored under this component, such as ways of preventing wastage of horticultural products at the primary level and marketing this to the community. Component 3: Project management and Monitoring and Evaluation (M&E) (planned: US$0.90 million, actual: US$0.19 million). This component financed management capacity for implementation of the project, including the management costs of the implementing NGOs and the development of an effective information, education and communication (IEC) strategy which were monitored for assessing behavior change. Rigorous monitoring and evaluation were supported, including household surveys to measure nutritional status, household knowledge and behaviors, and access to services. Routine reporting and monitoring were also be ensured under this component. This component also promoted knowledge dissemination with a variety of stakeholders through briefing notes and knowledge sharing workshops. Significant Changes during Implementation 12. There were two restructuring of the project that changed the implementation schedule: (i) First restructuring was approved on June 8, 2017 to extend the Closing Date from July 31,2017 to July 31, 2018. This was required to complete planned project activities. (ii) Second restructuring was approved on July 26, 2018, to extend the Closing Date for two additional months, from July 31, 2018 to September 28, 2018. The extension was required for the Ministry to develop a sustainability plan and to issue a letter of commitment on using domestic finance and resources to sustain the interventions after project completion. II. OUTCOME A. Relevance of PDO Rating: High 13. The project was and remains highly relevant to the country’s development priorities. The project Page 8 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) consistent with the National Nutrition Strategy 2018 10 , which aims to: (1) to prevent and reduce undernutrition (underweight prevalence) in children (0- 3 years) by 3 percentage points per annum from National Family Health Survey 4 (NFHS 4) levels by 2022, and (2) to reduce the prevalence of anemia among young children, adolescent girls and women in the reproductive age group (15- 49 years) by one third of NFHS 4 levels by 2022. The government had in recent years significantly expanded resources for public programs in nutrition, health, rural development and social protection. This project added value to existing efforts, by pilot testing an innovative approach with potential systemic impact on the way social programs are directed in addressing nutrition outcomes. 14. The project directly contributed to the World Bank India Country Partnership Strategy (CPS) FY13-1711, particularly on Engagement Area 3 (Inclusion), and on Outcome 3.3 (improved child nutrition delivery systems). The project is also well aligned to the India Country Partnership Framework (CPF) FY18–2212. The CPF FY18–22 includes three focus areas for the World Bank Group’s engagement. These are (1) promoting resource-efficient growth, (2) enhancing competitiveness and enabling job creation, and (3) investing in human capital. The project is consistent with focus areas (2) and (3). B. Achievement of PDO (Efficacy) Rating: High Overall Efficacy assessment 15. The achievement of PDO is rated as High because the project exceeded all of its PDO and IO targets. (see Table 1). The project met its objective of targeting the poor and vulnerable groups. The commonly used beneficiary selection criteria were bottom 40% of households or those lived below the poverty line (BPL), although consultations with local stakeholder revealed that there might be more poor households in need of nutrition interventions, as the two project blocks were the most backward blocks in the state of Karnataka where poverty and malnutrition were most prevalent. The project therefore adopted a systematic approach for more accurate beneficiary identification (see details in Annex 3). Briefly, the project used data from the 2011 Socio Economic and Caste Census (SECC) and applied twelve socio-economic parameters (automatic inclusion criteria included households without shelter, primitive tribal groups, households whose main occupation was scavenging and legally released bonded laborers; and 8 sets of deprivation indicators were used to rank the remaining households) to identify target households. This beneficiary identification exercise found 70% of residents in the project areas were poor, far above the previously estimated 40%. Following a data validation process, which included a household survey, 29,786 households (with 37,781 beneficiaries, including children 0-36 months of age, pregnant and lactating women and adolescent girls) were finally selected under the pilot project. Among project beneficiaries, 51% were adolescent girls, 13% were pregnant and lactating mothers, and 35% were 7-23 months-old children. Around 53% of all project beneficiaries and 50% of VNVs were from the traditionally disadvantaged Scheduled Castes and Scheduled Tribes (SC/ST) communities, which far surpassed the 30% requirement in the Tribal and Vulnerable Groups Development Plan. In addition, 50% of the beneficiaries were from socially and economically disadvantaged groups. 10 http://niti.gov.in/writereaddata/files/document_publication/Nutrition_Strategy_Booklet.pdf 11 http://documents.worldbank.org/curated/en/207621468268202774/India-Country-partnership-strategy-for-the-period- FY13-FY17 12 http://documents.worldbank.org/curated/en/277621537673420666/India-Country-Partnership-Framework-for-the-Period- FY18-FY22 Page 9 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) 16. All PDO targets and IO targets were exceeded (see Table 1). In addition, feedback from key stakeholders and beneficiaries had also revealed the positive impact of this project on changing behaviors and improving nutrition status of beneficiaries. Table 1. Project Achievements by Indicators Indicator Final target Actual value Achievement PDO indicators 1. Percentage of targeted under-three children, adolescent girls and pregnant and nursing mothers who receive nutritious 80 95 Surpassed (119%) supplementary foods produced and supplied by the project 2. Percentage of targeted households who utilized social sector programs with a potential impact on nutrition (specifically ICDS, 60 89 Surpassed (148%) health services, and water and sanitation services) 3. Percentage of targeted pregnant and lactating women who 60 100 Surpassed (167%) practice core child nutrition and health care behaviors Intermediate results indicators 4. Percentage of targeted households receiving counselling on 90 100 Surpassed (111%) improved child care and feeding behaviors 5. Percentage of target beneficiaries weighed monthly by 90 95 Surpassed (106%) nutrition volunteers 6. Percentage of targeted households who receive information 90 100 Surpassed (111%) about available social programs from nutrition volunteers 7. Percentage of planned nutrition volunteers in place 90 107 Surpassed (119%) • PDO indicator 1: 32,882 of out of the 34,431 targeted beneficiaries (95%) had received nutritious supplementary foods produced supplied by the