Opening-Up the Lower-Middle Income Housing Market in South Africa: The Role of Demand-Side Subsidies Marja C Hoek-Smit and Liza Cirolia A Report for the Department of the National Treasury City Support Program Republic of South Africa Table of Contents EXECUTIVE SUMMARY .........................................................................................................................................2 1. OBJECTIVE ....................................................................................................................................... 7 2. CONTEXT ........................................................................................................................................... 8 3. EXPERIENCE WITH DEMAND-SIDE SUBSIDIES ..................................................................... 9 4. WHY DEMAND-SIDE SUBSIDIES? ............................................................................................. 12 5. CURRENT SUPPLY CONSTRAINTS TO A DEMAND-SIDE SUBSIDY APPROACH ........ 16 6. POSSIBLE NEXT STEPS ................................................................................................................ 26 LIST OF TABLES 1. Budget and Total FLISP Subsidies Approved by Province and Year…………………………………………………………….12 1 Executive Summary This report provides preliminary findings of a consultancy about the perceived need for finance- linked demand-side household subsidies to expand urban homeownership in South Africa for the income group between R3,500 and R15,000/month (often referred to as the ‘affordable market’) and the current hindrances to their implementation.1 The main question the consultancy sought to answer was whether a well-designed demand-side subsidy can work to expand the affordable housing market in South Africa. Our analysis suggests that there is capacity and an interest in this market segment in the private sector, and that a demand-side subsidy linked to credit is considered a critical requirement to open up this market. It also suggests that current inefficiencies in the affordable housing market can and should be addressed by better policies and regulations before a demand-side subsidy can be effective (e.g. more efficient and less costly housing standards and permitting procedures, improved infrastructure systems, and improved financial sector regulations and systems to mitigate risks and costs of lending for the affordable market). Such agenda will require comprehensive policy and regulatory reforms that go beyond the demand-side subsidy itself. It will also require a robust administrative and operational system for subsidy implementation. Different parts of such agenda are already being discussed. Reasons to Address South Africa’s Affordable Market There are approximately 3 to 3.5 million urban households within the R3, 500 and R15, 000/month income segment. With an urban household growth rate of 2 % per year (World Bank 2014), and assuming a similar income distribution, at least 60,000 to 70,000 new housing solutions will be needed to accommodate new households and stop the increase of the housing backlog for this income segment. Supply for this market is heavily constrained. South Africa has witnessed a substantial decline in the private construction of affordable units. The private sector produced approximately 6,500 new urban units per annum over the past decade for households with incomes below R15,000, not including the RDP/BNG housing. Lack of affordability was one of the reasons - the average cost of a new-build affordable unit was over R380,000 in 2015. The potentially huge resale market for this income segment did not fill the gap despite much greater affordability, because of issues with titling and sales restrictions on township and subsidized (RDP/BNG) units. With a combination of savings, credit and a relatively small demand-side subsidy, these prices could be affordable for a large proportion of the target market. 1The below R3,500/month income segment benefits from the large-scale RDP/BNG program, which provides fully subsidized ownership housing for free to qualifying households. More than 3 million units have been supplied since 1996. 2 Benefits of Using Demand Side-Subsidies A demand-side housing subsidy is a transfer made to a household with the specific purpose of increasing their willingness and ability to consume better housing or housing of a particular type. Demand-side subsidies can: broaden housing choices for the consumer and address the varied housing needs and constraints; leverage households’ own contributions through credit and savings and, therefore, reduce the subsidy amount per household; be applied by local authorities to incentivize the provision of different types of housing suitable for their housing markets; be applied to households acquiring houses in the secondary market; and gradually move the private sector down-market by targeted households at the margins. However, demand-side subsidies are only effective if supply markets can respond to the increased demand that the subsidy program unleashes. Past Efforts to Address South Africa’s Affordable Housing Market Since the early 1990s, many efforts have been made in South Africa to address the affordable market with and without the use of demand-side subsidies. Most notably, the 2004 Financial Services Charter (FSC) included the commitment to invest R42 billion over 10 years towards stimulating the provision of and access to housing for households in the affordable market. In this context, the Department of Housing (now called the Department of Human Settlements) collaborated with the Banking Council to formulate the first “Finance Linked Individual Subsidy Programme� (FLISP) in 2005, which provided down-payment support. The FSC showed that many lower-income households were creditworthy, and could handle repayments on long-term mortgage loans. It also showed that financial institutions could make such loans profitably and deal with the modestly higher credit risk. This was achieved with a minimal application of the FLISP subsidy, which lenders preferred not to use because of potential limitations on foreclosure and the perception that the subsidy could impact borrower repayment behavior. Recognizing some of the design challenges of FLISP, the National Department of Human Settlement (NDHS) launched a revised finance linked subsidy program in 2012 that: i) set income qualification at R3,501-R15,000 per month; ii) established a maximum house-price of R300,000 (removed subsequently); iii) allowed both newly constructed houses and resale houses to be included; iv) increased the maximum subsidy for the lowest income segment to R87,000; v) removed the savings/down-payment requirement; vi) gave households below the R7,000 income level the option to apply for a free serviced stand; vii) appointed the National Housing Finance Corporation (NHFC) as program administrator; and, viii) developed nine nationally driven FLISP delivery projects. These changes reflect NDHS’s understanding of the need for a flexible subsidy product and efficient administration. The uptake of the new FLISP has, however, been extremely low yet again. A total of 2,793 subsidies were approved between April 2012 and March 2015. 3 Moreover, the banks use the current subsidy simply as an additional credit risk-insurance – allocating subsidies to borrowers who already qualify for a loan of sufficient size to pay for the house – rather than using the subsidy to deepen or expand the affordable mortgage market. One of the reasons for this poor performance is that no national level awareness or publicity campaign was ever conducted, nor were the necessary administrative systems put in place. However, our research found that there are more fundamental supply-side constraints that hinder FLISP’s successful implementation. Understanding the Systemic Constraints and Disincentives A core reason for the failure of FLISP is that the main parties that should drive affordable housing and loan supply lack the incentives to do so. Below is an explanation of why three key actors, municipalities, developers, and banks, have little incentive to expand this market. 1. Municipalities are hesitant to expand the fully-subsidized RDP/BNG and affordable housing sectors because of budget implications, misaligned incentives, and real capacity and other constraints. The burden of having to service new housing areas without being able to collect rates, is an example of budget and incentive challenges. There is a political component as well, reflected in the overwhelming perception that housing delivery targets can only be met through new building rather than through the secondary market. Subsidy beneficiaries who sell their homes are believed to return to inadequate housing and informal settlements. The result is a lack of new projects, and little support for titling programs that might expand secondary market development in the affordable segment. 