Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004655 IMPLEMENTATION COMPLETION AND RESULTS REPORT CREDIT NUMBER 5283-IN ON A CREDIT IN THE AMOUNT OF SDR 66.1 MILLION (US$100 MILLION EQUIVALENT) TO THE REPUBLIC OF INDIA FOR A LOW INCOME HOUSING FINANCE PROJECT September 19, 2019 Finance, Competitiveness and Innovation Global Practice South Asia Region CURRENCY EQUIVALENTS (Exchange Rate Effective {July 12, 2019}) Currency Unit = Indian Rupees (INR) US$1.00 = INR 68.56 FISCAL YEAR April 1 - March 31 Regional Vice President: Hartwig Schafer Country Director: Junaid Kamal Ahmad Regional Director: Zoubida Kherous Allaoua Practice Manager: Esperanza Lasagabaster Task Team Leader(s): Anuradha Ray ICR Main Contributor(s): Yunfan Gu, Varsha Marathe Dayal ABBREVIATIONS AND ACRONYMS CLSS Credit Linked Subsidy Scheme DFS Department of Financial Services EWS Economically Weaker Section GoI Government of India GRIDS Grievance Registration and Information Database System HFC Housing Finance Company ICR Implementation Completion and Results Report IDA International Development Association IFC International Finance Corporation INR Indian Rupee IUFR Interim Unaudited Financial Reports LIG Low Income Group M&E Monitoring and Evaluation MFD Mobilizing Finance for Development NHB National Housing Bank NPL Nonperforming Loan PAD Project Appraisal Document PDO Project Development Objective PIU Project Implementation Unit PLI Primary Lending Institution PSB Public Sector Bank QPLI Qualified Primary Lending Institution SDR Special Drawing Rights SEDD Social and Environment Due Diligence US$ United States Dollar WB World Bank WBG World Bank Group TABLE OF CONTENTS DATA SHEET .......................................................................................................................... 1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 5 A. CONTEXT AT APPRAISAL .........................................................................................................5 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) .......................................9 II. OUTCOME .................................................................................................................... 11 A. RELEVANCE OF PDO ..............................................................................................................11 B. ACHIEVEMENT OF PDO (EFFICACY) ........................................................................................11 C. EFFICIENCY ...........................................................................................................................19 D. JUSTIFICATION OF OVERALL OUTCOME RATING ....................................................................21 E. OTHER OUTCOMES AND IMPACTS (IF ANY)............................................................................21 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 22 A. KEY FACTORS DURING PREPARATION ...................................................................................22 B. KEY FACTORS DURING IMPLEMENTATION .............................................................................22 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 23 A. QUALITY OF MONITORING AND EVALUATION (M&E) ............................................................23 B. ENVIRONMENTAL, SOCIAL AND FIDUCIARY COMPLIANCE ......................................................24 C. BANK PERFORMANCE ...........................................................................................................25 D. RISK TO DEVELOPMENT OUTCOME .......................................................................................28 V. LESSONS AND RECOMMENDATIONS ............................................................................. 28 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 30 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ......................... 37 ANNEX 3. PROJECT COST BY COMPONENT ........................................................................... 40 ANNEX 4. EFFICIENCY ANALYSIS ........................................................................................... 41 ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS ... 44 ANNEX 6. SUPPORTING DOCUMENTS .................................................................................. 45 The World Bank India Low-Income Housing Finance (P119039) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P119039 India Low-Income Housing Finance Country Financing Instrument India Investment Project Financing Original EA Category Revised EA Category Financial Intermediary Assessment (F) Organizations Borrower Implementing Agency National Housing Bank, National Housing Bank, Republic of India Department of Financial Services, Ministry of Finance Project Development Objective (PDO) Original PDO To provide access to sustainable housing finance for low income households, to purchase, build or upgrade their dwellings. Page 1 of 45 The World Bank India Low-Income Housing Finance (P119039) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing 100,000,000 100,000,000 93,050,362 IDA-52830 Total 100,000,000 100,000,000 93,050,362 Non-World Bank Financing 0 0 0 Borrower/Recipient 0 0 0 Total 0 0 0 Total Project Cost 100,000,000 100,000,000 93,050,362 KEY DATES Approval Effectiveness MTR Review Original Closing Actual Closing 14-May-2013 20-Nov-2013 09-Dec-2015 31-Dec-2018 31-Dec-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 03-Dec-2018 93.05 Change in Components and Cost Reallocation between Disbursement Categories KEY RATINGS Outcome Bank Performance M&E Quality Satisfactory Satisfactory Modest RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 11-Oct-2013 Satisfactory Satisfactory 0 02 14-Apr-2014 Satisfactory Moderately Satisfactory 4.95 Page 2 of 45 The World Bank India Low-Income Housing Finance (P119039) 03 08-Nov-2014 Moderately Satisfactory Moderately Satisfactory 8.07 04 27-May-2015 Satisfactory Moderately Satisfactory 10.55 05 14-Jul-2015 Satisfactory Moderately Unsatisfactory 10.55 06 09-Feb-2016 Moderately Satisfactory Moderately Unsatisfactory 10.55 07 12-Apr-2016 Moderately Satisfactory Moderately Satisfactory 32.87 08 11-Nov-2016 Moderately Satisfactory Moderately Satisfactory 55.79 09 23-May-2017 Moderately Satisfactory Moderately Satisfactory 55.79 10 02-Dec-2017 Moderately Satisfactory Moderately Satisfactory 87.27 11 29-May-2018 Moderately Satisfactory Moderately Satisfactory 89.05 12 27-Dec-2018 Satisfactory Satisfactory 93.05 SECTORS AND THEMES Sectors Major Sector/Sector (%) Public Administration 3 Central Government (Central Agencies) 3 Financial Sector 97 Banking Institutions 48 Other Non-bank Financial Institutions 49 Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Finance 65 Financial Stability 65 Financial Sector oversight and policy/banking 65 regulation & restructuring Public Sector Management 2 Public Administration 2 Transparency, Accountability and Good 2 Governance Page 3 of 45 The World Bank India Low-Income Housing Finance (P119039) Urban and Rural Development 33 Urban Development 33 Services and Housing for the Poor 33 ADM STAFF Role At Approval At ICR Regional Vice President: Isabel M. Guerrero Hartwig Schafer Country Director: Onno Ruhl Junaid Kamal Ahmad Director: Sujata Nitin Lamba Zoubida Kherous Allaoua Practice Manager: Ivan Rossignol Esperanza Lasagabaster Task Team Leader(s): Michael Markels III Anuradha Ray ICR Contributing Author: Yunfan Gu Page 4 of 45 The World Bank India Low-Income Housing Finance (P119039) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL Context 1. At project appraisal, India was experiencing rapid urbanization and mass migration of its population to cities. Between 2001 and 2011, the urban population in India had increased by 91 million to 377 million urban dwellers. 1 The number of metropolitan cities with a population of one million and above had risen from 35 in 2001 to 50 in 2011. With fast growth in sight, the urban population was projected to reach 600 million by 2031. 2 2. Severe housing shortage existed, especially for the low-income households, many of which earned their income in the informal sector. It was estimated that a third of urban dwellers lived in slums. The urban housing backlog was estimated at 19 million in 2012. 3 More than 90 percent of the housing shortage was borne by low-income and poor families working in the informal sector and/or living in dwellings without formal property titles, as “decent” housing in the formal sector was beyond the reach of the vast majority of these households. 4 3. Shortage of housing for low-income households was the result of both demand and supply side factors, and lack of access to affordable housing finance was a key constraint. On the demand side, many low-income families could have afforded simple, “decent” units if they had access to finance under reasonable terms. 5 However, the low-income segment of the population was not a commercial target for mainstream financial institutions. 6 The lack of housing finance reduced the effective demand and disincentivized developers from building affordable housing to scale. On the supply side, constraints including a dearth of land serviced by utilities and a lack of formal property rights artificially restricted the availability of housing. 4. Many low-income households were excluded from the formal housing finance market due to the informality of their income and dwellings. The households who worked in the informal sector and earned informal income needed customized financing instruments tailored to their circumstances, but there was a lack of capacity to design or implement such instruments in housing finance institutions. In 1 Census of India, 2011 2 The High Powered Expert Committee for Estimating the Investment Requirements for Urban Infrastructure Services, Report on India Urban Infrastructure and Services, March 2011. 3 Ministry of Housing and Urban Affairs, “Report of the Technical Group on Urban Housing Shortage (TG-12) (2012-2017)”, September 22, 2012. 4 Decent housing here is defined as one that meets the basic requirements per United Nations-Habitat guidelines in terms of structural quality and access to basic services (water, sanitation, electricity). 5 “Expanding the Housing Finance Market to Cover Lower Middle Income Segments in India”, prepared by Monitor Group for the FIRST initiative and the World Bank, 2007. 6 Households with annual household income between INR 60,000 to INR 200,000 represented one third of the total urban population but accounted for just three percent of formal housing finance lending in 2012. See project appraisal document. Page 5 of 45 The World Bank India Low-Income Housing Finance (P119039) addition, as in many other low and middle-income countries, the majority of households in low-income neighborhoods in India generally built and occupied houses before being granted property title and formal approval of development plans and these houses could not be used as collateral. Due to the lack of formal sources of financing, these low-income households frequently took out short-term loans from informal sources such as money lenders, which had very high costs. 7 5. The market failure addressed by this project was the exclusion of low-income households from the housing finance market caused by the informality of their income 8 and dwellings. 9 The project aimed to provide access to sustainable housing finance for low-income households with annual incomes, including informal incomes, below INR 300,000, 10 and/or living in informal property. Its objective was to implement market mechanisms to effectively leverage the expertise of private sector financial institutions to deliver affordable housing finance to informal and low-income households. 6. The project supported the Government of India’s (GoI’s) strategy to foster inclusive urban development, and was well aligned with India’s Country Partnership Strategy (CPS) at appraisal. Two of the three strategic engagement areas of the CPS (FY13-17) were Transformation and Inclusion, consistent with the GoI’s objectives in its twelfth Five-Year Plan (2012-17). The project supported the Government’s efforts in both areas by addressing the affordable housing problem brought about by rapid urbanization and supporting financial inclusion for informal and low-income households. 7. The National Housing Bank (NHB), the project counterpart, is the apex institution for housing finance in India. The NHB functions as a liquidity facility providing short-and long-term refinancing to primary lending institutions (PLIs), including banks, housing finance companies (HFCs) and other institutions that provide housing finance. 11 In addition, the NHB also regulates and supervises HFCs. As a development finance institution, the NHB has a public policy objective to develop long-term housing finance in India and deepen financial markets. Accordingly, it has played a catalytic role in developing India’s mid-market mortgage finance industry. At project appraisal, the NHB had evolved to refocus on serving the housing finance needs of low-income households. 