project’s production units, far surpassing the target of 80%. In collaboration with Global Alliance for Improved Nutrition (GAIN), the project established two Energy Dense Foods (EDFs) production units licensed by the Food Standards and Safety Authority of India (FFSSAI) and used local farm products to produce 3 types of high-energy nutrition supplements for children, adolescent girls and lactating and pregnant women, respectively. The supplements were distributed by over 600 VNVs to targeted under-three children, adolescent girls and nursing mothers. The intake of nutrition supplements was expected to reduce malnutrition. • PDO indicator 2: After the Project became effective, household counselling on good nutrition behaviors conducted by VNVs creating significant increased demand for social programs that could improve nutrition outcomes, such as applications for construction of toilets, and demand for Iron and Folic Acid tablets (IFA). By March 2018, 24,731 of 27,484 (89%) targeted households utilized one or more social sector programs with a potential impact on nutrition (specifically ICDS, health services, and water and sanitation services), far better than the pre-set target of 60%. By August 2017, 1,210 toilets had been constructed in Chincholi block and 1,424 in Devadurga block taking benefit of the Swatch Bharat Mission Open Defecation Free (ODF) scheme. • PDO indicator 3: 100% (4,379 out of 4,379) of targeted pregnant and lactating women reported practicing all 6 core child nutrition and health care behaviors according to Health Information System data. Core child nutrition and health behaviors were defined as: (1) initiation of breastfeeding within one hour of birth; (2) exclusive breastfeeding up to 6 months of age; (3) ensuring recommended immunizations; (4) complementary Page 10 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) feeding which includes breastfeeding up to 2 years of age and feeding with 3+ food groups a minimum number of times per day; (5) diarrhea management; and (6) handwashing. 17. IO indicators monitored by the project • Intermediate results indicator 1: All 27,484 (100%) of households received counseling from VNVs on improved child care and feeding behaviors. • Intermediate results indicator 2: 95% (32,882 out of 34,431) targeted beneficiaries were weighted monthly by VNVs. • Intermediate results indicator 3: 100% (27,484 out of 27,484) households received information from VNVs on the availability of different social schemes related to reducing malnutrition and how to access them. • Intermediate results indicator 4: 418 Village Nutrition Volunteers (VNVs) were planned for under the project, and by February 2016, 447 had been recruited in all 421 target villages in Chincholi and Devdurga Blocks (107% of planned), surpassing the target of 90%. The volunteers were trained and carried out various important functions of the project. C. Efficiency Rating: Substantial Economic Analysis 18. An economic analysis was conducted as project preparation stage based on global evidence. This section provides a more detailed analysis. Although health outcomes were measured at project baseline and endline, there were no comparable data at the two-time points. Therefore, the economic benefit of the project was estimated and assessed using global and India-specific evidence on cost-effectiveness of nutrition interventions. 19. Improving nutrition among children, adolescents and pregnant and lactating mothers have life-long consequences for health, education, employment, and economic growth. In the short term, undernutrition increases the risk of mortality and morbidity among newborn children, young children and adolescents. In the longer term, the consequences of malnutrition and stunting extend to adulthood, increasing the risk of impaired cognition that results in poor school performance, poor pregnancy outcomes that pass on to next generation, reduced economic productivity and earnings, and future risk for overweight contributing to the rise in non-communicable diseases, such as hypertension and cardiovascular diseases; and increases the potential for intergenerational transmission of stunting and poverty. These consequences add up to 0.2 to 3.6 years less of schooling, and as much as 22% loss of early income in adulthood.13 Childhood anemia alone is associated with 2.5% drop in adult wages. 14 Reductions in stunting are estimated to potentially increase overall economic productivity, as measured by GDP per capita, by 4–11 percent in Africa and Asia15. 13 Grantham-McGregor, S; Cheung, YB; Cueto, S; Glewwe, P; Richter,. L; Strupp, B. 2007. Developmental potential in the first 5 years for children in developing countries. Lancet 369 (9555): 60-70 14 Horton S and Ross J. 2003. The Economics of Iron Deficiency. Food Policy 28:517-5 15 Horton, S., and R. Steckel. 2013. “Malnutrition: Global Economic Losses Attributable to Malnutrition 1900–2000 and Projections to 2050.� In the Economics of Human Challenges, edited by B. Lomborg, 247–72. Cambridge, U.K.: Cambridge University Press. Page 11 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) 20. The cost of malnutrition to the economy is significant in India. It has been estimated that in India productivity losses (manual work only) from stunting, iodine deficiency, and iron deficiency together are responsible for an annual loss of 3 percent of GDP.16 It is further suggested that micronutrient deficiencies alone may cost India $2.5 billion annually, about 0.4 percent of India’s annual GDP.17 One estimate suggested that the productivity losses in India associated with undernutrition, iron deficiency anemia, and iodine deficiency disorders, in the absence of appropriate interventions, would amount to about US$114 billion between 2003 and 2012 (India’s GDP in 2012 was US$ 1,841 billion).18 A more recent analysis indicated that stunted children earn as much as 20% less than their counterparts, costing the global economy potentially US$125 billion annually by 2030 when children born now reach working age, with India accounting for nearly US$46 billion of this amount.19,20 21. In Karnataka, nutrition deficiencies among children under 5 were estimated to cost at least 0.3% of Karnataka’s GDP. Nutrition deficiencies were the 6th top cause of deaths under 5 (causing 439 deaths in 2016), and the 4th leading cause of Disability-adjusted life years (DALYs) (causing 232,813 DALYs in 2016)21. Applying the commonly used societal willingness to pay of 1x GDP per capita per DALY ($2,500 in 2017), nutrition deficiency among children under 5 results in economic cost of US$582 million, or 0.3% of Karnataka’s GDP in 2017. 