2. Developers see a great market opportunity in the affordable sector. However, high housing costs relative to incomes (worsened by price pressures from the free-housing RDP/BNG segment), and bureaucratic uncertainties and delays to obtain permits make this market only feasible at a very large scale and for a few large players who benefit from gaining access to public land and infrastructure grants. Developers avoid building in municipalities with poor regulatory and permitting records despite high demand for affordable housing. A further constraint is the lack of easily accessible mortgage finance for their customers and the dearth of creditworthy applicants. 3. Banks face real risks and costs constraints, but are currently the only type of financial institution allowed to participate in the program. While they profess to be interested in expanding the mortgage market to this potentially large segment of the population, banks consider the risks and costs of moving into the subsidized affordable market too high: i) transaction costs of underwriting and servicing lower-income applicants are high; ii) credit risk is higher for this segment relative to the conventional market - both the probability of default and the loss given default; iii) expanding long-term lending increases funding mismatch in banks that are predominantly deposit funded, and funding costs if funding sources other than deposits need to be utilized; iv) regulatory requirements and costs are relatively high for mortgage lending. These constraints result in the lack of drive to open up 4 the affordable sector and in the “tacit� agreement amongst banks to use the FLISP to reduce credit risk of currently qualifying borrowers rather than deepening the mortgage market. The analysis provides detailed recommendations for actions at both the national and the sub- national (provincial and local level) to address the supply constraints experienced by the principle actors in Section 5. The Way Forward The consultancy recommends the following next steps to implement the CSP demand-side subsidy project: 1. Reach consensus among stakeholders to detail and prioritize the specific actions required to alleviate constraints in each of the supply-side and administrative areas. 2. Identify leading institutions and individuals to spearhead the reform actions in each main area - administrative reforms, including how municipalities can play a role in implementing demand-side subsidy programs, real supply side issues at the municipal and national level and national financial sector reforms. 3. Decide on the next phase of the CSP demand-side subsidy project. We recommend a two- pronged approach: a. Provide support to improve national demand-side programs in collaboration with NDHS that includes comprehensive procedural and operational reforms and addressing identified supply bottlenecks. This includes support both for the new Human Settlements White Paper process and for revisions to the Housing Code. b. Design CSP pilots in selected municipalities that focus on local level improvements in the supply-chain for new and existing affordable housing and link these to an improved and revamped set of national demand-side programs (for rental and ownership housing, new and existing housing, finished units and home- improvement). Options Description Inputs and Outcomes Comprehensive Providing support to improve national • Will result in structural redesign of demand-side programs that include i) improvement, efficiency national comprehensive legal, procedural and and expansion of the demand-side operational reforms and specification of affordable housing market programs different applications of demand-side and greater efficiency of subsidies, ii) addressing identified real the subsidy program. supply bottlenecks with NDHS, NHBC, NT, • Requires collaboration of municipalities; iii) addressing lending NDHS and several constraints with NT, RBSA, banks and non- national and local entities. bank FIs, iv) developing consumer debt 5 repair & savings programs and other • May require external TA household support and public outreach and possible funding for programs with NDHS, NT, Municipalities, development of NGOs. administrative and operational systems. City pilots Designing CSP pilots in selected • Will provide important linked to municipalities that focus on local level lessons for the role of national improvements in the supply-chain and link municipalities in the redesign of these to an improved and revamped set of successful demand-side programs national demand-side programs, including implementation of a national level changes to expand mortgage diverse set of demand- lending and borrowing for the target groups. side subsidy programs, while avoiding the cost of stand-alone pilots. • Option depends on the appetite for reforms at the national level. 6 1. Objective This report provides preliminary findings of a consultancy about the perceived need for demand- side (and finance-linked) household subsidies to expand urban homeownership in South Africa for the income group between R3,500 and R15,000/month (often referred to as the ‘affordable market’). Specifically, it addresses the limited take up of the “Finance Linked Individual Subsidy Programme� (FLISP) that is currently in place. The consultancy is carried out for the City Support Program (CSP) of the National Treasury (NT), Government of South Africa and is focused on the needs of the metropolitan municipalities with which CSP works. The premise of the analysis is that demand-side subsidies have the potential to greatly improve the housing situation in South African cities but that existing programs have not been sufficiently explored to address the varied requirements for housing in this income bracket. In particular, the FLISP instrument has remained underutilized. It is considered likely that the lack of successful outcomes may be due to specific demand and supply constraints for which solutions can be found through a concerted effort by core parties – national government, local government, lenders, developers and realtors. It is specifically hypothesized that major blockages and reasons for lack of take-up have their roots at the municipal/metro level. This consultancy analyzed the deterrents to the effective implementation of demand-side approaches and engaged in a process to define workable solutions. At the same time, the consultancy looked into possible constraints in the administration of the current FLISP program at the national and provincial levels, and analyzed structural issues that hamper the use of FLISP. These constraints were discussed with the major stakeholders during the second and third missions. The CSP may, based on the outcomes of this consultancy, consider several pilot projects of demand-side subsidy approaches in selected municipalities. However, since the constraints are both national and local in nature, careful thought has to be given to the structuring of potential pilots. The focus on demand-side subsidies for the affordable market has several reasons. First, the unfulfilled housing demand in this large segment of the market is growing rather alarmingly and has many negative effects on households and the broader urban environment. Second, this segment holds an enormous economic opportunity if it can be unlocked. Households in this income bracket can pay towards their own housing solution, but require some assistance to access the formal housing market. Third, demand-side subsidies that leverage households’ own contributions to solve their housing problems, whether through credit, savings and/or labor, are fiscally more efficient than supply-side subsidies. Fourth, demand-side subsidies allow for great flexibility in the type of housing options that can be supported (rental and ownership, new and existing, finished and home-improvement) which is important for households and municipal governments alike. CSP representatives and consultants visited four metros/municipalities (Ekurhuleni, Tshwane, Cape Town, and Mangaung) and their Provincial housing offices and met with senior government officials from the National Treasury, the National Department of Human Settlement, the National Housing Finance Corporation which administers the FLISP, the Housing Development Agency, the Estate Agency Affairs Board, representatives of the Non-Governmental sector, representatives of the four main banks and the South Africa Banking Association and other housing finance professionals, and with developers involved in the affordable housing market. This report gives the results of the analysis of the data and information from two consulting visits and provides preliminary recommendations. 7 The consultants want to give special thanks to Yasmin Coovadia of the CSP, who was an integral part of the team and provided the consultants with all necessary background information, prepared the visits and interviews and made herself available throughout. 