8. The project was designed as a pilot (US$100 million) to test the market frontier in lending to informal and low-income households. The project design was based on solid analytical work and 7 Informal money lender is an integral part of the social network and economy in the community of informal and low-income households. Moneylenders frequently lend at high interest rate reaching 5-7 percent interest rate per month, which comes to over 60 percent per year. 8 “Informal income” for the purpose of grant of housing loan under this project refers to income derived from low income economic activities and that meets one of the following criteria: 1) self-employment in any low-income business, profession or occupation; 2) casual, temporary, irregular or multiple jobs; and/or 3) employment in the unorganized sector. 9 “Informal dwelling/housing” is characterized by lack of records and/or contravention of municipal development plans and land regulations. The violations imply that the Government is legally empowered to evict slum dwellers. Under Indian law, title and evidence of title is made up of some of the following elements: 1) land survey record of ownership, 2) record of right or pattadar pass book issued by the revenue department to land owners on the basis of revenue record, 3) document of allotment of land, 4) registered sale deed, and 5) document of inheritance. 10 The initial income threshold was INR 200,000 at appraisal stage, and later revised up to INR 300,000 to reflect inflation and wage growth in the years after project appraisal. The targeted households make up about 70 percent of the urban population. Compared with households with annual income above INR 200,000 at appraisal stage, the households with income below INR200,000 are half likely to have salaried employment or business as the primary source of their income. (Data from India Human Development Survey) 11 The PLIs include the HFCs, urban cooperative banks, state-owned commercial banks, regional rural banks and at the other end of the spectrum, microfinance institutions and nongovernmental organizations (NGOs). Page 6 of 45 The World Bank India Low-Income Housing Finance (P119039) extensive market consultation. Before the project was in place, the financial institutions serving the segment lacked capacity, and lending to the segment was generally perceived to be high risk and remained untested in India and other low-income markets. Theory of Change (Results Chain) 12 9. The project focused on addressing the market failure caused by informality using a market- based solution, with the Mobilizing Finance for Development (MFD) principle at its core. The project’s lending was highly targeted towards informal and low-income households and was delivered using a market-based mechanism through the PLIs. Leveraging the NHB’s apex role in housing finance market, the project also strongly supported the development of necessary market infrastructure to enable private sector financial institutions to expand lending to low-income households against informal incomes and informal property titles. The focus on a market-based solution and collaboration with private sector PLIs ensured the sustainability of the project’s outcomes and made the project an MFD operation at its core. Figure 1: Theory of Change 12 There was no requirement for a theory of change at the project design stage. The theory of change is constructed from the project appraisal document for this Implementation Completion and Results Report (ICR). Page 7 of 45 The World Bank India Low-Income Housing Finance (P119039) Project Development Objective (PDO) The Project Development Objective (PDO) was to provide access to sustainable housing finance for low- income households, to purchase, build, or upgrade their dwellings. Key Expected Outcomes and Outcome Indicators The Key Expected Outcomes were: • Improved access to sustainable housing finance for informal and low-income urban households. • Development of a sustainable housing finance market for informal and low-income urban households. • Improved capability of informal and low-income urban households to purchase, build, or upgrade their dwellings. Key Outcome Indicators were: • The increased number of primary lenders active in the Economically Weaker Section (EWS) and Low- Income Group (LIG) segments. 13 • The increased volume of lending in the EWS and LIG segments by primary lenders which are refinanced by the NHB. • The annual percentage increase in the number of borrowers in the EWS and the LIG income segments receiving loans from primary lenders which are refinanced by the NHB. • The percentage of replacements of non-performing with performing sub-loans by Qualified Primary Lending Institutions (QPLIs) required for pools of loans which have been refinanced by the NHB. 14 Components The project comprised the following components funded by the International Development Association (IDA): • Component 1 (Special Drawing Rights (SDR) 1.32 million / US$2 million at appraisal) – Capacity Building. Under this component, activities were planned to strengthen the capacity of the NHB, Qualified Intermediary Institutions, and QPLIs. The component envisaged the design and roll-out of pilots with state/municipal agencies, experts and practitioners and selected PLIs. The aim of these pilots and capacity building exercises was to develop new financial products, loan standards, risk management tools, and financial literacy and consumer protection capacity. This component also included the roll-out of a final impact evaluation to assess the social and household-level impact of the project. 13 EWS: less than INR 100,000 household income per annum at project appraisal stage. LIG: less than INR 200,000 household income per annum at project appraisal stage. The official income threshold for EWS and LIG segments were adjusted up multiple times during project implementation. The NHB eventually used loans with individual loan size below INR 1,500,000 as lending towards EWS and LIG segments. Indicators using lower loan size thresholds were also collected by the task team to complement these PDO indicators. 14 This indicator is a proxy for nonperforming loan (NPL) ratio. The NHB takes the corporate credit exposure on the PLI and the PLI bears the risk on the retail mortgages. The PLIs provide part of their mortgage loan portfolio as collateral for the NHB loan. Should any of the loans provided as collateral become delinquent, they are replaced with a performing loan. This ensures that the NHB minimizes its direct credit risk to the mortgage market. Page 8 of 45 The World Bank India Low-Income Housing Finance (P119039) • Component 2 (SDR 64.12 million / US$97 million at appraisal) – Financial Support for Sustainable and Affordable Housing (Refinancing). This component was designed to support the NHB to refinance, directly or indirectly through Qualified Intermediary Institutions, low-income housing loans made by the PLIs to Primary Borrowers to purchase, build or upgrade their dwelling. • Component 3 (SDR 0.66 million / US$1 million at appraisal) – Project Implementation. This component included the establishment of a Project Implementation Unit (PIU) within the NHB to help implement the project’s fiduciary responsibilities, carry out monitoring and evaluation (M&E), be responsible for legal issues and grievance redressal, oversee and monitor social and environmental due diligence, and carry out any procurement planned under the project. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) Revised Components 10. A level 2 project restructuring was implemented in December 2018 to reallocate project funds. The restructuring reallocated SDR 1.97 million (SDR 1.31 million from Component 1 and SDR 0.66 million from Component 3) to the refinancing line (Component 2). This essentially amounted to an allocation of the entire IDA financing of SDR 66.10 million to Component 2 of the project, which originally was envisaged to be SDR 64.12 million. The financing for Components 1 and 3 was supported through counterpart financing by the NHB. Table 1: Disbursement by Component Components Planned Planned Disbursement Amount Actual Final Disbursement after Restructuring in 2018 Disbursement Amount Amount (SDR million) (SDR million) (SDR million) Component 1 1.32 0 0 Component 2 64.12 66.1 66.1 Component 3 0.66 0 0 Total 66.1 66.1 66.1 Rationale for Changes and Implications for the Original Theory of Change 11. The NHB preferred to continue with the momentum on disbursements to informal and low- income segments using project funds, and use counterpart funds for the remaining activities under Components 1 and 3. The NHB implemented several key capacity building activities with counterpart funds. The NHB also continued broader capacity building initiatives, including on strengthening credit infrastructure and consumer protection frameworks at a sectoral level, given its development mandate. Capacity building of the PLIs on topics such as good risk management practices and the social and environmental due diligence (SEDD) framework were periodically conducted by the NHB. Audits conducted by the NHB were largely satisfactory, with the SEDD being mainstreamed into overall credit Page 9 of 45 The World Bank India Low-Income Housing Finance (P119039) processes by the PLIs. These audits were used, among other things, to train and sensitize PLI staff on environmental and social safeguard parameters, including grievance redressal mechanisms. These capacity building activities were also used to address relevant issues raised by beneficiaries and the PLIs. At the same time, certain activities planned under Component 1 (capacity building) and Component 3 (M&E), including the project impact evaluation and a third-party audit of the SEDD framework, were delayed. Although the NHB remained committed to carrying out these activities, there were delays at key decision points due to internal management changes and NHB’s strong preference to use limited internal resources for these activities. 12. The GoI was supportive of the NHB’s proposal to reallocate the amounts from Components 1 and 3 to the refinancing component provided that the NHB pursued key M&E and capacity building activities as per project design. At the Tripartite Portfolio Review meeting of the World Bank’s lending portfolio in January 2018, the GoI advised the NHB to proactively consider closure of project activities before end-date of December 2018 in the context of its good disbursement performance. 13. The restructuring did not affect the initial Theory of Change. The PDO remained unaltered throughout the project life. The IDA credit line targeting informal and low-income households was fully disbursed. While not funded by the IDA credit, key capacity building activities were carried out by the NHB and the PLIs using counterpart funding, and these agencies made significant progress in their capacity to serve informal and low-income households (see also para. 19). Page 10 of 45 The World Bank India Low-Income Housing Finance (P119039) II. OUTCOME A. RELEVANCE OF PDO Assessment of Relevance of PDO and Rating 14. The PDO was highly relevant at appraisal and remains highly relevant in the current country context of India. India’s cities will be home to 600 million people by 2031, more than twice the level observed in 2001. 15 Amid this urbanization trend, informal and low-income households face a large housing shortage, and affordable housing has been among the GoI’s top priorities. In 2015, the GoI launched the Housing for All initiative to provide affordable housing for low-income households (see para. 23). This project strongly supported the GoI’s affordable housing agenda by directly addressing the market failure caused by informality, and aimed to create a functioning housing finance market for informal and low-income households through an MFD approach. 15. The PDO remains well-aligned with the World Bank’s key focus areas in the current Country Partnership Framework (CPF FY18-22). By providing access to sustainable housing finance for the informal and low-income households, the project directly contributed to the first focus area of the CPF, “Promoting Resource-efficient Growth,” under which a key objective is to “improve livability and sustainability of cities” through new engagement on affordable housing. The project also directly contributed to the second focus area of the CPF, “Enhancing Competitiveness and Enabling Job Creation,” under which a key objective is to “increase resilience of the financial sector and financial inclusion” by supporting the development of credit products and providing financing to facilitate long-term funding to the affordable housing sector. The project was also well aligned with the CPS (FY13-17) at appraisal (para. 6). 