22. The benefit of this project was estimated at US$14 million. Many nutrition-improving interventions in the project have been shown to be very cost-effective and having high return to investment by global and India evidence, such as nutrition supplements in pregnant and lactating mothers, children and adolescents, exclusive breastfeeding, and multi-sectoral interventions. A global nutrition investment study showed that the return to investment to a comprehensive package of nutrition interventions for every $1 invested is between $4 to $3522. Roughly dividing the total project cost by total number of project beneficiaries yielded a per beneficiary cost of around US$100, which was higher than the reported costs of previous nutrition programs in India (US$5-29 per person). The higher per capita cost was likely due to the comprehensive program coverage, high quality standard of the production units, good village nutrition volunteer compensation, extra resources invested to better reach and follow up with the high proportion of migrant population in the areas, and the Information Education Communication campaign conducted. Given the higher per capita cost, the project return to investment was assumed to be on the lower end of the literature range (US$4 per US$1 invested). Giving the cost of the project at US$3.5 million, the economic benefit of the project is US$14 million. Therefore, this project was expected to have high economic return. Implementation Efficiency 23. There were delays in project implementation which required two extensions, reducing its implementation efficiency somewhat. After the project became effective on 15 July 2014, there was a 16 Horton S (1999) “Opportunities for investments in nutrition in low-income Asia,� Asian Development Review 17: 246-273. 17Alderman H (2005) “Linkages between Poverty Reduction Strategies and Child Nutrition: An Asian Perspective," Economic and Political Weekly 40(46): 4837-42. 18 CARE India and Linkages India (2003) “PROFILES for India 2003.� 19 Save the Children (2013) “Food for thought: Tackling child malnutrition to unlock potential and boost prosperity�. Save the Children, UK. 20 Horton S et al. (2010) “Scaling Up Nutrition: What will it cost?� World Bank. 21 GDB India Compare 22 An Investment Framework for Nutrition. 2017. World Bank Group Page 12 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) significant delay in implementation. Three main causes for this delay included: (1) The project had a slow start due to delays in contracting the implementing NGO, which was critical to carry out the comprehensive services and activities. After the selection of implementing NGO was made, clearances from the Government of Karnataka took longer time. As it was considered a high-value contract, higher than usual levels of approval from the State Government were required. The contract between Karnataka State Rural Livelihood Promotion Society (KSRLPS) and the NGO was signed in June 2015. (2) The selection of beneficiaries took additional time, as the team switched from a simpler Below Poverty Line approach to a more comprehensive beneficiary selection process as requested by KSRLPS. The new methodology required the use of Socio Economic and Caste Census (SECC) data which were only released in Karnataka in August 2015, after which additional survey-based verification had to be conducted for accurate selection of beneficiaries. The selection of beneficiaries was fully completed in both blocks only by December 2015. (3) A local election and the Election Code of Conduct was in operation from early January 2016 to February 2016. As a result, several preparatory activities got delayed as no new project activities or appointments/selections could be done during this period. Soon with the joint effort of the Karnataka State Rural Livelihood Promotion Society (KSRLPS), KCNM and the Bank, the project gained momentum, and rapidly achieved its targets. However, the 14-month delay in the project implementation timeline, resulted in two Closing Date extensions (in June 2017 and July 2018, respectively). 24. While the project has achieved its development objectives, the project did not fully disburse the allocated grant amount, reflecting both efficiency gains from mobilizing other resources as well as some forgone opportunities. Of the US$4.55 million grant from JSDF, US$3.48 million was disbursed, resulting in a disbursement rate of 77%. The partial disbursement was due to two factors: 1) The project team managed to acquire donations from GAIN to provide free machines that were used in the two EDF production units to produce nutrition supplements, resulting in savings as machine procurement was no longer being funded under the Project; and 2) The planned project impact evaluation was not implemented by KSRLPS and KCNM as planned, due to administrative and logistics considerations. The evaluation was replaced by intensive monitoring and evaluation and two smaller scale surveys. The US$1.04 million outstanding balance could have been re-allocated to other activities. However, due to the delay and low fund disbursement in the early phase of the Project, the KSRLPS and KCNM teams decided not to implement additional activities in order not to further delay project completion. Overall Outcome Rating Rating: Satisfactory 25. The overall outcome rating is Satisfactory based on the ratings for relevance, efficacy and efficiency. While an HS rating would be derived based on IPF rating guidelines, the project rating is downgraded to S to take into account some implementation delays and minor inefficiencies. Relevance Efficacy Efficiency Overall Outcome High High Substantial Satisfactory D. Other Outcomes and Impacts 26. The project has significantly improved the management capacity of the recipient, particularly in financial Page 13 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) management (FM), procurement, and monitoring and evaluation. By establishing two EDF production units and setting up operation procedures, the project also introduced good manufacturing practices and standards to the community. 27. The project successfully leveraged multi-sectoral linkages to achieves its objectives. Within the health and nutrition sector, coordination mechanisms between VNVs, anganwadi (rural health center) workers, Accredited Social Health Activist (ASHA), medical officers, and auxiliary nurse midwife (ANM) were established at the grassroots level to promote good team work between the various functionaries. An anemia health check-up camp was also organized in collaboration with the Health Department. In the sanitation sector, the project leveraged on the Swatch Bharat Abhiyan ODF, which constructed over 2,600 toilets for the project areas. The project leveraged the education sector by orienting 204 high school teachers and 69 middle school teachers to bring hygiene practices to schools and ensure regular distribution of IFA tablets. Local governments (gram panchayat/Taluk panchayat) members were oriented about the benefits of the project activities so that they could motivate the beneficiaries to consume EDF and adopt good hygiene practices. 