2. Context The productive, inclusive and sustainable development of South Africa’s largest cities is increasingly critical to prospects for national economic growth and poverty reduction. The United Nations estimates that over 71% of South Africans will live in cities by 2030, and nearly 80% by 2050. South African cities are responsible for the majority of national economic activity. Metropolitan municipalities account for a rising share of this activity, hosting over 62% of economic activity and nearly 58% of all job opportunities. Despite their importance, cities do not capture the benefits of their growing social and economic significance. The National Treasury’s Cities Support Programme works with cities to enhance their expertise in core areas that will make cities more efficient. The objective of the Human Settlements Support (HSS) component of the Cities Support Programme is to assist cities in their efforts to ensure the availability of affordable accommodation at scale for all residents, within more inclusionary, integrated and efficient human settlements. This consultancy forms part of the CSP’s goal to assist cities in “Improving the performance of fiscal instruments that can “incentivise the private sector to increase participation in affordable housing in better-located areas, and pilot demand side subsidy instruments.� South Africa’s large-scale supply-driven housing subsidy programme (which has produced over three million fully subsidised housing units) has focused on households at the bottom of the income distribution, i.e., incomes of R3,500 or less per month. Housing policy has, however, given insufficient attention to the affordable housing market made up of urban households with incomes between R3,500 to R15,000, for whom housing and housing finance markets also do not work. There are approximately 3 to 3.5 million urban households within this income segment, and at an urban household growth rate of 2 % per year (World Bank 2014), a total of 60,000 to 70,000 new housing solutions are needed to accommodate new household formation and stop the increase of the housing backlog for this income segment. At the same time, the country has witnessed a substantial decline in the private construction for this market. The private sector produced approximately 6,500 units per annum in urban areas over the past decade for households with incomes below R15,000.2 As an example, the metro of Ekurhuleni projects that by 2019 they need 320,000 new units for this income bracket, of which half to accommodate new households and the other half to resettle households from informal settlements, while there is currently hardly any affordable stock produced or even in the project pipeline. The challenge of supplying affordable units is exacerbated by constraints in the resale market for affordable housing, including a backlog in titling and sales restrictions on township and subsidized units. Yet the resale market has the potential to offer housing that is more affordable for the lower income segment than new housing. 2The private sector has constructed just over one million units in the last twenty one years (excluding RDP/BNG) and delivery has been on the decline with as few as 26,000 units constructed annually in recent years (ABSA 2015). 8 In 2015, CAHF found that the cheapest new build house in South Africa was R R370 000, (CAHF, 2015)3. In the same report it is noted that in the formal resale market, RDP/BNG housing costs between R70 000 – R250 0004. With a combination of savings, credit and a relatively small demand- side subsidy, these prices are affordable for the target market.5 Indeed, the finance-linked demand- side subsidy program (FLISP) was meant to increase the private housing supply for this market segment and expand the secondary market for affordable housing. However, given the high cost of mortgage origination and lending, it was always considered a challenge to extend the program to the lower part of the targeted income group and the delivery of FLISP subsidies has remained limited (CAHF, 2015) Attempts to address the challenges of the affordable housing market have received considerable housing policy attention and South Africa’s history is replete with examples of efforts to enroll the private sector in the provision of affordable housing and housing credit, in particular for those unable to qualify for a fully subsidized unit.6 The following section briefly outlines this history. 3. Experience with Demand-side Subsidies Since 1995, various national and local initiatives were introduced to stimulate effective demand of the lower-middle income groups, i.e., households who need a small subsidy to supplement their own savings, finance or sweat equity.7 At the time, it was argued that households earning between R1,500 and R3,500 were only in need of a partial subsidy. The demand-side subsidy instruments inherited from the previous dispensation, such as the first-time home buyer interest subsidy scheme and the state-assisted home-owner saving scheme, were phased out. Between 1995 and 2000, new demand-side subsidies were introduced, including the rural voucher subsidy, credit and non-credit linked individual subsidies, the consolidation subsidy and subsidies allocated through the Peoples Housing Process (PHP). The first two are truly individual demand-side household subsidies that allocate funds (or a debit card that has to be utilized for the purchase of building materials) to owner-builders who already have their own site (e.g., a rural plot of land, a site in an original site-and-services scheme, or a regularized plot in an upgrading area), or as in the case of the individual subsidy, the household has found a qualifying finished house and applies for a subsidy. Other such subsidies, like the Consolidation Subsidy and the Enhanced PHP (EPHP), use “demand organizers� in the form of NGOs that manage the subsidy funds and assist and supervise owner-builders. 3 Centre for Affordable Housing Finance in Africa (2015) “Understanding the challenges in South Africa’s Gap Housing Market and opportunities for the RDP Resale Market�. Page 3. 4 Ibid. page 8. 5 CAHF (2015) estimated that a households earning R3,501 could qualify for a loan of approximately R83,000. With a current FLISP subsidy of R87,000, the households would be able to afford a unit of R170,000 even without a down-payment. Households earning R15,000 could qualify for a loan of approximately R357, 000. 5 With a FLISP subsidy of R20,000, the household would be able to purchase a unit of R377,000 without a down-payment. This calculation assumes an amortization period of 20 years, an interest rate of 11.25% and a payment of 25% of a household’s monthly income. It also assumes that households do not have major credit indebtedness issues. With prime interest rates currently (2016) at 10.5%, these numbers will have shifted. 6 The Financial and Fiscal Commission (FFC) (2013) recommended that “government should consider investing more resources in funding or scenarios that are likely to stimulate additional funding from the private sector (banks and private developers) as well as household contributions towards housing delivery. These include investment incentives using tax rebates, upgrading of backyard rentals (with incentive), and housing vouchers� 7 The nominal income figure of R3,500 has not changed since 1995 and as a result the real income group qualifying for subsidized housing is at a much lower income percentile of the population. 9 Some of these programs were credit linked, requiring households to access various forms of finance. Most were owner-builder programs and required that households already have access to land or a serviced site. Assessments of the individual consolidation or voucher subsidies showed that in cases of old site-and- services schemes, where beneficiaries already have a core house on a serviced lot, these voucher schemes have been quite successful (Forster and Gardner, 2014). However, in the individual subsidy programs, studies found that funds were often used for non-housing purposes, e.g. to start businesses, pay school fees etc. defeating the public objective of improving housing consumption8. Indeed these programs have been difficult to supervise. Over this same period, a Mortgage Indemnity Fund, aimed at de-risking lending in township and subsidized housing areas, was established and dismantled (Tomlinson, 2002). While the fund did support increases in lending in township areas, high default rates put extreme pressure on the state. A major private financial sector driven effort to expand housing finance availability for the latter market segment is important in this context. The Financial Services Charter (FSC) of 2004 was a “voluntary� agreement by the financial sector to expand access to financial services to excluded population groups. The FSC included the commitment to invest R42 billion over 10 years towards stimulating provision of and access to housing for households earning between R1,500 and R7,500 (with provision for an annual increase in line with the South African Consumer Price Index)9. In this context, the Department of Housing collaborated with the Banking Council to formulate the first finance linked individual subsidy program in 2005. The program provided down-payment support ranging from R3,369 to R23,584 (depending on income) for qualifying