16. The relevance of the PDO is rated High. As discussed above, the project is highly relevant in the development context of India and remains well aligned with focus areas in the India CPF (FY18-22). B. ACHIEVEMENT OF PDO (EFFICACY) Assessment of Achievement of Each Objective/Outcome 17. The project focused on addressing the market failure caused by informality, and strongly supported the development of a sustainable housing finance market for informal and low-income households using an MFD approach. The project’s achievements towards development outcomes will be discussed in this Section. To sum up, the NHB set up an innovative Special Refinance Scheme that was dedicated to disbursing the project’s financing to the targeted informal and low-income households. The NHB also carried out capacity building for the PLIs, including market-based pilots, as well as regulatory reforms, including institutionalization of the SEDD framework to ensure building safety, access to basic 15 India Systematic Country Diagnostic (2018) Page 11 of 45 The World Bank India Low-Income Housing Finance (P119039) hygiene facilities and consumer protection. The stable funding provided by the Special Refinance Scheme, together with an enabling environment created by the NHB, incentivized the PLIs (mostly HFCs) to invest in new lending tools that catered to informal and low-income households, such as field-based income verification. More than 90 percent of the loans financed by the IDA credit line reached low-income households that earn informal income and/or live in houses with informal property titles. Led by the NHB as India’s apex housing finance institution, the project’s financing was disbursed to 15,541 households across 17 states through 17 HFCs, creating a substantial demonstration effect in the wider economy and further mobilizing financing to the informal and low-income households. Outcome: Improved Access to Sustainable Housing Finance for Informal and Low-income Urban Households 18. The project provided affordable and long-term housing finance to informal and low-income households. • The IDA credit line was effective in reaching underserved informal and low-income households. The IDA credit line reached 15,541 informal and low-income households across 17 states in India. Forty- five percent of the IDA credit was delivered to low-income households with informal property titles, and 95 percent of total loans were delivered to low-income households with informal income (Table 2). Due to the focus on the informal segment, the average annual household income of the end- beneficiaries was INR 196,700 (US$2,900), well below the project’s targeted income threshold of INR 300,000. 16 The average loan size was also small, at INR 435,200 (US$6,400). Table 2: Disbursement Matrix by Quadrants Particulars Formal Dwelling Informal Dwelling Formal Income 5% 0% Informal Income 50% 45% Note: Most households living in informal dwellings (right column in the table) have informal income. • The cost of the housing loans was reduced and the tenor profile of the loans improved for informal and low-income households relative to financing terms received prior to the project. Before the project, the weighted average interest rate for housing loans of households below the targeted income threshold was 27 percent, as these households frequently borrowed from informal money lenders who charged high interest rates and offered mostly short-term loans. 17 In comparison, the average interest rate for disbursed loans under the IDA credit line was 15 percent. 18 The average loan tenor under the IDA credit line was also long at 128 months, or more than 10 years. 16 The data are reported by the NHB. Lending income thresholds and informality criteria are enforced by the NHB and also checked on a sample basis by the project audit. Incentives for households to under-report income is very low. In fact, households tend to over-report their income as the project financing is market-based, and higher income households could get loans at cheaper rates. 17 The 27 percent weighted average borrowing cost is for households with income below INR 200,000 at project appraisal. As discussed in Para. 5, the low-income threshold is set at INR 300,000 in the later years of this project to take into account inflation and wage growth. 18 Data on the IDA credit line is from NHB’s project implementation report. The baseline interest rate (government bond yield) remained relatively stable in India in the past decade, ensuring certain comparability between the rates before and after the project was in place. Page 12 of 45 The World Bank India Low-Income Housing Finance (P119039) • The risk of lending for informal and low-income housing was higher relative to lending for formal income and title, but the risk was well-managed. The average loan-to-value ratio (LTV) in the IDA credit line was 41 percent, which ensured a sufficient owner’s equity stake. The average loan size (INR 435,200) was just 2.2 times the average annual household income (INR 196,700), providing assurance that the housing loans did not push the households into over-indebtedness. In light of the growing portfolios of PLIs, the percentage of replacements of their non-performing with performing sub- loans—a PDO indicator and a proxy for loan delinquency ratio under the IDA credit line—fell from 1 percent (baseline) to 0.86 percent by the end of the project, significantly exceeded the end target of less than 1.5 percent. The overall gross non-performing asset ratio for the HFCs also remained stable during the project (data in Tables 3 and 4). Outcome: Development of a Sustainable Housing Finance Market for Informal and Low-income Urban Households 19. The project strongly supported the development of a sustainable housing finance market for informal and low-income urban households. • The NHB set up an innovative Special Refinance Scheme dedicated to this project and provided stable funding at market-based rates for the informal and low-income segment. The “Special Urban Housing Refinance Scheme for Low Income Households” had four key features: (a) it offered stable long-term funding targeting the informal and low-income households; (b) it imposed no cap on on- lending interest rates; 19 (c) the PLIs provided collateral against the refinance loan and borne the mortgage credit risk; and (d) the refinance interest rate offered by the scheme was at 8.50-8.95 percent, higher than the rate of other NHB refinance schemes at 6.50-7.00 percent. The first feature (stable long-term funding) incentivized the PLIs to enter into the underserved market and provided longer-term housing loans. The second feature (no interest rate cap) allowed the HFCs to better recover the cost of serving informal and low-income households, due to smaller loan sizes and the need to develop new products with innovative appraisal methods. The third feature (the PLIs bear retail mortgage risk) incentivized the PLIs to keep risk under control and avoided putting borrowers into over-indebtedness. The fourth feature (higher refinance interest rate) ensured the long-term sustainability of the scheme, as the scheme was market-based and not subsidized and reflected the higher risk in lending to the informal and low-income segment. 20 The Special Refinance Scheme’s eligible income threshold of INR 300,000 was much lower than other contemporary NHB refinance schemes such as the Urban Housing Fund, which had an income threshold of INR 600,000 and no special focus on informality. The NHB indicated that the Special Refinance Scheme would be kept open after the project closure. 21 19 Other NHB schemes include an on-lending interest rate cap of 200-275 basis points above the refinancing rate. 20 Other important innovations of the scheme include: 1) it extends refinancing not only for housing loans which are secured through collateral of the property but also to those that are alternatively secured and 2) it allows refinancing of microfinance institutions through PLIs. 21 Details on the operation of the Scheme after project closure are still being finalized. The NHB is also currently in discussion with the World Bank on a follow-up lending operation. Page 13 of 45 The World Bank India Low-Income Housing Finance (P119039) • The NHB implemented other important upstream regulatory reforms, including the SEDD framework, and conducted capacity building and market-based pilots with the PLIs. The SEDD framework was developed under this project to set specific requirements to ensure the safety and sanitation of the financed properties as well as consumer protection. 22 The SEDD framework is now institutionalized by the NHB and the PLIs, with the NHB conducting training and sample-based loan- level monitoring. 23 Lenders who had fully adopted the SEDD framework and fully integrated it as part of their lending policy reported being satisfied with the process and commented on the value of the approach in terms of protecting the value of the loan security against risks such as poor construction or flooding. The NHB also carried out other important reforms and actions that facilitated the operation of the PLIs in the informal segment, including conducting market-based pilots with the HFCs, 24 providing a definition of informal income for the first time in India, and providing capacity building activities on risk management and the SEDD framework for lending towards informal and low-income households. 25 The HFCs informed the World Bank task team that the NHB’s strong support towards lending to informal and low-income households sent a strong signal and provided the much-needed enabling environment, confidence and reassurance for them to enter into the informal and low-income segment. • The stable funding provided by the Special Refinance Scheme and upstream regulatory reforms and capacity building activities carried out by the NHB incentivized the PLIs to invest in innovative lending techniques catering to informal and low-income households. These lending techniques require large upfront investments to upgrade technology and build up the product teams, as the traditional credit appraisal technique is inadequate when applied to informal beneficiaries, given their volatile income status and lack of transaction histories. The HFCs suggested that the long-term funding of the Special Refinance Scheme and the enabling environment created by the NHB strongly supported their decision to invest and improve on these new lending tools. One lender, which made extensive use of the Special Refinance Scheme, invested an amount equal to 0.2 percent of its asset base in lending technology, a considerably higher amount than other comparable lenders, in order to have a fully scalable platform with a reduction in customer acquisition and servicing costs. 26 Field- based income verification methods were pioneered and used by the HFCs. Digital technology was 22 In the SEDD, key Environment Due Diligence include 1) avoiding high risk and hazardous location, 2) ensure the financed units have access to water supply and sanitation, 3) avoiding risky household activities such as mass storage of fuels and hazardous waste management facilities, 4) ensure safety on type of construction through mandatory compliance with safety standards and building regulations stipulated by local authorities and technical evaluation of structural safety by qualified personnel. Key Social Due Diligence Requirements include: 1) No discrimination, 2) Encourage co-borrowing, 3) Ensure ability to pay and avoid over indebtedness, 4) Protection of borrowers from third parties, 5) Facilitation of borrowers and 6) Security of Tenure. The average loan size is on average just 2.2 times of annual household income, indicating that the housing loans do not put the households into situation of over indebtedness. 23 Citing Chief Executive Officer (CEO) of an HFC: "(The Special Refinance Scheme) is a wonderful scheme, and we feel that this will definitely improve the conditions under which these kinds of customers currently live especially in shanties and dilapidated structures, on rented accommodation where there is no proper hygiene." 24 Four pilots were launched under the NHB’s Special Refinance Scheme. AU Housing Finance and India Shelter piloted largely in Rajasthan against informal incomes and informal tenure. The NHB also piloted a line through Ujjivan, a microfinance institution intermediates through Tata capital, to reach low income borrowers. 25 Given NHB’s strong preference to use internal resources, the capacity building activities and the SEDD framework are financed using NHB’s own funding rather than the World Bank financing. 26 This was presented at the 8th Global Housing Finance conference in Washington DC, June 2018 by CEO of an India HFC under the Special Refinance Scheme. Page 14 of 45 The World Bank India Low-Income Housing Finance (P119039) used to store relevant identity and income data and automate the loan decision process. The World Bank task team documented the following example of new lending tools being used to serve informal households: a borrower couple were daily hawkers in the weekly street market, with daily earnings of INR 500-800. The PLI assessed the risk of the borrower by talking to the suppliers and customers of the hawker couple to understand their cash flow and set a monthly payment that would not burden the family. Further, the PLI also demonstrated flexibility when the co-owner was undergoing medical treatment at the hospital. • The project created a substantial demonstration effect in crowding in more private sector financing for low-income housing. The NHB’s engagement sent a strong market signal. The project required the NHB to expand the number of PLIs serving the housing finance needs of informal and low-income households, and also piloted on-lending from HFCs to microfinance institutions. The IDA credit line was delivered through 17 HFCs and reached households in 17 states, creating a substantial demonstration effect. AU Housing Finance and India Shelter, which piloted lending largely in Rajasthan against informal incomes and informal tenure using the Special Refinance Scheme, built a robust pipeline. In addition, HFCs such as Gruh Finance and Equitas Housing Finance, both lenders under the Special Refinance Scheme, targeted small and micro borrowers across Gujarat, Rajasthan, Madhya Pradesh, Maharashtra and other states enabling low-income households to access long-term housing finance against informal income, albeit with formal titles. As the CEO of the NHB said, “This (project) is going to be a game changer as it is going to establish a different trend in the housing industry in our country.” 