28. The project managed to mobilize additional financial and human resources. As mentioned above, the project team managed to acquire donations from GAIN to provide free machines that were used in the two EDF production units to produce nutrition supplements, resulting in savings. To obtain additional human resources for community outreach activities, VNVs mobilized 1138 Adolescent Girls groups at the village level and educated them on the importance of personal hygiene, and the harm of child marriage and teenage pregnancy. In addition, 2,305 Self-help groups (SHG) leaders and 10,621 SHG members have been oriented on the benefit of EDF and helped distribute them. 29. The project funded two innovative mini-projects: 1) A study to explore the feasibility of commercialization EDF production units and introduction of low-cost EDF to the market was conducted. This was considered crucial to improve the sustainability of the model after project completion. Findings suggested that there was demand for low-cost EDF; and that advertising and promotion, taste and availability of products would be the major factors that lead consumers to purchase them. 2) A 30-hour module on food/nutrition security and public policy was developed by KCNM to be used in post-graduate programs in public policy, nutritional sciences and social work. The module had been introduced as a multi-disciplinary course in five universities in Karnataka and was expected to raise more professional and policy awareness of nutrition. 30. The project had a positive impact on improving the nutrition practices of women and girls, and also empowering women. Majority of project beneficiaries were female. Improved nutrition practices could improve their own nutritional outcomes and have positive inter-generational effects on their children. All village nutrition volunteers were women. Having a role in helping the community made these volunteers feel empowered, especially when they were provided bicycles and supplies to distribute supplements to the beneficiaries. Production unit workers were also women. They formed a self-help group (SHG) in the first year to learn how to operate the unit. They also drove trucks carrying materials to the factory. In the district of Gulburga, it was rare to see a woman driving a car. Therefore, a woman driving a truck, signaled women breaking the glass ceiling and gender stereotypes. 31. The good experiences of the project resulted in several best practice examples that can potentially be replicated in other rural areas, including: 1) use of a comprehensive beneficiary identification methodology to identify the group in need, 2) delivering nutrition-specific interventions through extensive community Page 14 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) outreach by Village Nutrition Volunteers, 3) promoting convergence by leveraging services and programs in other sectors that have positive impacts on nutrition outcomes of the beneficiaries, and 4) strengthening local production capacity and job creation. III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME 32. Factors that positively affected implementation were as follows: • A well-designed project with easily measurable Results Framework and clear activity plans for each component of t. • Joint commitment by KSRLPS and KCNM strengthened the implementation capacity and brought additional resources to the Project. With very strong technical capacity and leadership, the KCNM significantly accelerated the project implementation. • Good financial and non-financial incentives to the VNVs after adjusting based on VNV feedback was able to motivate and inspire VNVs to carry out the work in the villages. • Extensive community outreach activities by VNVs and mobilization of SHGs and Adolescent Girls Groups to expand Project coverage. • Simple yet effective mechanism for facilitating multi-sectoral convergence that focused on the beneficiary households and tracked if the household was benefitting from all available relevant nutrition specific and sensitive interventions and schemes offered by the different line departments (e.g. water, sanitation, education) in their block. 33. Factors that presented challenges to implementation: • Delayed procurement of the NGO (KHPT) to implement the project significantly delayed implementation. The process was accelerated after strong effort from KRSLPS, KCNM, and the Bank. • The planned robust impact evaluation study could have enhanced the validity of the evidence on the health impacts of the project. The initially designed survey-based difference-in-difference impact evaluation study was not carried out, due to the project recipient’s unwillingness to spend large amount of fund on impact evaluation, concerns on loss of project intellectual property and data ownership, and the complex approval procedures needed for high-value procurement contract. The impact evaluation and the accompanying process evaluation could have informed and helped further improve project implementation and provided additional data on the nutrition and health outcomes of beneficiaries that resulted from the behavioral changes observed in the project. • Establishing and operating of the EDF production units encountered many operational challenges, from machine acquisition and set up, to obtaining licenses, developing the SOP procedures, supply of ingredients, equipment maintenance, sub-optimal production quantity, and supply chain, and timely and sustainable funding. Strong support from the KSRLPS, KCNM, World Bank and GAIN jointly helped resolve most of the challenges and improved the operation of the EDF production units. • Coordination challenges also occurred between the implementation agency KSRLPS and technical advising agency KCNM. The project required extensive coordination on technical and management tasks. However, the separation of technical advisor role and implementer role did not occur, resulting in two agencies leading the implementation. Some challenges arose from different opinions, requirements of procedures, and ambiguity in roles and responsibilities on some specific matters. An Page 15 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) important lesson learnt is to either have a single entity to provide both implementation and technical oversight, or to separate and reinforce the specific roles and responsibilities of different agencies involved. IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. Bank Performance Rating: Satisfactory 34. Bank Performance at project preparation is rated as Satisfactory. The task team worked closely with KSRLPS, KCNM and other technical partners to ensure that the project was aligned with key nutrition strategies and targets at the National and State levels. The team maintained a good balance between introducing multi-sectoral global best practices in nutrition and meeting the specific needs of the target population in the Devadurga Block and Chincholi Block. The social context of high portion of disadvantaged groups was also taken into consideration in project preparation. To reach the population most in need of nutrition services, the project introduced an innovative beneficiary identification process, to achieve better beneficiary targeting compared with commonly used methods. The task team worked with the government team to add several innovative interventions in the project, including the establishment of EDF production units that used local farm products and employed workers to produce nutrition supplements, and development of a nutrition and policy course for five universities, contributing to the local economy and professional education. 