beneficiaries. In response to this program, a number of banks launched affordable housing finance products for households earning less than R7,500 per month. Interestingly, while the FSC commitment was successfully fulfilled well within the set time-line, the FLISP remained largely unimplemented. Banks rather followed their own programs (mortgages for the higher end of the FSC income bracket, pension-backed loans and unsecured loans for smaller loan amounts), without relying on FLISP subsidies. The main reason for this lack of take up was the provision in the Housing Act (2001 Amendment) that disallows the sale of a subsidized unit for a period of 8 years, making repossession of units legally contestable. In addition, it was feared that loans linked to a government subsidy, might negatively impact repayment behavior, and that administrative procedures were inefficient. The FSC showed, however, that many lower-income households were creditworthy, and could handle repayments on long-term mortgage loans, and that financial institutions could make such loans profitably and deal with the modestly higher credit risk.10 The FSC has since been incorporated into the 2013 Financial Sector Code, which applies to the entire financial sector and provides financial institutions points based on the extent to which they serve low-income customers (incomes are indexed and the maximum income to which the Code applies is currently R18,500 per month; just above the current FLISP income cut-off of R15,000 per month). The points are translated into a BBBEE score for each institution and government business is allocated to those with acceptable scores. The code provides, 8 See the Public Service Commission’s 2003 Evaluation of the National Housing Subsidy Scheme. 9 The Banking Association of South Africa claims that the financial sector has exceeded its FSC goals, originating almost four hundred thousand loans worth over seventy billion between 2004 and 2014. 10 FSC mortgage loans during those early years did not display dramatically higher default rates than mortgages for the non-FSC market; loan performance has since deteriorated somewhat relative to non-FSC loans, and NPLs are noticeably higher for loans written since 2009 (Melzer 2009, 2015). 10 therefore, an incentive for financial institutions to expand lending into the affordable housing market, although no specific housing targets are set. The number of mortgage loans made to the affordable sector has not noticeably increased, however, and appears to remain in the 25,000 to 30,000 per year range, just enough to show some commitment on the side of the banks, but of a scale that is not related to needs in the affordable housing sector. To support demand in the affordable market, the National Department of Human Settlement launched a revised finance linked individual subsidy program in 2012 that increased the minimum qualifying criteria to R3, 501 and maximum qualifying income to R15,000 per month, set the maximum house-price allowed under the program at R300,000, allowed the subsidy to be applied to newly constructed houses and resale houses, increased the maximum subsidy for the lowest income segment to R87,000, and removed the savings/down-payment requirement. In addition, households with incomes below R7,000 per month were given the option to apply for a free serviced stand under an Integrated Residential Development Program in lieu of the demand subsidy linked to a mortgage loan. Administrative constraints were addressed by bringing on board the National Housing Finance Corporation (NHFC) to administer the subsidy program.11 In a further adjustment to improve the flexibility of the program, the house-price ceiling of R300 000 was recently removed and the allowable house price became a function of the “affordable� loan size and subsidy amount. Nine nationally driven FLISP delivery projects were launched. While there appears to be widespread agreement amongst developers, lenders and local government officials that a demand-side subsidy program linked to credit is critical for large scale housing delivery to this market segment, uptake of the new FLISP has been extremely low yet again. Table 1 (below) shows that a total of 2,793 subsidies has been approved between April 2012 and March 2015, for a total amount of approximately R140 million at an average subsidy amount of R50, 000.12 It is not possible to compare this spending against a total FLISP budget.13 No figures have been compiled on the proportion of FLISP subsidies used for new construction or resale houses, but spot checks show that while most subsidies are used for new housing and are processed through developers, FLISP also serves the secondary market through real estate brokers. 11 Provinces have to agree to release the FLISP part of their Human Settlements Development Grant (HSDG) to NHFC for centralized disbursement. Only five out of the nine Provinces has done so. 12 Approvals and committed budget amounts can be used as proxies for disbursement, since most of the approvals were disbursed. Disbursements that happened in another financial year from approvals were excluded to avoid ‘double counting’. 13 It is difficult to ascertain the total budget which has been allocated for FLISP over the past three years. This is partly due to provinces’ continual shifting of spending between housing programs to ensure that budget is spent. The National Treasury’s 2013 Estimates of National Expenditure and 2014 Estimates of National Expenditure (Vote 31: Human Settlements), do not explicitly discuss FLISP. It is therefore not possible to get nationally compiled budgets for FLISP. 11 Table 1: Budget and Total FLISP Subsidies Approved by Province and Year Total delivery by FY 2012/13 FY 2013/14 FY2014/15 province PROVINC # VALUE # VALUE # VALUE # E EC 5 R 183 225 260 R 12 258 882 324 R 16 718 400 589 GP 28 R 1 265 925 755 R 31 912 032 420 R 17 390 100 1203 KZN 28 R 1 766 625 155 R 8 457 225 207 R 11 165 580 390 LP 32 R 1 291 950 87 R 4 704 750 58 R 3 088 500 177 MP 0 R0 1 R 44 475 8 R 374 312 9 NC 0 R0 1 R 68 100 0 R0 1 WC 118 R 6 980 850 203 R 11 423 850 168 R 8 729 448 489 NW 7 R 179 454 3 R 122 625 0 R0 10 218 R 11 668 029 1478 R 69 572 139 1195 R 57 951 060 2891 The total FLISP budget allocated for all years and all provinces FY 12/13-FY 14-15 =R 139 191 228 The current FLISP delivery targets under the Medium Term Strategic Framework (2014-2019) aim at a disbursement of 70,000 FLISP subsidies as part of Minister Sisulu’s goal of providing 1.5 million houses by 2019. These targets would require drastically upscaling the provision of FLISP. The questions we will address in the next sections relate to the reasons for the persistent poor performance of the demand-side, finance linked subsidy programs for this critical market segment. Several studies have been conducted on the programs which our analysis tries to integrate with the aim of creating a comprehensive evaluation of issues and priorities for reform. We begin the analysis with a brief discussion of the reasons for South Africa to move to demand-side subsidy programs and the general requirements for their effective design based on international experience. We assess the experience with FLISP against that background. Lastly we make recommendations for a different way of implementing demand-side subsidies at both the municipal and the national level. 4. Why Demand-side Subsidies? A demand-side housing subsidy is a transfer made to a household with the specific purpose of increasing their willingness and ability to consume better housing or housing of a particular type. Increasing the demand for housing can be done through housing allowances or housing vouchers for rental or owner-occupied housing, through up-front grants tied to housing finance or savings for housing, or through interest rate subsidies or tax-benefits that lower the effective recurring cost of housing payments. Finance-linked upfront subsidies, which are considered the most efficient and equitable, can be applied to closing costs, the down payment, the premium for private mortgage insurance, or the loan amount. Since the beneficiary makes the choice to buy or rent a particular house, demand-side subsidies are considered more efficient at improving a household’s satisfaction with their housing situation than supply-side subsidies, which are instead linked to specific subsidized housing solutions or implicit subsidies to the finance sector of which beneficiaries may not even be aware. Also, supply-side subsidies to developers or through the financial system are known to distort housing and finance markets, and in case of developer subsidies, these are typically only partially transferred to the beneficiary in the form of better housing products or lower prices. Moreover, suppliers will limit production to the number of available subsidies, rather than explore the opportunities offered by a growing lower-middle income 12 market. For these reasons, many countries have shifted away from supply-side subsidies and have chosen an up-front grant or other demand-side subsidy system