27 20. With project support, the housing finance market for informal and low-income households in India developed quickly. • Low-income households in the EWS and Low-Income Groups were increasingly served by HFCs during project implementation (Tables 3 and 4). The number of active PLIs in the low-income segment more than doubled, from 25 in 2013 to 52 in 2018, and a new group of “affordable HFCs” emerged to serve informal and low-income urban households. The combined loan book of these affordable HFCs rose from INR 10 billion in 2013 to INR 270 billion in December 2017, with an average loan size of INR 930,000, facilitating the ownership of more than 230,000 affordable homes. 28 The total volume of loans made by HFCs with an individual loan size below INR 1,500,000 (US$20,000), which the NHB categorizes as loans towards the EWS and LIG segments, 29 increased from INR 268 billion in 2013 to INR 489 billion in 2018, showing a compounded annual growth rate of 13 percent. • Lending growth was fastest in small-size loans, which were more likely to be taken out by informal households (Table 4). The total volume of housing loans made by HFCs with individual loan size below INR 500,000 (US$7,000) rose from INR 25 billion in 2013 to INR 56 billion in 2018, showing a 27 This is recorded in the video made by the World Bank to disseminate the experience of the project. 28 “State of the Low-income Housing Finance Market”, 2018. Data is based on 26 Affordable Housing Finance Companies. 29 Volume of loans with sizes below INR 1,500,000 is used by the NHB as a proxy for lending towards informal and low-income households. However, the loans will also include lending to formal households and households earning annual income higher than the project threshold of INR 300,000. Alternative indicators using different thresholds are presented in Table 4. Page 15 of 45 The World Bank India Low-Income Housing Finance (P119039) compounded annual growth rate of 17 percent. 30 Very small loans below INR 300,000 (US$4,000), which were most likely to be taken out by informal and very low-income households, increased at an even faster compounded annual growth rate of 27 percent, from INR 8.8 billion in 2013 to INR 29.0 billion in 2018. In addition to the volume of loans, the number of low-income borrowers financed by the HFCs also increased quickly; the number of borrowers with individual loan size below INR 300,000 increased from 105,001 to 240,913, a compounded annual growth rate of 18 percent. 31 The total amount of housing loans towards very low-income individuals with an annual income below INR 120,000 (INR 10,000 per month), 32 who most likely have informal income and live in informal properties, increased from INR 2.6 billion in 2013 to INR 4.1 billion in 2018, showing an annual compounded growth rate of 10 percent. 33 21. The PDO results indicators did not fully capture the MFD impact of the project. 34 The project strongly supported the mobilization of private sector financing towards the EWS and LIG segments from the HFCs, which grew at a much faster pace than the NHB’s refinance towards the segments. Two of the PDO results indicators (Table 3), the annual increase in the volume of lending and the number of beneficiaries in the EWS and LIG segments by the PLIs refinanced by the NHB, fell below targets. For example, the PDO indicator on the annual increase in volume of lending (second indicator in Table 3) was INR 98.4 billion in 2018, versus an end target of INR 155.4 billion for that year. However, these targets were not achieved because lending refinanced by the NHB grew much more slowly than the overall housing finance market for the EWS and LIG segments. This picture is consistent with the NHB’s catalytic role in housing finance market development, where most of funding is mobilized from various private sources including banks and capital markets. 35 The PDO indicators are complemented by supplemental indicators (see Table 4) to better capture the project’s outcomes, especially in MFD. Notice also that these two indicators reflect the annual increases, so the outstanding volume of lending and number of borrowers refinanced by the NHB still increased over time, albeit at a slower pace than the broad housing finance market. For the other two PDO indicators, one indicator on the number of primary lenders active in the EWS and LIG segments (52) significantly exceeded the end target (35), partly reflecting the MFD impact of the project. The other indicator, the percentage of replacements of non-performing with performing sub-loans by the QPLIs, a proxy for the NPL ratio of lending to EWS and LIG segments refinanced by the NHB, fell well below the maximum targeted level and even below the 2013 baseline figure, indicating sound risk management practices. The gross NPL ratio for the HFCs also remained stable over project implementation (Table 4). 30 The same nominal threshold of INR 500,000 is used without adjusting for inflation. If the adjustment for nominal loan threshold is performed, the increase in small-ticket loans will be even faster. Similar quality data before 2012 is not available, as the NHB changes the reporting format in 2012. 31 The number of borrowers with individual loan size below INR 500,000 also increased quickly as shown in Table 4. The NHB does not report the number of borrowers for other income thresholds. 32 This threshold is reported by the NHB in “Report on Trends & Progress in Housing in India”. The amount of loans towards households below the targeted income threshold of INR 300,000 is not reported by the NHB. 33 The income threshold is kept unchanged at INR 120,000 (INR 10,000 per month). Hence it includes a much smaller share of urban households in India over time given growth and wage inflation. Assuming that wages grow at the rate of CPI inflation plus real GDP growth, households that made INR 120,000 in 2013 would be making INR 214,000 in 2018. 34 Discussion of the results indicators can be found in Annex I. 35 The NHB typically funds around 5 percent of new lending in the India housing finance market, but the share has been declining and is volatile as a large part of the NHB’s program comes from government allocations or guarantees for bond issuances under the Urban Housing Program and the Rural Housing Program with amounts varying year by year depending on government budgets. Page 16 of 45 The World Bank India Low-Income Housing Finance (P119039) Table 3: PDO Indicators in the Results Framework End Target Achievement 2013 2014 2015 2016 2017 2018 (for year of Targets 2018) The number of primary lenders active Exceeded / 25 42 44 50 52 52 35 in the EWS and LIG* Achieved segments. The annual increase in the volume of lending in the EWS and LIG segments by primary 103.6 123.6 115.6 80.0 79.6 98.4 155.4 Not Achieved lenders which are refinanced by the NHB (INR billion) The annual increase in the number of borrowers in the EWS and LIG income segments receiving 506,301 725,544 478,086 136,359 375,128 127,484 759,452 Not Achieved loans from primary lenders which are refinanced by the NHB. The percentage of replacements of non- performing with performing sub-loans Less by the QPLIs required Exceeded / 1.00 0.70 0.69 0.60 0.58 0.86 than for pools of loans Achieved 1.50 which have been refinanced by the NHB. (percent, a proxy for NPL ratio) Source: The NHB. Note: The two indicators on the annual increase in volume of lending and number of borrowers are reported in normalized terms in the results framework where the value in 2013 was normalized to 100 (percent). The normalization was done as actual figures were not available during project effectiveness. In this table, following the various Aide Memoires, the actual (non- normalized) volume of lending and number of borrowers are presented. The targets are the expected increase in volume of lending and number of borrowers for the last year of the project (2018). See Annex I for more discussions on the results indicators. *Note: As per the NHB, the lending in the EWS and LIG segments is proxied by loans with individual loan size below INR 1.5 million (US$20,000). Complementary data using alternative loan size thresholds are presented in Table 4. Page 17 of 45 The World Bank India Low-Income Housing Finance (P119039) Table 4: Supplemental Results Indicators (Complementing Indicators in the Results Framework) 2012 2013 2014 2015 2016 2017 2018 The volume of lending in the EWS and LIG segments* disbursed by the 215.9 268.4 292.8 296.5 332.3 389.1 489.0 HFCs (INR billion) The volume of lending disbursed by the HFCs with individual loan size below 22.7 25.3 28.2 31.5 40.2 52.4 55.6 INR 500,000 (INR billion) The volume of lending disbursed by HFCs with individual loan size below INR NA 8.8 10.9 13.8 NA 27.7 29.0 300,000 (INR billion) The number of borrowers serviced by the HFCs with individual loan size below NA 161,566 139,399 159,625 NA 267,763 306,848 INR 500,000 The number of borrowers serviced by the HFCs with individual loan size below NA 105,001 87,351 108,415 NA 209,385 240,913 INR 300,000 Gross Non-Performing Asset Ratio for the 1.23 1.11 1.18 1.08 1.09 1.11 1.31 HFCs (%) Source: The NHB, various issues of “Report on Trends and Progress of Housing in India”. *Note: As per the NHB, the lending in the EWS and LIG segments is proxied by loans with size below INR 1.5 million (US$20,000). Outcome: Improved Capability for Informal and Low-income Urban Households to Purchase, Build or Upgrade their Dwellings 22. The housing loans to low-income households have been mostly used for building and upgrading dwellings through self-construction. 36 The SEDD framework was critical in ensuring safeguards in these self-constructed houses. The extensive use of housing loans for self-construction was in line with the rampant trend of self-construction in the low-income neighborhoods. 37 Given the average loan size of INR 435,200, it can be inferred that the loans were used for purposes ranging from minor repairs and upgradations (cost as little as INR 30,000) to multi-story room new construction (cost as much as and more than INR 600,000). 38 However, exact data on usage of loans are not available due to a delay in the final impact evaluation (see below). 23. The project benefited from a healthy and fast-growing housing finance market and was highly complementary to the GoI’s affordable housing policies, most notably the Housing for All initiative. 39 36 Self-construction also includes hiring a contractor to construct the house. The mis-use of funds are unlikely as the HFCs usually tranche the loan disbursement against satisfactory completion of construction phases. (e.g. 30 percent disbursement on completion of housing foundation.) 37 “Self-Construct: Enabling Safe and Affordable Housing” by Micro Home Solutions, 2011. 38 The estimated cost of minor repairs (INR 25,000) and multi-room construction (INR 500,000) are from “Self-Construct: Enabling Safe and Affordable Housing” by micro Home Solutions, 2011. More comprehensive data on the use of funds by end-beneficiaries will come from the final impact evaluation, which has been considerably delayed and remain incomplete 6 months after project closure. 