35. The Bank supervision is rated as Satisfactory. The Bank team comprised of technically strong specialists in relevant areas and well informed on Bank policies, including experienced nutrition specialists, health specialists, a consultant with expertise in food safety and quality, a financial management specialist, a procurement specialist, and a safeguard specialist. Having one team member with deep understanding of the context of Karnataka was an advantage and improved project oversight. The Bank team conducted five supervision missions and many additional consultation visits. In each mission, the team monitored the progress, identified challenges in Project implementation including underlying causes, discussed solutions and action points with a feasible timeline, and provided extensive capacity building support on implementation and project management, all of which were documented in detail in Aide Memoires. The team also conducted four field visits to the EDF production units and many consultation visits to the project sites and the community. The EDF production unit visit allowed the team to provide detailed feedback on improving the operation of the unit to meet the demand, address the root causes of operational challenges, and institutionalize environmental safeguards (such as the development of SoPs to improve hygiene and safety). During the field visits, there were extensive interactions with the VNVs, beneficiaries and local leaders to gather feedback and improve project Implementation. The FM and procurement specialists also provided on- the-job capacity building. In addition, Bank funded the baseline M&E survey to collect necessary data for project implementation after a significant delay in client’s procurement of M&E agency. However, this led to early exhaustion of Bank’s supervision budget. B. Quality of Monitoring and Evaluation (M&E) Page 16 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) Rating: Substantial 36. M&E design is rated as Substantial. The Project Results Framework indicators reflected the PDO. The RF consisted of a good combination of input indicators, output indicators and behavior change indicators, that were measurable and relevant. Baseline values, target values and data sources were also clearly identified. 37. The M&E design included three major efforts: (1) M&E through the development of an electronic information system: an electronic information system was to be developed which village health volunteers would use to collect and upload data on the indicators in the RF and other program indicators; (2) Routine M&E through project supervision: to provide qualitative and quantitative monitoring, evaluation, and feedback on project implementation and outcomes; and (3) A survey-based impact evaluation study: using a difference-in-difference methodology and two surveys (at baseline and endline) to collect data on the health status and behaviors of different beneficiary groups and non-beneficiaries (control group) before and after the project. While the project RF centered around service utilization and behavior change among beneficiaries, which was the common practice for nutrition programs, this study was expected to complement RF indicators, and provide additional information on project development impacts in terms of nutrition and health outcomes. In addition, it would also provide the opportunity to triangulate and verify routine monitoring data with survey data. 38. M&E Implementation is rated as Modest as there were some shortcomings in part of M&E implementation. In terms of the M&E through information system, an electronic information system, MIS, was developed as planned, and village health volunteers collected and uploaded data on the RF and other program indicators using the system. Despite 3-4 months of delays in collecting data manually and entering to the information system, the system is functional. The routine M&E through supervision too was relatively well implemented. The task team produced five ISRs, five Aide-Memoire, and two field-visit reports during project implementation. Achievements of PDO indicators and Intermediate Results Indicators were reported in all ISRs. Findings and recommendations on implementation were well documented. However, the impact evaluation study was not carried out as planned. Concurrent monitoring agency procurement was classified as high-value procurement, and the client had significant challenges and hesitation with procurement of an M&E agency. Specific reasons included the unwillingness to spend large amount of funding on impact evaluation, concerns on loss of project intellectual property and data ownership, and the complex approval procedures needed for high-value contract. After several delays, the client decided not to hire a concurrent monitoring agency and decided not to conduct the impact evaluation as planned. The lack of robust impact evaluation through well-designed surveys incurred three risks: (a) the lack of an independent data source to verify and triangulate implementer-reported data in the electronic information system; (b) the lack of detailed data on various health and nutrition practices and status among the beneficiaries and non-beneficiaries, and (c) challenges in quantifying the project impact on key nutrition and health outcomes, such as stunting rate and anemia rate. 39. To mitigate the risks of not conducting the planned impact evaluation study, a simplified impact study was carried out, and two surveys were conducted in the beginning and upon completion of project. In the first year, a baseline survey was funded by the Bank budget in Project blocks and control blocks to assess 27 Indicators under 4 categories: (1) Nutritional and Health outcomes (such as BMI, Hemoglobin, low birth weight, height for age), (2) Knowledge (such as knowledge for nutrition during pregnancy, child feeding practice, hand washing, menstrual hygiene, immunization schedule), (3) Behavior change (such as ANC/PNC Page 17 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) visits, early initiation of breast feeding), and (4) Uptake of Schemes. Upon project completion, a small-scale survey was done by the National Institute of Nutrition (NIN). The survey found the intervention blocks to have better outcomes than the control blocks in different domains, including: lower stunting rate of children, lower anemia rate in adolescent girls, better awareness of nutrition, health and sanitation related issues and utilization of various government programs. However, the scale and design of the end-line impact study was not as robust as planned in project preparation stage. 