tied to the individual or household. In general, supply-side subsidies are efficient only when input markets do not work well and do not respond to regulatory or policy incentives to deliver specific types of housing or housing finance. This was, for example, the case for the market segment for the lowest income group in South Africa, after 1994. However, there are now many reasons why a shift to demand-side subsidies should be pursued, at least for the affordable market segment. The first reason for South Africa, and many other countries, to move to demand-side subsidies and away from supply-side approaches is the need to broaden housing choices for the consumer and address the varied housing needs of those households who need partial state assistance to improve their housing situation.14 In addition, such subsidies can be applied by local authorities to incentivize different types of housing such as smaller infill developments, existing lower middle income housing markets in low-income areas, experimental housing that allows for expansion of units for rental housing, employer-supported housing, etc. rather than just large-scale new construction. The housing demand profiles of households currently in the affordable market segment are diverse and not just in levels of income and type of employment. It is a younger population, with smaller household sizes and different housing needs over the household life-cycle. They operate in a more unstable labor market and in a low economic growth environment, but with considerable pressure on house-prices. It is imperative that they are able to move to employment opportunities, or to schooling opportunities for their children, and different house-types in different stages of life. Demand-side subsidies offer a wider choice of locations and house-types to beneficiaries and help create mobility in the housing market. Households can choose to buy an existing house in a neighborhood of their choice or buy a new house if they can afford to do so, increasing the range of households that can become part of the formal ownership market.15 The bulk of the affordable housing stock in South African cities is found in ‘RDP/BNG neighborhoods’ and older townships. Demand-side subsidies can be applied to households acquiring houses in the secondary market, which are in general of a lower price than newly constructed houses of a similar quality. Households selling their property could “filter up� to housing of a better quality because of the equity they built up in their existing house (see State Subsidized Housing as Assets, NDHS and The Presidency, 2014). Importantly, selling households could move to newly constructed housing if such housing becomes available. They may do so with or without subsidies. The important issue here is that by subsidizing lower-income households to buy existing houses in neighborhoods of their choice, new construction of houses of a higher price will be stimulated in response to increased demand by the selling households and a more dynamic and efficient housing market will develop.16 This filtering model is more likely to work in South Africa than in most other emerging market countries with 14 Outcome 8 states the “need to diversify our approach to include alternative development and delivery strategies, methodologies and products including upgrading of informal settlements, increasing rental stock, and promoting and improving access to housing opportunities in the gap market�. 15 Demand-side subsidies may, at a later stage, be extended to rental housing. 16 The impact measurement of demand-side subsidies should therefore be the number of households assisted, rather than new units constructed, which is a supply-side measurement. When the affordable housing market begins to work well with the help of flexible demand-side subsidies, the subsidies provided for the resale market should translate in increased new construction and a gradual decrease in new informal housing. These effects should be monitored and evaluated. 13 an urbanizing population because of the enormous investment in low-income housing in the past that can be included in the formal market. The second reason to move to a demand-side subsidy approach is the efficiency these subsidies afford to leverage households’ own contributions through credit and savings and, therefore, reduce the subsidy amount per household. Furthermore, they can address different types of household constraints. Households in the affordable market segment can and are willing to pay for housing that is affordable and in the right locations (see above), but they often require support to access credit (come up with a down-payment and pay for upfront costs, receive assistance to repair their credit, or deal with uneven income flows because of informal employment) or to pay for part of the house in areas where house-prices are high relative to incomes. As is done in South Africa, demand-side subsidies can be designed to decrease with higher incomes to improve equitability and avoid cliff effects17, and they can be more easily adjusted when costs of construction or cost of credit change. Overall outlays by the state are, therefore, generally lower than in the case of supply-side subsidies for similar housing products. The third reason to move to demand-side subsidies is that by using private sector suppliers and subsidies that effectively target households at the margin of the different sub-markets, the private sector can gradually move down-market. The subsidy will allow suppliers of credit and housing, and estate agents in the secondary market to develop the systems and know-how to serve a successively lower income segment and serve an increasing number of households that were previously outside the reach of the market, opening up enormous economic opportunities. However, demand-side subsidies are only effective if supply markets can respond to the increased demand that the subsidy program unleashes. When demand-side subsidies are applied in an environment where developers or lenders are unable or unwilling to respond to the demand expansion for certain housing or loan products, the beneficiary’s choice remains limited and the “so-called� demand- side subsidy will effectively become a supply-side subsidy for developers and/or lenders. This has been the case in South Africa where supply elasticity in the affordable market is very low, both on the developer side and on the credit side. FLISP has not attracted large numbers of developers to the affordable market segment, since supply constraints make this market not very profitable. A few large developers control the subsidy allocation and incorporate the subsidy amount (and other subsidies) into their pricing model of affordable units. The same conclusion can be drawn for the credit side. The four banks that are allowed to use FLISP, use the subsidy as an additional risk-insurance for borrowers who already qualified for a maximum affordable loan of sufficient size to pay for the house they bought. Once the FLISP subsidy is disbursed it is simply used to reduce the approved, affordable outstanding loan amount, and therefore reduces the LTV and credit risk for the lender.18 The FLISP subsidy does, therefore, currently not deepen or expand the affordable mortgage market to include currently non-qualifying or lower income households, and the number of mortgages made to the affordable segment has remained static for a decade. Of course, the subsidy does not solve all major demand-side constraints households face, only that of low-income relative to house-price. Many households are over-indebted, have poor credit scores, 17 This cliff effect refers to the drop in subsidy amount between qualifying and non-qualifying households when hard income brackets are used, e.g., the difference between a free house for households with incomes R3,500. 18 A mortgage insurance product could do this much more cheaply (see section 4). 