39 Housing for All (Urban), or Pradhan Mantri Awaas Yojana was launched in June 2015 with an aim to provide affordable housing for urban poor, targeting households with income below INR 600,000 per annum. Housing for All (Urban) is proposed to enable 20 million urban poor to get their own houses in urban areas by the year 2022 through a financial assistance of INR 2 trillion (US$29 billion) from central government. Housing for All has four components: 1) “In Situ” Slum Redevelopment, 2) Affordable Housing through Credit Linked Subsidy, 3) Affordable Housing in Partnership with private and public sector, and 4) Subsidy for Beneficiary-led House Construction/Enhancement. For detailed description, see “Housing for All (Urban) Scheme Guidelines”, Ministry of Housing & Urban Poverty Alleviation, the GoI (2015). Page 18 of 45 The World Bank India Low-Income Housing Finance (P119039) • On the demand side, the Credit Linked Subsidy Scheme (CLSS) under the GoI’s Housing for All initiative provides interest rate subsidies of up to 6.5 percent on eligible housing loans. Compared with the project, the CLSS targets a much wider income spectrum. The project focused on creating the housing finance market for the informal and low-income households and paved the way for additional public and private funds to more effectively serve these households using the CLSS. • On the supply side, the sustainability of the project crucially depends on key reforms in the Housing for All initiative, including slum development, the construction of affordable housing through public- private partnerships, and subsidies for beneficiary-led house construction and enhancement. The project also served to reinforce property rights, as anecdotal evidence suggests that increased housing finance for informal titles incentivized local administrations to formalize titles of informal lands/properties. Justification for Overall Efficacy Rating 24. The Efficacy of the project is rated Substantial. Designed as a pilot to test the frontier in lending to informal and low-income households, the project strongly supported the development of the informal and low-income housing finance market in India. In particular, the project not only provided housing loans to the targeted households at much lowered cost and longer tenor, but more importantly, also supported an MFD approach to develop the housing finance market and crowd-in private sector financing to more sustainably serve the informal and low-income households. The market was developed through a combination of stable financing (IDA credit line), market-based mechanisms (Special Refinance Scheme), regulatory reforms (including the SEDD framework), capacity building activities and market-based pilots. Though not fully captured by the PDO indicators, the MFD approach was highly effective in achieving the project’s intended outcomes and also highly complementary to the GoI’s Housing for All initiative. Acknowledging both the direct impact of the IDA credit line to end beneficiaries and the substantial MFD and market development impact of the project, the Efficacy is hence rated Substantial. C. EFFICIENCY Assessment of Efficiency and Rating This section analyzes how economically resources and inputs were converted to project results. An economic rate of return was not calculated in the PAD at appraisal and was not calculated in this ICR. Further quantitative analysis and an efficiency analysis in line with the economic analysis in the project appraisal document (PAD) are provided in Annex 4. 25. The IDA credit line, a key project input, was highly efficient in providing a stable source of funding towards informal and low-income housing finance. • The IDA credit line was unique among all of the NHB’s refinance schemes in its strong focus on serving low-income households and in addressing market failures caused by informality: The income Page 19 of 45 The World Bank India Low-Income Housing Finance (P119039) threshold of INR 300,000 was much lower than contemporary NHB refinance schemes, and 95 percent of the IDA funding was allocated to households with informal income and/or informal title (Table 2). 40 • The IDA credit had longer tenor than other sources of funding for the NHB and the HFCs. Such stable long-term funding significantly reduced funding uncertainty and incentivized the NHB and the PLIs to break into the informal and low-income segment. 41 The cost of the IDA credit line was also competitive compared to other sources of funding, which provided another incentive for the NHB to take the credit line, though the funding was channeled to the PLIs with market-based interest rates. 42 • The size of IDA financing was also appropriate as a pilot to test the market frontier. The funding was small relative to the entire housing finance market and was easy for the NHB to absorb. At the same time, the IDA funding size was significant relative to the informal and low-income segment of the market so as to create a substantial impact. Given the innovative nature of the project, the financing also served as a pilot to put in place the needed upstream reforms and establish market channels of financing. 43 26. The NHB effectively implemented an MFD approach to crowd-in private sector financing, multiplying the impact of limited project resources and improving project efficiency. The stable IDA financing provided through the NHB’s Special Refinance Scheme, combined with capacity building, piloting and upstream regulatory reforms, unleashed the potential of private sector PLIs to serve the targeted households (paras. 19 and 20 and Table 4). This MFD approach created a multiplier effect for resources deployed in the project and improved the project’s efficiency. However, frequent changes in the PIU team and in NHB management did cause some frictions in project implementation, and certain M&E activities envisaged to be taken up by the NHB using its own resources, including a final impact evaluation and a third-party audit of the SEDD framework, were delayed (para. 41). 27. Project implementation by the World Bank team was efficient, including in addressing key bottlenecks in the project (see para. 48 in the Bank Performance section). 28. The Efficiency of the project is rated Substantial. Notwithstanding the delay in certain M&E activities and frequent PIU changes, the IDA credit line was used in a highly efficient manner to provide a 40 The NHB’s Urban Housing Fund has eligible income threshold of INR 600,000. No other NHB’s refinance scheme has a special focus on informality. The low-income households targeted by the project were highly disadvantaged in the housing finance market at project appraisal. Before the project was in place, only 36 percent of the targeted low-income households obtained their housing loans from banks and government programs, two of the cheapest sources of credit, versus 76 percent for households with income above the threshold. On the contrary, 14 percent of these low-income households received housing loans from money lenders, versus only 1 percent for households with income above the threshold. For the low-income households with informal income and informal title, which was the target of the project and the IDA credit line, the access to finance was even worse. (Data from India Human Development Survey) 41 The NHB does not have access to such stable long-term funding like the IDA credit in domestic borrowing. For HFCs, they used to have most (80 to 90 percent) of their funding from short term borrowing, with only 10 to 20 percent of their funding from more long-term borrowing. 42 The cost advantage of IDA credit line, reflected by its low interest rate, is diluted at certain times of the project due to volatility in foreign exchange market. At appraisal stage, the net all-in reduction of the PLI’s cost of funds is estimated to be around 200-300 basis points. However, volatility in the foreign exchange market has at times increased the hedging cost of the IDA credit line, which reduce the cost advantage of the IDA credit line to lower level than estimated at project appraisal. The NHB confirmed that despite the adverse movement, the long maturity of the loan together with the broader support package still makes this an important project to pursue. 43 Additional World Bank financing is currently under discussion with the NHB. Page 20 of 45 The World Bank India Low-Income Housing Finance (P119039) stable source of funding that was well targeted towards the underserved informal and low-income segment. The NHB and the World Bank also ensured the efficient use of resources with an MFD approach by setting up the market infrastructure to leverage private sector PLIs to serve informal and low-income households. D. JUSTIFICATION OF OVERALL OUTCOME RATING 29. The overall outcome rating is assessed as Satisfactory based on the preceding assessment and detailed in the table below. Table 5: Summary of Outcome Ratings Relevance of PDO Efficacy Assessment Efficiency Assessment Outcome High Substantial Substantial Satisfactory E. OTHER OUTCOMES AND IMPACTS (IF ANY) Gender 30. The project promoted the financial inclusion of women. The project supported PLIs catering to female-headed households. 44 Several HFCs also typically included women as co-borrowers or guarantors to secure credit risks. The project helped households access formal financing to purchase a house, with several lenders offering a discount of 0.5 percent on the interest rate if a woman was on the title for the property, due to their improved risk profile and better repayment record. These terms meshed well with several local government initiatives of reducing the stamp duty for women purchasers or joint purchasers; in Delhi, for instance, the stamp duty was reduced from 6 to 4 percent if a woman was on the deed. These measures contributed to improving women’s empowerment in their communities. Anecdotally, beneficiaries have also reported additional employment opportunities, particularly for female family members, once they own a house. Institutional Strengthening 31. The project’s new approach towards the SEDD framework has become the model for other World Bank housing finance projects. The new approach focused on the macro/institutional level, combined with sample-based loan-level monitoring, and emphasized the importance of institutionalizing rather than micromanaging better practices. As part of this approach, primary lenders were asked to review their policies regarding social and environmental issues and adopt due diligence measures as appropriate to each of them. The PLIs reported being satisfied with the process and commented on the value of the approach in protecting the value of the loan security against risks such as poor construction or flooding. This model has been subsequently adopted by World Bank housing finance projects in Nigeria and Kenya. 44 Sewa Grih Rin, one of the PLIs, primarily lends to women borrowers. Page 21 of 45 The World Bank India Low-Income Housing Finance (P119039) III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 32. The project was innovative and untested in India and other low-income markets, and the World Bank assumed a calculated risk in approving the project (see para. 8). The project preparation took a long time and faced close internal scrutiny within the World Bank. The World Bank eventually approved the project based on its strong developmental impact in potentially creating new market and pushing new frontiers in low-income housing finance. 33. The project design adopted an MFD approach to effectively leverage public and private resources (para. 9). 34. Relying on strong preparatory work, the project design targeted the appropriate end beneficiaries. 45 Preparatory work and studies identified the market failure for low-income housing finance caused by the informality of income and informality of title. Preparatory studies also informed the choice of income threshold for project intervention, avoiding setting either too high an income threshold where the housing finance market already functioned or too low an income threshold where market mechanisms may not have worked effectively. The income threshold of market-based lending to informal households was also continuously tested during project implementation (para. 48, bullet point 1). B. KEY FACTORS DURING IMPLEMENTATION Factors subject to Government or implementing entities’ control 35. The project kept its strong focus on serving informal and low-income households throughout implementation. The NHB had requested to raise the income threshold for targeted households to INR 400,000 per annum and increase lending to formal households. The midterm review in 2015 was used strategically to reassert the relevance and appropriateness of the project design. Taking into account inflation and wage growth over time while being consistent with the project’s objective of reaching down to lower-income households, the income threshold eligible for refinancing by the NHB under the IDA credit line was adjusted upward from INR 200,000 to INR 300,000 per annum, while the income threshold for incremental housing remained at INR 200,000. Importantly, the project remained focused on lending to households with informal income and informal property title. 36. Initial disbursements under the project stagnated due in part to internal GoI budgetary requirements and procedural issues with fund flows. As a result, the NHB could not receive any transfers of IDA funding from the Department of Financial Services (DFS) during the first two years of 45Prior to the project, a Low-Income Housing Finance Technical Assistance to the NHB was implemented, funded by FIRST. One key output is the report on ‘Expanding the Housing Finance Market to Cover Lower Middle-Income Segments in India’, prepared by Monitor Group. Page 22 of 45 The World Bank India Low-Income Housing Finance (P119039) implementation. The World Bank team worked closely with the NHB and the DFS to address the funding flow issues by project midterm and ensured smooth disbursement of the IDA credit subsequently. 37. Provided with an enabling environment, India HFCs pioneered innovative lending tools targeting informal and low-income households (para. 19, bullet point 3). 46 Factors subject to World Bank control 38. Proactive supervision and implementation support from the World Bank Group (WBG) was critical for the project’s efficacy (see para. 48). Factors outside the control of the Government and/or implementing entities 39. Self-construction was rampant in low-income neighborhoods in India, and most of project’s loans were used for building or upgradation through self-construction (para. 22). IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. QUALITY OF MONITORING AND EVALUATION (M&E) M&E Design 40. While some results indicators aligned well with the Theory of Change (Figure 1, constructed from the PAD for this ICR), the results framework could have been strengthened to better capture the MFD impact of the operation. Potential adjustments could have included, for example, adding indicators on total volume of lending and number of loans from HFCs to the informal and low-income households (not necessarily refinanced by the NHB). 