40. M&E utilization is rated as Substantial. Findings from M&E were used to improve project implementation. Challenges and issues identified during routine M&E and supervisions were documented in field visit reports by the Bank team and shared with key stakeholders and implementers. Recommendations and actions plans were made, reinforced and monitored to improve project implementation. C. Compliance 41. Environmental Compliance: Although Safeguard Policy (OP4.01 Environment Assessment) was triggered, due to the establishment and operation of the EDF production units, the environmental performance was considered satisfactory as adequate measures had been implemented as proposed in the Environment and Social Management Plans (ESMP) to address identified stakeholder concerns. Concerns identified during four supervision visits to the EDF production units by the safeguard specialists (such as cleanliness of the EDF production unit, the need for Standard operating procedures for hygiene and safety, the need for independent quality testing for raw materials and final products, and pest control) were communicated to EDF management team and KSRLPS for close follow up. As a result, SOPs for hygiene and safety were developed and environmental safeguards mainstreamed. 42. Social Accountability: Social Safeguard Policy (OP 4.10 Indigenous Peoples) was triggered by the Project. Adequate measures had been taken to address identified concerns in the implementation. The two blocks covered by the project have a high proportion of residents living in the ST/SC communities. The Project design included various actions to ensure that these disadvantaged groups benefit from the project. These include: a comprehensive beneficiary identification exercise that treated ST/SC as one of the automatic inclusion criteria, hiring additional VNVs to provide services to households living in hamlets (tandas) away from the main villages and where these communities reside, and hiring VNVs from ST/SC communities. As a result, around 50% of VNVs and beneficiaries were from the ST/SC communities. 43. Financial Management: During appraisal, the FM risk rating for the grant was rated as Low. During implementation, the task team conducted routine supervision on FM aspects. Some capacity issues were identified, such as the FM staff having no background in accounting, and delays in submitting audit reports. The FM specialist made detailed recommendations and conducted extensive capacity building activities to address the identified issues. As a result, the FM staff in KSRLPS was able to maintain separate books in Tally accounting software system for the project, submit IUFR claims within a reasonable time limit, and manage the FM system as per World Bank guidelines. 44. Procurement: During appraisal, the initial procurement risk rating was rated as Substantial, but after mitigation measures were implemented, it was rated as Moderate. The primary risk was the implementing agency’s limited experience with procurement under World Bank guidelines. A procurement officer was designated and trained. Bank’s procedures for procurement was implemented. However, there were Page 18 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) significant delays in procurement by KSRLPS, particularly of 1) the implementing NGO, 2) the concurrent monitoring agency for the impact evaluation (which was later dropped), which led to some delays in Project implementation. D. Risks to development outcome 45. Overall, the risks to development outcomes is estimated to be low. The project has significantly improved: 1) awareness of nutrition-related knowledge, social programs for nutrition, health and hygiene, as well as good nutrition practices, and 2) the capacity of the VNVs, SHG members, and other community health and social workers in providing nutrition knowledge and delivering services that can impact nutrition outcomes. These are likely to last beyond project completion. 46. The project has had some success in sustaining its activities beyond completion and there is evidence of possible interest in scaling up these efforts in other regions the State. The two EDF production units that were initially closed soon after project closing have restarted production of nutrition supplements in February 2019 with funding from the Government of Karnataka. The project team completed an innovative mini-study to assess the feasibility of commercializing the EDF production unit and market the nutrition supplements produced and identified the key factors that affected consumer purchasing behavior. There are also commitments from the Government on scaling up the project in other regions in the State. V. LESSONS LEARNED AND RECOMMENDATIONS A. Key lessons learned during grant implementation 47. The project showed that a comprehensive approach is required to address the issue of malnutrition. Improving nutrition is a complex task involving changing various behaviors (including nutritional, health, sanitation, and other practices) among different population groups (such as pregnant women, mothers, young children, and adolescent girls) at different levels (household and community levels). There were also many awareness and access barriers. It is therefore important to fill the knowledge gaps about good practices that have an impact on nutrition outcomes and then proceed to bridge utilization gaps by providing food supplements and improving access to nutrition-related (specific and sensitive) services in other sectors through government programs (such as immunization, sanitation and safe drinking water). 48. The pilot project demonstrated a life cycle approach to nutrition, that is, focusing on adolescent girls, pregnant and lactating women and children. This should be highlighted as most nutrition programs focus on pregnant women and children but miss out on a key group - adolescent girls. 49. Village Nutrition Volunteers (VNVs) hired by the project played an instrumental role in motivating and assisting beneficiaries to improve health and dietary practices. The project hired 447 VNVs (all women) who were trained and empowered so they could be active participants in community engagement, and in bringing about behavior change. On the other hand, the social standing and attitude to life of VNVs were also changed through their involvement in the project. 50. It is critical to establish a multi-stakeholder multi-sectoral convergence mechanism to improve Page 19 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) nutrition outcomes. (1) Within the health and nutrition sector, coordination mechanism between several key community players VNVs, anganwadi (rural health center) workers, Accredited Social Health Activist (ASHA), medical officers, and auxiliary nurse midwife (ANM) were established at the grassroots level to promote good team work between the various functionaries. An anemia health check-up camp was also organized in convergence with the Health Department. (2) In the sanitation sector, the project leveraged on the Swatch Bharat Abhiyan ODF, which constructed over 2,600 toilets for the project areas. (3) The project leveraged the education sector by orienting 204 high school teachers and 69 middle school teachers to bring hygiene practices to schools and ensure regular distribution of IFA tablets. (4) Local governments (gram panchayat/Taluk panchayat) members were oriented about the benefits of the project activities so that they could motivate the beneficiaries to consumed EDF and adopt good hygiene practices. 