14 or work in informal employment with uneven income streams. For these reasons they will not qualify for a mortgage loan and will require different financial products and support. We return to these supply and demand constraints in section 5. The main question we tried to answer during this consultancy was, therefore, whether a demand- side subsidy with the advantages described above, can work to expand the affordable housing market in South Africa given the supply-side constraints, or whether, supply-side subsidies to developers or lenders might be the only way to address the housing needs of this market segment. Our analysis suggests that there is capacity and an interest in this market segment in the private sector. It also suggests that inefficiencies in the affordable housing market can and should be addressed by better policies and regulations, e.g. more efficient and less costly standards and permitting procedures, improved infrastructure markets, and improved risk mitigation institutions and increased competition in the finance sector. These interventions stand in contrast to subsidizing the higher costs resulting from supply inefficiencies -as supply-side subsidies do. This approach will require a more comprehensive policy agenda that goes beyond the demand-side subsidy itself and includes policy and procedural reforms on the supply side. Different parts of such agenda are already being discussed. The other question we focused on was how to improve the current inefficient administrative procedures to make demand-side subsidies feasible. Demand-side subsidies require fundamentally different types of operational and administrative procedures and systems than supply-side subsidies that focus on developers and lenders. Demand-side subsidy programs have to assess and approve individual households and provide them with a “voucher� on the basis of which they can approach a lender of their choice who can “underwrite� the household for a loan. For a demand-side subsidy to work the following operational procedures, information and administrative systems need to be in place: 1. Clear, unambiguous operational rules and procedures on how the subsidy works and is to be applied, including:19 a. Borrower eligibility b. Purchase and loan processing c. Mortgage agreement, Loan closure, Subsidy disbursement d. Loan servicing (procedures in case of delinquency, early payment, loan transfer) 2. Standard administrative procedures, including: a. Borrowers identification and validation b. House/unit selection (qualified programs) c. Lender selection d. Registration of new lenders under the scheme (master agreement) 3. Local/national level publicity and awareness campaigns and publicly available information systems about the various subsidy options, qualification criteria, tools for self-evaluation, etc. 4. Effective pre-screening systems (including computer-based self-screening programs), timely pre- qualification mechanism of applicants using a transparent system of prioritization when the application pool reaches scale 19For example, the current FLISP description and application are unclear. “The mortgage -linked subsidy is a capital grant that can be applied to lower the loan amount in order to make the monthly loan repayments “affordable� over the repayment term; and/or, augments the shortfall between the “qualifying� loan amount and the total house -price. The terms “qualifying� and “affordable� are not specified and have given rise to confusing interpretations on how to apply the subsidy. 15 5. National level unified subsidy data-base that can be checked for previous subsidy allocations and efficient national administrative system 6. Real time electronic communication system with lenders on case files/ approvals/qualifying subsidy amounts/loan performance, etc. 7. Master agreement with lenders 8. Timely (at time of mortgage closure) disbursement procedures of subsidy amount to lenders/developers. 9. Monitoring and evaluation systems that can be used to make regular subsidy program adjustments. None of these administrative prerequisites currently exists in South Africa. There is lack of clarity in the respective roles of the Province (which holds the budget) and the NHFC. NHFC as the national administrator does not have the budget or incentives to set up the required operational, administrative and publicity/ information systems. Under current conditions, it would be unfeasible to scale up the current FLISP program. In addition, the current program ignores the critical role of municipalities in information provision about housing programs and in program administration. Housing markets are local. Municipalities are the first contact for residents’ housing queries, they maintain waiting lists, understand the local housing market, both for new and existing houses, and know the local players in the real estate market. They have relationships with local employers and other parties who can play a role in the demand-side subsidy programs. In conclusion, the efficiencies and increased flexibility that demand-side subsidies can offer in the South African housing market, are critically important. Past supply-side approaches to subsidize housing have shown their limitations for all but the poorest segments of the market. However, the success of demand- side subsidies hinges on the existence of responsive supply-side systems both in development and housing finance, and on an effective subsidy administration system. All three systems need major reforms. 5. Current Supply Constraints to a Demand-side Subsidy Approach A fundamental problem in the implementation of the current FLISP program is that the main parties that should drive this program and assist in expanding the affordable housing market through the use of demand-side subsidies, lack the incentives to do so. Neither municipalities/ metros, nor private sector agents such as developers, real estate agents and lenders, see much benefit in expanding the production and resale of affordable housing or increasing lending to the sector given current conditions. In addition, the general public lacks information on FLISP and many income-qualifying households face credit constraints and they would for those reasons not even apply. What can be done to motivate the critical implementing partners to deliver housing and credit at scale for the affordable segment when demand-side subsidies are available to make it feasible to reach this market? Where to start? Based on existing studies and our own analysis of the core supply constraints and “agency� problems, we formulated initial suggestions for policy, administrative and program reforms to make demand-side subsidies work. We received feedback from major stakeholders during the second and third missions and discussed the proposed strategic actions in some detail. In the 16 following sections we summarize our core findings organized according to the perspective of the main implementing entities of demand-side subsidies – municipalities/metros, developers and lenders. • Municipalities and Metros are at the forefront of providing shelter and services for their residents, facilitate the planning and construction of affordable housing and subsequently service these areas. They should perceive demand-side subsidies for the affordable market as a useful tool to reach a segment of their residents hitherto excluded from the housing market. • Developers and Realtors must perceive the affordable housing market as a profitable business to engage with and the demand-side subsidy program as a way to expand their business and achieve scale. • Lenders must weigh the benefits, costs and risks of expanding credit which is linked to demand- side subsidies for a lower-income and more risky segment of the population than currently served by them. The demand-side subsidy should be a powerful incentive to safely expand their mortgage lending operation and increase their client base. In reality each stakeholder faces mayor constraints in their participation in demand-side subsidy programs. We summarize these constraints and provide suggestions for policy and programmatic actions at the national and local level that would improve the incentive structure for each partner. This summary was used as the basis for a workshop with stakeholders to decide on the best way forward which was held in April, 2016. For the sake of completeness, all of the suggestions are listed below. 1. Municipalities are hesitant to expand the RDP/BNG and affordable housing sectors because of budget implications. However, they know that the outcomes of non-action are worse, leading to an increase in the number of informal settlements and backyard dwellings. There is a political component as well. a. New construction: Municipalities/metros have focused on providing RDP/BNG and other projects and lack capacity and budget to expand the affordable market segment themselves. • Most municipalities appear to lack capacity to work directly with developers on mixed or affordable large scale projects and this has made negotiation and approval times unacceptably long. • Municipalities are required to provide land in most cases and use part of their infrastructure grants (USDG in the case of metros) for these projects, while the projects have a low return for them: o Most of the low-income housing stock falls below the cut-off point for taxes/rates (R200,000 is commonly used as a cut off, however it varies between municipalities). Currently, the frequent lack of rating means these properties, do not offer a return on the long-term investment in services and infrastructure for municipalities. In addition, many properties in the RDP/BNG and affordable market lack title in the first place and are not included in the property tax rolls, nor do they pay for services.20 20Although precise numbers of titling backlogs in RDP/BNG neighborhoods are not yet available, the Title Restoration Project estimates that approximately 60 to 65 % of RDP/BNG units received a title over the past 5 years. Conveyancing costs are high and there often is a mismatch between the beneficiary list and occupants at time of registration, which makes conveyancing difficult. 