47 M&E Implementation 41. M&E Implementation • Key project data were regularly reported by the NHB. Both the World Bank team and the NHB made an effort to collect performance indicators that were not included in the initial results framework, including the interest rate and tenor profile of loans. These indicators were tracked in addition to the indicators in the result framework to better inform the progress of the project. A restructuring to adjust the results indicators was proposed at midterm review, as documented in various aides- mémoires, but was not implemented due to challenges of internal management coordination within 46 Information on some of the technology in reaching the informal and low-income households were shared in the Global Housing Finance Conference: https://www.worldbank.org/en/events/2018/05/30/8th-global-housing-finance-conference 47 For example, the supplemental indicators in Table 4, which capture the overall housing loans for the informal and low-income households rather than just loans refinanced by the NHB, could be used as supplements. Page 23 of 45 The World Bank India Low-Income Housing Finance (P119039) the NHB to agree on the issue. The World Bank team also conducted frequent field visits to the PLIs and end-beneficiaries to understand the development of the low-income housing finance market. • Two planned M&E activities, the final impact evaluation and third-party audit of SEDD, were delayed and not yet implemented. The World Bank team constantly communicated with the NHB on the importance of these M&E activities, and intensively so in the last two years of project implementation, as documented in aides-mémoires. However, while the NHB remains committed to these activities, internal management and PIU changes and the strong preference to use internal resources for these M&E activities contributed to the delays. M&E Utilization 42. Monitoring data were regularly analyzed to track performance and identify problems in implementation. The collected project data, including the affordable interest rate, small loan size, and large numbers of PLIs and end-beneficiaries give assurance on the efficacy of the project. In addition, the World Bank team’s frequent field visits to the PLIs and particularly to end-beneficiaries (informal and low- income households) brought first-hand knowledge to inform project implementation and also partially compensated for the delay in final impact evaluation. The World Bank team also made a video to capture important lessons from the project, which has been widely disseminated internally within the World Bank and externally to the general public. However, the delay in the final impact evaluation lessened the opportunity to disseminate some further lessons. Justification of Overall Rating of Quality of M&E 43. The overall M&E is rated Modest. The M&E system as designed and implemented informed project implementation, and the World Bank team actively collected data outside the results framework. Frequent field visits to PLIs and end-beneficiaries also informed project implementation. However, the original results framework could have been augmented to better capture the MFD impact of the project. In addition, two M&E activities—the final impact evaluation and third-party audit for the SEDD—were delayed, resulting in the Modest rating. B. ENVIRONMENTAL, SOCIAL AND FIDUCIARY COMPLIANCE 44. The project complied well with the World Bank’s safeguards policies. 48 The SEDD framework, which focused on the good practice application of social and environmental safeguards during credit appraisal, became the institutional framework for the NHB and the PLIs. The NHB carried out periodic internal audits regarding the safeguards compliance of the PLIs, and recorded their satisfactory performance. With regard to third party audit of the SEDD framework, the NHB could not secure a relevant consultant, as the audit job was relatively small. Given this practical difficulty, the NHB has informed the World Bank that the audit aspects have been integrated as part of their annual financial audit, which is 48The Project did not trigger OP 4.12 and OP4.10, since no civil works were financed under the project and the project area did not have the presence of indigenous characteristics as mentioned in the World Bank’s OP 4.10. Page 24 of 45 The World Bank India Low-Income Housing Finance (P119039) generally finalized during November to December. While the third-party audit is important, it is more important that the NHB internalizes the procedures integrating safeguards risks and gender inclusion aspects in their regular procedures. Considering this, the safeguards rating is maintained as Satisfactory at project closing. 45. The NHB has an institutional mechanism for grievance redressal (Grievance Registration and Information Database System, or GRIDs) as well as a Fair Practice Code for the HFCs, which were drafted and are supervised, respectively, by the Department of Regulation and the Department of Supervision. As part of the grievance redressal process, the NHB periodically conducts training activities, such as the training on “Prevention of Fraudulent Practices in Housing Finance” conducted in March 2018. In addition, the NHB is also a member organization of the Centralized Public Grievance Redress and Monitoring System run by GOI’s Department of Administrative Reforms and Public Grievances. 46. Overall, the project complied well with fiduciary agreements. The financial management performance rating remained Satisfactory throughout the project implementation period, except for the initial few years when it was downgraded to Moderately Unsatisfactory due to (a) the non-availability of budget funds to the NHB from the GoI; (b) administrative and procedural bottlenecks at the DFS, resulting in long disbursement delays in submission of Interim Unaudited Financial Reports (IUFR) to the World Bank; and (c) non-submission of the first year project audit report. These financial management issues were subsequently addressed by the end of December 2015 through (a) the creation of a separate budget line for the project in the budget books of the DFS and release of funds to the NHB; (b) streamlining the IUFR submission and approval processes in the NHB and the DFS; and (c) reaching an agreement on terms of reference for project audit between the NHB and the World Bank. Apart from these initial budget and fund flow issues, there were no other financial management issues during project implementation. Most of the project audit reports for later years were submitted ahead of the due date, and no accountability issues were reported by the auditor. On procurement, the NHB had a strong preference to conduct capacity building and impact evaluation activities in house and was less open to hiring external consultants to supplement internal capacity, as per agreed implementation arrangements. C. BANK PERFORMANCE Quality at Entry 47. The World Bank worked effectively during project preparation. • The World Bank team proactively supported a series of preparatory technical assistance studies which were led by the NHB and provided the rationale for the engagement in this segment. The technical assistance analyzed the market failure caused by informality and identified the appropriate end beneficiaries of the project (para. 34). The close engagement with the NHB during project preparation was highly important, given its strategic role in the housing finance market in India and the fact that this was the first time the NHB operated a lending project with financing from the World Bank. Page 25 of 45 The World Bank India Low-Income Housing Finance (P119039) • The project design had the MFD principle at its core. The project was designed before the MFD approach was announced by World Bank management as a core principle for development, and was at the cutting edge of World Bank housing finance operations at the time of design. The project’s public sector counterpart, the NHB, was the apex institution for housing finance in India and was strongly committed to affordable housing finance. While the project relied on the NHB to create an enabling operating environment, the design aimed to incentivize private sector PLIs and leverage their expertise to deliver credit to end-beneficiaries. Such a design was well aligned with the MFD principle and was the key to achieve the project’s intended outcomes. • The World Bank took on a well-calculated risk in approving the project. Informal and low-income households were largely underserved by the formal housing finance market, and the capacity of financial institutions to serve the segment was lacking at project appraisal (para. 8). Based on strong preparation work, the World Bank eventually concluded that the risks were justified given the project’s potential impact and demonstration effect, and approved it to advance the market frontier in informal and low-income housing finance. Quality of Supervision 48. The World Bank’s active supervision was critical for the project to achieve its intended outcomes. • The proactive engagement of the World Bank team ensured that the project tested the income threshold of market-based lending to informal households. At various stages of the project, the World Bank faced resistance from the counterpart to tap into the informal and low-income housing finance market, due to perceived high risk and difficulty in lending to the segment. Requests to increase the income threshold and increase lending against formal dwellings were made several times by the NHB. Such challenges were exacerbated by changes in NHB management and the PIU team. 49 The World Bank task team, with strong support from the World Bank management, including from the Country Management Unit, collaborated closely with the NHB and other stakeholders and conducted continuous market-sounding activities, technical assistance and market-based pilots. The World Bank supervision missions engaged with numerous lenders across India and in different states, including meetings in Nagpur, Delhi, Mumbai, Bangalore, Bhopal, Jaipur and Ahmedabad. Each of these visits was coordinated with the NHB, often including the NHB’s state-level representative, and provided an opportunity for dialogue among lenders, the NHB and the World Bank. The missions also assessed the role of the NHB in addressing informality and how the challenges varied from state to state. Through these activities, the NHB gradually improved its capacity to serve the informal and low- income segment and performed necessary upstream reforms to enable market-based solutions (para. 19). • The World Bank team provided extensive support in critical areas of the project, including close handholding during the implementation. This included design and implementation of the Special 49 Repeated communications, usually from the very basics, are needed when changes in management and PIU takes place. Page 26 of 45 The World Bank India Low-Income Housing Finance (P119039) Refinance Scheme and the SEDD framework, and in-depth analysis into informal property and informal income in India. For example, the supervision team worked closely with the NHB’s refinance team and credit departments to design the criteria for the Special Refinance Scheme, and also provided extensive inputs into the SEDD framework. In addition, the Bank team proactively addressed a delay in the flow of IDA funds from the Department of Financial Services to NHB (para. 36).50 • The World Bank worked effectively with the International Finance Corporation (IFC) to provide a holistic package of support. IFC played a key role in fostering the development of a new group of lenders who saw value in leveraging technology to make lending to lower-income households economically viable. Many of these lenders benefited from the project and would not have grown at the rate they did without access to project funding and the supportive environment created for lending to the informal sector. Alongside the investment activities, the IFC also complemented the program with some capacity building for the credit bureau, CIBIL, and the mortgage asset registry system, CERSAI, both of which are important parts of the infrastructure supporting India’s mortgage market. The close collaboration among the World Bank, the IFC and its private sector clients (HFCs) also helped bridge the knowledge gap between the NHB and the HFCs and facilitated the design and implementation of key market infrastructure, including the NHB’s Special Refinance Scheme. • The World Bank team made substantial efforts in project M&E, including by collecting data outside of the results framework, conducting frequent field visits to PLIs (mostly HFCs) and particularly end- beneficiaries. These efforts significantly informed project implementation. In light of issues with the results indicators (unable to capture full project impact), the slow disbursement for capacity building activities, and the NHB’s strong preference to fund these activities using internal resources, the World Bank team had actively proposed a project restructuring to (a) make modifications to the results framework to better reflect project outcomes; and 2) reallocate resources in favor of the refinancing component. The restructuring was discussed during multiple supervision missions since 2015. However, the request for restructuring by the NHB was delayed due to challenges of internal management coordination within the NHB to agree on the issue. Eventually, the reallocation of funds was agreed to and implemented in 2018. Justification for Overall Rating of Bank Performance 49. The World Bank performance is rated as Satisfactory. As discussed above, the proactive and effective work of the World Bank during both entry and supervision was critical for the project to achieve its intended outcomes. 