51. Innovative methodology for improved beneficiary identification (see Annex 3) enabled better and probably more accurate identification of the poorest people in the blocks that could be targeted by pilot interventions. As project blocks were the most backward blocks in Karnataka, the proportion of poor households in need of nutrition services might be higher than the conventional estimate of 40%. The project adopted a systematic approach for more accurate beneficiary identification. Project used data from the 2011 Socio Economic and Caste Census (SECC) and applied fourteen socio-economic parameters (automatic inclusion criteria included households without shelter, primitive tribal groups, households whose main occupation was scavenging and legally released bonded laborers; and 8 sets of deprivation indicators were used to rank the remaining households) to shortlist target households from this dataset. This beneficiary identification exercise found 70% of residents in project areas were poor, far above the previous estimated 40%. Following a data validation process, which included a household survey, 29,786 households (with 37,781 beneficiaries) were finally selected under the pilot project. B. Recommendations 52. Capacity building on project management in the early phase of the project was crucial to ensure successful project implementation. These include capacity building on procurement, financial management, data collection, and supervision. 53. The project and other similar projects would benefit from robust Monitoring and Evaluation (including impact evaluation) to better assess performance and achievements. However, robust M&E plans sometimes face implementation challenges related to procurement, data, and administrative processes. Therefore, it is . important to identify these challenges early on and develop strategies to tackle them. Page 20 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 1. Percentage of targeted Percentage 0.00 80.00 80.00 95.00 under-three children, adolescent girls and pregnant 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 and nursing mothers who receive nutritious supplementary foods produced and supplied by the project Comments (achievements against targets): Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 2. Percentage of targeted Percentage 0.00 60.00 60.00 89.00 Page 21 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) households who utilized social 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 sector programs with a potential impact on nutrition (specifically ICDS, health services, and water and sanitation services) Comments (achievements against targets): Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 3. Percentage of targeted Percentage 0.00 60.00 60.00 100.00 pregnant and lactating women who practice core child 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 nutrition and health care behaviors Comments (achievements against targets): A.2 Intermediate Results Indicators Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 4. Percentage of targeted Percentage 0.00 90.00 90.00 100.00 Page 22 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) households receiving 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 counselling on improved child care and feeding behaviors Comments (achievements against targets): Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 5. Percentage of target Percentage 0.00 90.00 90.00 95.00 beneficiaries weighed monthly by nutrition volunteers 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 Comments (achievements against targets): Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 6. Percentage of targeted Percentage 0.00 90.00 90.00 100.00 households who receive information about available 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 social programs from nutrition volunteers Comments (achievements against targets): Page 23 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion 7. Percentage of planned Percentage 0.00 90.00 90.00 107.00 nutrition volunteers in place 27-Jul-2015 31-Jul-2017 31-Jul-2018 30-Mar-2018 Comments (achievements against targets): Page 24 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) B. ORGANIZATION OF THE ASSESSMENT OF THE PDO Objective/Outcome 1 1. Percentage of targeted under-three children, adolescent girls and pregnant and nursing mothers who receive nutritious supplementary foods produced and supplied by the project 2. Percentage of targeted households who utilized social sector Outcome Indicators programs with a potential impact on nutrition (specifically ICDS, health services, and water and sanitation services) 3. Percentage of targeted pregnant and lactating women who practice core child nutrition and health care behaviors 1. Percentage of targeted households receiving counselling on improved child care and feeding behaviors 2. Percentage of target beneficiaries weighed monthly by nutrition Intermediate Results Indicators volunteers 3. Percentage of targeted households who receive information about available social programs from nutrition volunteers 4. Percentage of planned nutrition volunteers in place Component 1: Increase consumption of nutritious foods and improve household nutrition-related knowledge and behaviors. • 32,882 of out of the 34,431 targeted beneficiaries (95%) had received nutritious supplementary foods produced supplied Key Outputs by Component by the project’s production units (linked to the achievement of the Objective/Outcome 1) • 100% (4,379 out of 4,379) of targeted pregnant and lactating women reported practicing all 6 core child nutrition and health care behaviors. • All 27,484 (100%) of households received counseling from VNVs on improving child care and feeding behaviors. Page 25 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) • By March 2018, 95% (32,882 out of 34,431) targeted beneficiaries were weighted monthly by VNVs. • Component 2: Improve access to multi-sectoral interventions with an impact on nutrition. • 100% (27,484 out of 27,484) households received information from VNVs on the availability of different schemes and how to access them. • 24,731 of 27,484 (89%) targeted households utilized one or more social sector programs with a potential impact on nutrition (specifically ICDS, health services, and water and sanitation services). 1,210 toilets have been constructed in Chincholi block and 1,424 in Devadurga block taking benefit of the Swatch Bharat Mission. Component 3: Project management and Monitoring and Evaluation. 418 Village Nutrition Volunteers (VNVs) were planned for the project, and 447 had been recruited in all 421 target villages in Chincholi and Devdurga Blocks. The volunteers were trained, and carried out various important functions of the Project. • Baseline survey conducted Impact evaluation survey conducted Page 26 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) . ANNEX 2. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Components Percentage of Approval (US$M) Closing (US$M) Total 4.55 3.48 77% 1. Increase consumption of nutritious foods and improve 3.51 3.16 90% household nutrition-related knowledge and behaviors 2. Improve access to multi- sectoral interventions with an 0.15 0.12 84% impact on nutrition 3. Project management, 0.90 0.19 22% monitoring and evaluation Page 27 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) ANNEX 3. INNOVATIVE BENEFICIARY IDENTIFICATION METHODOLOGY A key activity prior to program roll-out was the identification of beneficiaries to be targeted by KMNP. The original proposal consisted of rolling-out the intervention to 40% of the poorest households that live below poverty line. However, stakeholder consultations highlighted two limitations of this approach. First, this approach may underestimate the number of households in need of nutrition services as the two project blocks were the most backwards blocks in the state, and prevalence of poverty and malnutrition might be higher than 40%. Second, the project needed to target children, adolescent girls and pregnant and lactating women who were malnourished. Individuals from “non-poor� households might become malnourished if unable to afford additional nutrition to meet their needs. For example, women in households without a BPL card might become pregnant and malnourished due to greater nutritional needs during pregnancy. Thus, such women were at-risk of poor health outcomes even though they were from “non-poor� households. There was no simple way to do a nutrition ranking within the below poverty population. Therefore, a more systematic beneficiary identification methodology was used. The team used data from the 2011 Socio-Economic and Caste Census (SECC), commissioned by the Ministry of Rural Development, Government of India, to try and identify individuals who were at risk. The SECC survey collected data at the individual and household levels on the following: occupation, education, disability, religion, SC/ST status, name of caste/tribe, employment, income and source of income, asset ownership, housing, consumer durables and non-durables and ownership of land. Error! Not a valid bookmark self-reference. presents a list of criteria that was used to identify a set of individuals who were eligible for inclusion as beneficiaries for the scheme. Table 1: Criteria for Selecting Beneficiaries Sl. No Auto-Include Criteria Deprivation indicators used to Rank households 1 Households without shelter Only one room with kutcha walls and kutcha roof Households where main 2 occupation is manual scavenging No adult member between age 16 to 59 years Female headed households with no adult male 3 Primitive Tribal groups member between age 16 to 59 Disabled member and no able-bodied adult 4 Legally released bonded laborers member 5 SC/ST households 6 No literate adult above 25 years Landless households deriving major part of their 7 income from manual casual labour Monthly income of highest earning household 8 member, less than 1000 per month Source: Note circulated by KHPT Page 28 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) The algorithm to identify beneficiaries began with the SECC data for the two talukas and automatically selected all households that meet any of the auto-include criteria listed in Column 2 in The team used data from the 2011 Socio-Economic and Caste Census (SECC), commissioned by the Ministry of Rural Development, Government of India, to try and identify individuals who were at risk. The SECC survey collected data at the individual and household levels on the following: occupation, education, disability, religion, SC/ST status, name of caste/tribe, employment, income and source of income, asset ownership, housing, consumer durables and non-durables and ownership of land. Error! Not a valid bookmark self- reference. presents a list of criteria that was used to identify a set of individuals who were eligible for inclusion as beneficiaries for the scheme. Table 1. The set of eight criterion in Column 2 of The team used data from the 2011 Socio-Economic and Caste Census (SECC), commissioned by the Ministry of Rural Development, Government of India, to try and identify individuals who were at risk. The SECC survey collected data at the individual and household levels on the following: occupation, education, disability, religion, SC/ST status, name of caste/tribe, employment, income and source of income, asset ownership, housing, consumer durables and non- durables and ownership of land. Error! Not a valid bookmark self-reference. presents a list of criteria that was used to identify a set of individuals who were eligible for inclusion as beneficiaries for the scheme. Table 1 were used to give each household an ordinal rank based on a principal component analysis. Households were subsequently classified into terciles and households in the lowest tercile with an average income below Rs. 10,000 per month were given the highest priority for inclusion into the beneficiary list. The final eligible list was generated by placing indicative caps for each beneficiary group, that were calculated as the estimated number of children below the age of 36 months, adolescent girls in the 11 – 18 years of age, pregnant and lactating women to be expected in 40% of BPL households. The eligible list was further updated through field visits and explicit selection of beneficiaries by the VNV (along with their supervisors) to verify the list from the SECC analysis. The list was updated to reflect absence (people listed in the SECC exercises who had migrated, or were not identifiable based on SECC data, or had assets not listed in the SECC data) and inclusions (identifiably poor households such as those who are homeless i.e. meeting auto-inclusion criteria in The team used data from the 2011 Socio- Economic and Caste Census (SECC), commissioned by the Ministry of Rural Development, Government of India, to try and identify individuals who were at risk. The SECC survey collected data at the individual and household levels on the following: occupation, education, disability, religion, SC/ST status, name of caste/tribe, employment, income and source of income, asset ownership, housing, consumer durables and non-durables and ownership of land. Error! Not a valid bookmark self-reference. presents a list of criteria that was used to identify a set of individuals who were eligible for inclusion as beneficiaries for the scheme. Table 1). A summary of the final beneficiary list is presented in Table 2. The number of beneficiaries selected for inclusion was significantly larger than the number original estimated. Page 29 of 30 The World Bank Karnataka Multisectoral Nutrition Pilot (P149811) Table 2: Count of Beneficiaries identified by KHPT in pilot areas Settlement Adolescent Women Taluka Type # Children Girls Pregnant Lactating Total Chincholi Tandas 85 525 476 61 87 1149 Villages 135 3738 6351 623 1562 12274 Sub-Total 220 4263 6827 684 1649 13423 Devadurga Doddis 58 26 130 13 17 186 Tandas 62 214 197 16 66 493 Villages 170 3974 4808 528 1618 10928 Sub-Total 290 4214 5135 557 1701 11607 TOTAL 510 8477 11962 1241 3350 25030 Note: Tandas are “banjara� settlements i.e. settlements for a specific migratory castes/tribes. Doddis are small human settlements outside the main village boundaries and are not migratory and typically have a different caste/tribe composition than Tandas. Page 30 of 30