17 o Municipalities must provide many services for free even for those who should pay (an issue of the entitlement/non-payment culture). b. The secondary market: Municipalities lack the incentive to expand the secondary housing market in RDP/BNG and older neighborhoods by expanding titling systems.21 There are many reasons for this including: • Targets and KPIs which municipalities must meet place little value on resale housing. • Provinces pay rates on pre-1994 properties and municipalities receive these taxes without having to do the collection. • Most RDP/BNG houses were below the rate cut off point even if they have title, and collection for those above the cut-off has proven to be hard. • Informal transfers may continue even on titled properties. • Although properties 3,500 and <3,500 monthly income group housing support systems.23 The total subsidy amount across types of housing solutions should be the same in NPV terms for each income bracket. • Apply some maximum standards for new RDP/BNG complete houses (if this program will be continued) in order to prevent the escalation in quality and costs of the free house (Note: in some metros the quality/costs of an RDP/BNG house has been increased with disastrous consequences for the affordable housing market). Possible Actions at Local Level: • Understand the demand for different types of housing tenure and house-types for households on the waiting-lists, for both RDP/BNG and the affordable market. Update waiting-lists and prioritization systems regularly. Align supply initiatives and approvals for applicants in each group with the use of demand-side information. • Introduce changes in the RDP/BNG housing offerings within the municipality at the same time that the demand-side subsidy program is made public and rolled out. 3. Banks give confusing signals. While they profess to be interested in expanding the mortgage market to this potentially large segment of the population, the numbers have remained static. Banks contend that housing supply for the affordable segment has remained limited despite FLISP (see point 2 above), but other constraints including funding costs, credit risk and liquidity constraints, may explain their cautious attitude to mortgage portfolio expansion. Banks are currently the only financial institutions that are qualified to receive the demand-side subsidy linked to housing finance. There should indeed be powerful incentives for banks to expand mortgage lending to this potentially large market segment that is only just opening up: 23 Such schemes for the low-income segment may require additional supply-side incentives for private landlords. 21 • Banks can increase fee and interest revenue considerably.24 • Banks can expand cross selling of other products since the household will stay with the bank for a long time. • Banks receive points and rewards for complying with the Financial Sector Code 2013 (part of Broad Based Black Economic Empowerment) which requires that financial services are extended to households with incomes of R20,000or below25. These are all potential incentives which should drive innovation and competition for customers. However, the way in which banks currently apply the FLISP product (i.e., using the subsidy retroactively to lower already approved and affordable mortgage amounts and therefore lowering their credit risk, rather than moving the mortgage market to increasingly lower income groups) is either an indication that they consider the risks and costs of moving down-market too high or that there is a lack of competition and innovation in that market. Since mortgage lending for the middle market should be a profitable and growing sector for banks and other financial institutions to enter, it is worth exploring some of the possible constraints in that market in some detail. a. Risks and costs on the lending side: • Transaction costs related to loan originations in the FLISP market are exceedingly high. Banks are the entry point for FLISP applicants that have not received pre-approval by the subsidy authorities (NHFC or Provinces). For each qualifying candidate, more than 10 are rejected, because of high indebtedness, credit impairment, or low affordability levels. • Currently no effective credit repair programs are in place and a financial industry wide proposal to fund and implement a comprehensive credit repair/restructuring program was unfortunately stopped by the credit regulator a few years ago. • Banks are liable for credit regulator penalties in case they provide credit to non- qualifying households under the National Credit Act. The affordable market clientele more often has incomplete records or outstanding credit that is not recorded. This becomes a risk for the bank in case of delinquency when records have to be consulted. • The market is prone to political interference, which may affect fee and interest rate levels26, but more importantly, political interference (or fear thereof) and fear of negative publicity currently prevents foreclosures. In addition, the provision that subsidized houses cannot be transferred for a period of 8 years (Housing Act #107 of 1997/ 2001Amendment) creates legal uncertainty in case of foreclosure. The “losses given default� are therefore much more uncertain and definitely assumed to be higher than in “normal� markets. Another possible political risk is that government may enact another debt-forgiveness program, since over- indebtedness is still a major problem. • NPLs for the affordable book are rising modestly and, although no figures are available on “loss given default�, financial institutions operate under the assumption that these figures are relatively high for subsidized units (see above). The focus is, therefore, on reducing the “probability of default�, which makes for conservative underwriting, and, as was explained 24 In addition, low-income households do not prepay or refinance as frequently as middle- and high income households do in case of a downward rate change. 25 This threshold is adjusted periodically. 26 In April caps on fees were announced for the non-secured loan market. 22 above, the “wrong� use of the FLISP subsidy.27 The eviction problem also makes it hard to develop regular mortgage insurance to cover part of the credit risk and attempts to develop mortgage insurance were unsuccessful for these reasons and other institutional constraints. • Banks write mostly adjustable rate loans which increase the risk of non-payment. While some banks discuss the introduction of balloon type mortgage instruments that fix the rate for an initial period (e.g. 5 years) and would give comfort to households, such innovations have not been implemented because of funding risks (see funding below). b. Funding costs and risks: • Banks use mostly deposit funding (and wholesale loans) to fund the mortgage portfolio, since that is still the cheapest funding source. Expanding deposits to support a larger mortgage loan portfolio is expensive though and current interest margins have narrowed since 2009. Other types of funding (e.g., corporate bonds, securitization) are expensive and will drive up the average cost of funds and, therefore mortgage rates. Non-banks fund themselves through equity investors (including employers) or debt, which are more expensive than deposit funding and make it hard for non-banks to compete with banks. • Deposit funding, which is short-term, for long-term mortgage loans increases the funding mismatch. Banks already have a high exposure to mortgage loans in their credit book and may not be keen to expand long-term lending, certainly given current international regulations on the Solvency ratio and Liquidity ratios (under Basel III). While liquidity risk management of the long-term loan book is facilitated by the liquidity facility at the Reserve Bank (R66 billion), this is expensive and banks may be reluctant to use that window for other reasons as well. Securitization is available but it is regarded as deposit-taking by the regulator because true sale is not allowed and securitization does, therefore, not benefit from the regulatory relieve that comes with moving the mortgages off the balance sheet. • Reserve requirements are fairly high for mortgages (35 % and likely to go up to 45 %). Provisioning for higher NPLs of the low-income book adds additional costs. It is as yet not clear whether Basel III will provide relieve for loans with mortgage insurance, but BIS and FSB studies make that appear likely. Targeted mortgage insurance for higher risk borrowers may, therefore, persuade the Reserve Bank to reduce provisions for mortgages covered by such insurance. c. Structure of the industry: • Only the four main banks are allowed to participate in the FLISP program. Non-bank financial institutions such as mortgage- or finance-companies are excluded. With the current risk profile of the banks, innovation in this sector may be hampered. Non-banks are, on the other hand, interested to open up the sector to different types of lending products (e.g., non- mortgage-based capital lease products) and use securitization to off-load risks that might be better handled by investors. Results: These constraints have resulted in the lack of innovation to open up this sector and the “tacit� agreement amongst banks to use the FLISP program to reduce credit risk of currently qualifying borrowers rather than deepening the mortgage market by qualifying households at the margin. As a 27 Banks’ attempts to reduce payment risk by automatic payroll deductions has come under legal scrutiny. 