50 Faced with the NHB’s preference to fund planned capacity building activities using internal funds, the World Bank team has also continuously proposed new capacity building activities in line with evolving priorities, as documented in respective Aide Memoires. For example, in 2017, the World Bank task-team and the NHB agreed to potential new technical assistance activities to the NHB to strengthen its regulatory and supervisory mandates. The activities include the development of a housing information center, review of the NHB flagship publication on “trends & progress in housing”, stress testing of housing finance portfolios and institutions, and green housing finance. However, these activities have not materialized given the NHB’s preference to use internal resources for such capacity building activities. Page 27 of 45 The World Bank India Low-Income Housing Finance (P119039) D. RISK TO DEVELOPMENT OUTCOME 50. If complementary supply-side reforms are not well implemented, insufficient supply of affordable housing impose risks to the project’s outcomes. The project’s interventions, by developing the low-income housing finance market, targeted mainly the demand side of low-income housing. Amid this urbanization trend, supply-side constraints such as a dearth of land serviced by utilities, if not properly addressed, will limit the availability of decent affordable houses and push up housing prices. The GoI’s Housing for All initiative includes a comprehensive set of supply-side interventions (para. 23). However, implementation of these reforms remains challenging, as land governance falls under state governments and is fragmented. 51 If supply-side reforms are not properly implemented, supply-side constraints could significantly reduce the ability of low-income households to purchase, build or upgrade their dwellings. 51. Financial instability imposes risks to the project’s outcome. India’s overall banking sector has witnessed an increase level of NPLs, which threatens financial stability. 52 The GoI is aware of the risk and has taken steps to improve asset quality for banks and strengthen the balance sheet of financial institutions. 53 If not properly managed, fast-rising NPLs and instability in the financial system could significantly reduce the amount of credit in circulation, adversely affecting the nascent informal and low- income housing finance market. V. LESSONS AND RECOMMENDATIONS Lessons and Recommendations 1 Factors that Affected The project implemented an MFD approach to unleash the huge potential of Performance private sector PLIs to serve the informal and low-income households. The sustainable development of the informal and low-income housing finance Lessons market benefits from an MFD approach with a holistic package of interventions supported by the WBG. The WBG should consider adopting a comprehensive package of interventions to mobilize private sector financing for the sustainable development of the informal and low-income housing finance market. The interventions would vary by country but could include a stable funding source combined with Recommendations appropriate incentive mechanisms (IDA credit line delivered through the Special Refinance Scheme), upstream regulatory reforms (including the SEDD framework), and fostering of the HFCs (including capacity building and IFC direct investment). 51 The enactment of the Real Estate (Regulation & Development) Act, 2016 was seen as a game-changer that will bring discipline to a largely unregulated industry. It was observed nonetheless that in some states the establishment of the Real Estate Regulatory Authority (RERA) was delayed and its implementation was slower than envisaged due to administrative challenges and policy coordination failures. 52 The overall banking sector NPL level rose from around 4.3 percent in 2015 to more than 11.2 percent in 2018. 53 Reforms include for example the Insolvency and Bankruptcy Code (2016). See India – Financial Sector Assessment (2017) and Financial Sector Stability Assessment (2017) for detailed discussion. Page 28 of 45 The World Bank India Low-Income Housing Finance (P119039) Lessons and Recommendations 2 The approach of the project was innovative and untested in India and other Factors that Affected low-income markets, and the World Bank assumed a calculated risk in Performance approving the project as a pilot to test frontier of housing finance market for informal and low-income households. Creating new markets and pushing the market frontiers including in housing Lessons finance often involve innovations that requires taking on calculated risks based on strong preparatory work. The World Bank should encourage innovation and take on well-calculated risks Recommendations in order to create new markets and push new market frontiers such as housing finance for informal and low-income households. Lessons and Recommendations 3 The project’s intervention targeted mainly the demand side of low-income Factors that Affected housing by improving access to housing finance and complemented the supply Performance side reforms in the housing market notably the GoI’s Housing for All Initiative. Addressing challenges in low-income housing benefits from a holistic approach Lessons with complementary interventions on both demand and supply side. The World Bank should continue working closely with clients as well as Recommendations internally across global practices to take advantage of the complementarity between interventions on demand and supply side of low-income housing. Lessons and Recommendations 4 Factors that Affected The project’s approach towards the SEDD framework focused on Performance institutionalizing the framework for the NHB and the PLIs. Setting up the SEDD framework as an institutional framework between the Lessons NHB and the PLIs combined with sample-based loan-level monitoring is an efficient and sustainable way to ensure social and environmental safeguards. The World Bank team should adopt the SEDD framework as an institutional Recommendations framework to ensure efficient and sustainable safeguards in housing finance projects. Lessons and Recommendations 5 While several capacity building activities were implemented using counterpart Factors that Affected funding, the final impact evaluation was delayed, postponing the opportunity Performance to derive further lessons from the project. Important M&E activities can be subject to delays due to limited counterpart Lessons resources. The World Bank should consider exploring funding M&E studies with strong Recommendations “public good” nature directly instead of through the counterpart. . Page 29 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Provide access to housing finance for low-income households to purchase,build, or upgrade dwellings Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of Primary Lending Number 25.00 35.00 52.00 Institutions active in the target low-income segment 30-Jun-2013 31-Dec-2018 04-Dec-2018 Comments (achievements against targets): This indicator reports the number of primary lending institutions (PLIs) active in the Economically Weaker Section (EWS) and Low Income Group (LIG) segments. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Lending Volume Growth Percentage 100.00 150.00 95.00 Page 30 of 45 The World Bank India Low-Income Housing Finance (P119039) 30-Jun-2013 31-Dec-2018 04-Dec-2018 Comments (achievements against targets): The indicator reports the increase in the volume of lending in the Economically Weaker Section (EWS) and Low Income Group (LIG) segments by primary lenders which are refinanced by the NHB, with the number at the baseline normalized to 100. The increased volume of lending was INR 103,626 billion in 2013 and 98,397 in 2018, so the increased volume of lending in 2018 is 95 percent of the increase in the baseline year, falling below the 150 percent target. Notice that the indicator captures the increase in the volume of lending in the EWS and LIG segments refinanced by the NHB, so the total outstanding volume of loans refinanced by the NHB have increased, albeit at a slower rate than the broad housing finance market for low-income households (Table 4). Such picture is consistent with the NHB’s catalytic role in housing finance market development, with the private sector taking a larger and growing role in financing the EWS and LIG segments. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of Borrowers growth Percentage 100.00 150.00 25.00 30-Jun-2013 31-Dec-2018 04-Dec-2018 Comments (achievements against targets): The indicator reports the increase in number of borrowers in the EWS and LIG segments by primary lenders which are refinanced by the NHB, with the number at the baseline normalized to 100. The increased number of borrowers refinanced by the NHB was 506,301 in 2013 and 127,484 in 2018, so the increased number of borrowers in 2018 is 25 percent of the increased number in the baseline year, falling below the 150 percent target. Notice that the indicator captures the increase in the number of borrowers refinanced by the NHB in the EWS and LIG segments, so the total outstanding number of borrowers refinanced by the NHB have increased, albeit at a slower rate than the broad housing finance market for low-income households. Evidence suggests that the number of low-income borrowers (including those not refinanced by the NHB) have increased quickly during the Project implementation period, as evident by (i) the increasing number of borrowers from the HFCs Page 31 of 45 The World Bank India Low-Income Housing Finance (P119039) with individual loan size below INR 500,000 and INR 300,000 (both reported in Table 4 of the ICR), (ii) the fast-growing lending volume towards the EWS and LIG segments by HFCs (reported in Table 4 of the ICR), and (iii) the verbal feedback the World Bank task team received from the NHB and the HFCs. Such trend is consistent with the NHB’s catalytic role in housing finance market development, as the private sector has developed and takes a growing role in financing the EWS and LIG segments. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion NPLs (percentage of Percentage 1.00 1.50 0.86 replacements of non- performing with performing 01-Apr-2013 31-Dec-2018 04-Dec-2018 sub-loans by primary lenders required for pools of loans which have been refinanced by NHB) Comments (achievements against targets): The end target is for the value to fall below 1.50%. This indicator reports percentage of replacements of non-performing with performing sub- loans by the PLIs required for pools of loans which have been refinanced by the NHB. This indicator is a proxy for asset quality. These PLIs provide part of their mortgage loan portfolio as collateral for the NHB refinance. Should any of the loans provided as collateral become delinquent, they must be replaced with a performing loan. This ensures that the NHB minimizes its direct credit risk exposure to the mortgage market. Page 32 of 45 The World Bank India Low-Income Housing Finance (P119039) A.2 Intermediate Results Indicators Component: Component One - Capacity Building of the NHB, QII and QPLIs Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion NHB issues Lending Number 0.00 2.00 1.00 Guidelines for Housing Loans with Alternative Security 01-Apr-2013 31-Dec-2018 04-Dec-2018 Arrangements Comments (achievements against targets): Component: Component Two - Refinancing Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of Pilots Launched Number 0.00 4.00 4.00 01-Apr-2013 31-Dec-2018 04-Dec-2018 Comments (achievements against targets): Four pilots were launched under the NHB's Special Refinance Scheme. AU Housing Finance and India Shelter piloted lending largely in Rajasthan against informal incomes and informal tenure using the Scheme. The NHB also piloted a line through Ujjivan, a microfinance institution intermediated through Tata Capital, to reach low income borrowers. Page 33 of 45 The World Bank India Low-Income Housing Finance (P119039) Component: Componenet Three - Project Implementation Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Technical Assistance is Number 0.00 5.00 6.00 implemented according to agreed business plan 01-Apr-2013 31-Dec-2018 04-Dec-2018 Comments (achievements against targets): Technical assistance on the social and environmental due diligence framework, the design & implementation of market-based pilots and the NHB's Special Refinance Scheme were implemented. Page 34 of 45 The World Bank India Low-Income Housing Finance (P119039) A. KEY OUTPUTS BY COMPONENT Objective/Outcome 1: Improved access to sustainable housing finance for informal and low-income urban households. 1. The increased volume of lending in the EWS and LIG segments by primary lenders which are refinanced by the NHB. Outcome Indicators 2. The annual percentage increase in the number of borrowers in the EWS and LIG income segments receiving loans from primary lenders which are refinanced by the NHB. 1. Number of pilots launched. Intermediate Results Indicators 2. Number of technical assistance implemented according to agreed business plan. 1. Affordable lending delivered to targeted informal and low-income urban households. Key Outputs by Component 2. Increased capacity of the NHB to serve informal and low-income (linked to the achievement of the Objective/Outcome 1) households: the NHB established the Special Refinance Scheme to deliver the IDA credit line, and the SEDD framework was implemented to ensure safeguards. Objective/Outcome 2: Development of a sustainable market for housing finance for the informal and low-income urban households. 1. The increased number of primary lenders active in the EWS and LIG segments. Outcome Indicators 2. The percentage of replacements of non-performing with performing sub-loans by the QPLIs required for pools of loans which have been refinanced by the NHB. 1. Number of specific guidelines for sustainable lending to target Intermediate Results Indicators segments. 2. Number of pilots launched Page 35 of 45 The World Bank India Low-Income Housing Finance (P119039) 1. Increased participation from PLIs in serving the informal and low- income households. 2. The NHB’s Special Refinance Scheme to crowd in private sector Key Outputs by Component financing to informal and low-income households. (linked to the achievement of the Objective/Outcome 2) 3. The SEDD framework, which is used as an institutional framework by the NHB and PLIs to ensure building standards and consumer protection. Page 36 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Bertine Kamphuis Task Team Leader(s) Heenaben Yatin Doshi Procurement Specialist(s) Papia Bhatachaarji Financial Management Specialist Meera Chatterjee Social Specialist Sumriti Singh Team Member Suryanarayana Satish Team Member Barjor E. Mehta Team Member Vikram Raghavan Team Member Sita Ramakrishna Addepalli Team Member Sita Ramakrishna Addepalli Social Specialist Shiny Jaison Team Member Satya N. Mishra Social Specialist Simon Christopher Walley Team Member Supervision/ICR Anuradha Ray Task Team Leader(s) Heenaben Yatin Doshi Procurement Specialist(s) Arvind Prasad Mantha Financial Management Specialist Poorna Bhattacharjee Team Member Ashutosh Tandon Team Member Sakshi Varma Team Member Page 37 of 45 The World Bank India Low-Income Housing Finance (P119039) Rima Sukhija Team Member Kaushik Sarkar Team Member Simon Christopher Walley Team Member Ekaterina Grigoryeva Environmental Specialist Savita Dhingra Team Member Payal Malik Madan Team Member Sita Ramakrishna Addepalli Environmental Specialist Marius Vismantas Team Member I. U. B. Reddy Social Specialist Yunfan Gu Team Member A. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY10 18.923 131,064.12 FY11 26.707 156,669.35 FY12 43.540 263,578.12 FY13 34.372 210,223.43 FY14 0 0.00 Total 123.54 761,535.02 Supervision/ICR FY14 32.893 171,288.10 FY15 28.925 147,933.58 FY16 32.210 144,408.03 FY17 27.854 129,369.37 FY18 20.218 175,526.05 FY19 23.567 203,366.53 Page 38 of 45 The World Bank India Low-Income Housing Finance (P119039) FY20 .147 639.15 Total 165.81 972,530.81 Page 39 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval Actual at Project Percentage of Approval Components (SDR M) Closing (SDR M) (SDR M) Component 1: Capacity 1.32 0.00 0 Building Component 2: Financial Support for Sustainable and 64.12 66.1 0 Affordable Housing Component 3: Project 0.66 0.00 0 Implementation Total 66.1 66.1 0.00 Page 40 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 4. EFFICIENCY ANALYSIS The following quantitative analysis further demonstrate cost-effectiveness of the project. • First, the IDA credit line was highly focused in serving the informal and low-income households (all loans financed by IDA credit line delivered to low-income households, and 95 percent to households with informal income and/or informal title). Before the project was in place, many informal and low- income households were completely excluded from the housing finance market and cannot borrow for housing at all, and even the households who could borrow have to largely rely on informal money lenders and pay an average interest rate of 27 percent. In comparison, the informal and low-income borrowers under the IDA credit line only pay an average interest rate of 15 percent with an average maturity of more than 10 years. This means the IDA credit line of US$100 million managed to achieve a reduction of at least 12 percentage points in the interest rate paid by the targeted households, and also served households who were not able to obtain financing before the project was in place. • Second, the efficient use of resources is also ensured by the MFD approach of the project, as this US$100 million project leveraged substantial amount of private sector financing to serve the targeted households. As discussed in the ICR, one big impact of the project was to mobilize private sector financial institutions to serve the informal and low-income households at much lower cost and longer tenor, as evident by the quick rise of affordable HFCs and the increase in lending towards the informal and low- income households. For example, as discussed, the volume of lending disbursed by the HFCs with individual loan size below INR 500,000 has increased from INR 25.3 billion in 2013 to INR 55.6 billion in 2018. If we take a relatively conservative estimate and assume that 10 percent of the total increased disbursement from 2014 to 2018 are mobilized by this project to finance the informal and low-income households. This would amount to INR 81.4 billion (US$1.2 billion). This conservative estimate means that an additional US$1.2 billion of affordable housing loans have been mobilized by this US$100 million project to households who either were not able to receive financing or had to pay very high interest rate for financing before the project was in place. Such analysis showcases the efficient use of the resources and the strong multiplier effect of the project. Additional efficiency analysis in line with the economic analysis in the PAD is performed subsequently. The PAD grouped the expected impacts of the project along three main categories. The macroeconomic impact would be channeled through the project's demonstration effects by increasing the size of the mortgage market and making mortgage loans available to households who previously had no access to housing credit. The financial impact would operate by strengthening the balance sheets of the QPLIs and allowing the lenders to expand the availability of credit and/or increase the orientation of their portfolio towards mortgage loans. The microeconomic impact would take place at the household level through improvements in the welfare of households who obtain mortgages from the QPLIs, as well as any effects on the welfare of households who may benefit from improved access to credit in the future through the project's demonstration effects. The project addresses a market failure, as low-income people with informal incomes or property titles are not able to get access to housing finance, and the supply demand gap is worse for people in this segment. More than 15,000 beneficiaries were able to get housing finance, who had an income less than INR 300,000 per annum and most with informal income and informal title. Page 41 of 45 The World Bank India Low-Income Housing Finance (P119039) The macroeconomic impact is the increase in funds available for finance in this segment. The project represented an increment in the funds available, through its direct financing as well as crowding-in private funds through supporting market development. The NHB also carried out a project called Urban Housing Fund, running from 2013/14 to 2017/18. Refinance assistance was extended to PLIs with respect to their loans to 218,000 households. However, this catered to households with annual income not exceeding INR 600,000. This indicates that the project though unique in its target group, needs to be scaled up. Much will depend on how the NHB itself takes it forward, given the large scale of operation of the NHB. The project has demonstrated the possibility to address the informal segment. In terms of microeconomic impact, the social benefits are likely to be high because the gap between demand and supply of housing is large, and also because the beneficiaries are low-income people. Anecdotal evidence gathered by the project team also suggests that the benefits to the beneficiaries are considerable. The video made on the project also brings home the benefits of the project. Another example was a truck mechanic from Jaipur, who would be able to pay off his loan for his house in less than ten years. A resident of a neighborhood in West Delhi was happy that they now own their house and the owner was happy with the enhanced quality of life. To buy their house, they took a loan of INR 600,000, at 16 percent interest rate payable in 7 years, monthly mortgage payment is INR 11,918. If we assume that in the absence of this loan facilitated by the project, the household would have to pay an interest rate of 20 percent, payable in 7 years, monthly payment would have been INR 13,324. If as is not unusual the household was required to pay an interest rate of 25 percent, monthly payment would have been INR 15,187. Therefore, the project leads to quite substantial reduction (well over 20 percent) in interest paid by the beneficiaries, and since the beneficiaries are from the low-income segment, the social benefits are substantial. In addition, without the cheaper loans and the SEDD framework, many of these low-income households potentially cannot afford a decent housing, which means large welfare losses as owning a decent house brings health, social, and phycological benefits that can be hardly quantified in numbers. By demonstrating the possibility of intervening in a positive manner in the housing supply to this segment, scaling up would have great social benefits. The project significantly contributed to the development of a sustainable market for informal and low-income housing finance and helped crowd in private sector funding to this segment. The financial impact is also significant, as the enabling environment created by the NHB supported the financial institutions to develop their capacity to serve the informal and low-income households. This allows them to better intermediate funds and expand their business towards these previously under- served households. In terms of internal efficiency, the project disbursement improved in later years, though funding for components 1 and 3 was transferred to component 2 due to the NHB’s preference to use internal resources for capacity building activities. One aspect of the project is the relatively high ratio of the project preparation staff cost to supervision/ICR cost (see Table below). This was caused by the strong preparation work needed by this project and the long timespan of the project preparation, given the project was innovative and designed to test the frontier of lending to informal and low-income households. The investment made in the preparation of this project has not only helped catalyze the housing finance Page 42 of 45 The World Bank India Low-Income Housing Finance (P119039) market for informal and low-income segment in India, but the experience and learning has been used to inform World Bank housing finance operations in other countries including Nigeria and Kenya. Table: Preparation and supervision staff costs in this project (LIHF) and two other projects Preparation Supervision/ICR Ratio of preparation cost Project Time (A) Cost (B) Time (C) Cost (D) to supervision cost (B / D) India TEQIP II (P102549) 318895 132 2184078 0.15 Kosi (P122096) 39 231755 211 817203 0.28 This project: India LIHF (P119039) 124 761535 147 806318 0.94 Source for TEQIP II and Kosi: ICRs Page 43 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS Verbatim comments from the NHB (email): Subject: Implementation completion and results report – World Bank – LIH - NHB Please refer to your letter dated August 20, 2019 wherein you have forwarded the above captioned report. In this regard, we hereby state that we have perused the Report and found to be useful. We have no further comments to offer. We appreciate the efforts of World Bank for preparing the Report with key details as well as covering the other aspects with lessons and recommendation, Thanks for sharing the same with us. Page 44 of 45 The World Bank India Low-Income Housing Finance (P119039) ANNEX 6. SUPPORTING DOCUMENTS 1. Project Appraisal Document (April 2013) 2. Financing Agreement (August 2013) 3. Project Agreement (August 2013) 4. ISR Aide Memoire (Oct 2013) 5. ISR Aide Memoire (April 2014) 6. ISR Aide Memoire (November 2014) 7. ISR Aide Memoire (May 2015) 8. ISR Aide Memoire (July 2015) 9. ISR Aide Memoire (Feb 2016) 10. ISR Aide Memoire (April 2016) 11. ISR Aide Memoire (November 2016) 12. ISR Aide Memoire (May 2017) 13. ISR Aide Memoire (December 2017) 14. ISR Aide Memoire (May 2018) 15. ISR Aide Memoire (December 2018) Page 45 of 45