23 consequence the numbers of mortgages for affordable housing, including FLISP-linked mortgages, will likely remain low. Possible Actions at National Level: • Amend the Housing Act #107 of 1997 to remove or substantially change the preemptive clause that restricts the resale of a subsidized house for 8 years and gives the government the right of first refusal when a sale is necessary.28 • At the time of a large roll-out of a new demand-side subsidy or new subsidy application systems, the appropriate national entity should initiate an education campaign on rights and responsibilities of mortgage borrowers, including an explanation of foreclosure risk. This might be the National Consumer Financial Education Committee (in some countries it is the role of the Central Bank, in others it is the Consumer Protection Agency, or the Banker’s Association). At the same time NT could assist in setting rules on ways that interest rates are calculated and advertised to customers (e.g., using the APR). • New savings programs and credit repair programs could be initiated at the same time by the competent national entities, and be linked to the pre-qualification and application process. NT is working on improving several voluntary and statutory debt education programs (including a pilot with the Department of Justice) as well as national savings programs (including the retail savings bond program with tax incentives). Efforts to clear household debt, repair credit and options to rollover debt into the mortgage package should be an integral part of a demand-side mortgage- linked subsidy. NT could work with the financial sector to rekindle the earlier program proposals on credit repair in order to prepare it for use in combination with a finance-linked housing subsidy program. • Move the official pre-qualification of applicants to the national or local subsidy administration, with financial institutions playing a voluntary role only. Financial institutions should focus on the qualification of the household for a mortgage loan and coordination with the subsidy provider. The national (or local) subsidy administration needs to take responsibility for income and household level verification of program participants. This will alleviate the costs to developers and banks somewhat. • BASA to consider the introduction of a “trust-deed� mortgage (or title-based capital lease instrument) which assigns the title to the property to the lender until repayment of the loan has been completed. This instrument is already used in South Africa by non-bank financial institutions, but would merit consideration by the banks specifically for mortgage-linked subsidy programs. • Review earlier efforts to establish a PPP mortgage insurance. The demand-side subsidy should supplement the maximum affordable loan a household can afford (following credit act rules), rather than reduce the maximum loan amount. If further reduction in the credit risk for the lender is deemed necessary to serve the target group, mortgage insurance should be considered. The premium of the mortgage insurance can be paid for by the upfront subsidy for example. MI would be considerably cheaper than the use of an upfront subsidy for credit risk alleviation. For example, if we assume a 4 % upfront fee for MI over a guaranteed loan amount of R300,000, the 28 Political approval for such amendment will be contingent upon the availability of alternative housing options for those selling their house and the existence of a well-functioning existing housing market, in order to allay concerns that sellers do not receive a fair price for the housing asset and will move to inadequate housing. 24 cost would be R12,000, while an average FLISP subsidy is R50,000. Of course, this will only be feasible with political support for necessary foreclosures and due process, since it will be difficult to establish MI if the collateral cannot be repossessed. • NT and BASA may negotiate with the Reserve Bank a feasible reserve treatment of insured mortgages and the implementation of the Available Stable Funding Ratio. • NT may facilitate legal changes to the laws governing securitization in order to standardize the securitization framework and allow equitable assignment/true sale. (Improved securitization systems, might incentivize banks to offer mortgages with an initial fixed rate period.) • Increase competition and innovation in the affordable mortgage sector. Allow qualified non-bank players with proven sustained track-records of successful mortgage origination and servicing to participate in the demand-side subsidy program. • Increase the level of the demand-side subsidy to coincide with the Affordable Segment Level of R20,000 for the FSC. This will facilitate administrative systems for the financial institutions and make sure that the qualifying income level will be increased according to an index. In addition, the Charter Council should be asked to detail the allowable use of the demand-side subsidy in the Code to give it the necessary weight. • Start a process of structured discussions among the banking sector, NT and RBSA on the various national actions to stimulate mortgage lending mentioned above. Possible Actions Local Level: • Allow municipalities and metros to be involved in the subsidy application process and consumer education, using nationally prepared materials and computer tutorials. This will bring the demand-side subsidy closer to consumers. • Assist consumers in understanding the choice they have in debt providers participating in the demand program. • Network with local employers to have them play a role in assisting employees to get access to mortgage finance, provide guarantees, and other supporting measures. The program could be coordinated with existing incentives for employers to provide housing for low-income workers. 4. Summary of main supply constraints Two general constraints in the implementation of the current FLISP surfaced time and again in the many analytical studies that have been produced on the affordable market segment and in our own interviews: lack of supply of affordable mortgagable units (new build is limited and existing stock constrained by titling, quality issues, and high transfer related costs) and lack of access to or affordability of mortgage credit. And of course the two are connected– developers will not enter the affordable housing market for ownership if there is no guarantee that mortgages are easily available to off-load the stock upon completion, and banks cannot grow their mortgage portfolios when stock is not produced. The third constraint, has to do with the governance of the low-income housing sector where neither the municipalities and metros nor the NHFC have incentives to implement the current demand- side subsidy program. In addition, the municipalities are not fully informed about the flexibility which demand-side subsidies can provide them in addressing the varied housing problems of their residents; e.g., improve demand for affordable housing in greenfield development, as well as increase demand for housing in the lower-cost existing stock, support affordable housing development in well located infill 25 areas, and use of demand-side subsidies in upgrading areas for home-improvement. Municipalities could also be encouraged to think about the use of rental demand vouchers, to attract more rental investors to the affordable sector (but that is outside the scope of this consultancy). The initial analysis of the supply and demand constraints and the administrative issues was confirmed during the second and third missions and there appeared to be broad agreement that these constraints are real and need to be addressed. The analysis also shows that reforms that are needed are not that complicated and, if addressed, will make the affordable housing market more efficient overall, not just the subsidized segment. 6. Possible Next Steps During the second and third missions, stakeholders not only confirmed the analysis of constraints, but worked with the mission team to detail different solutions and actions. The results of these discussions was incorporated into section 5 above. The main question now is what the best way would be to implement the necessary changes that will facilitate the implementation of finance-linked demand- side subsidies? The identified reforms are varied and involve many different government and private sector parties. A first step would be to reach consensus among stakeholders – including politicians, officials, financial institutions, and developers - to detail and prioritize the specific actions in each of the supply-side and administrative areas. NDHS and municipal level actions on program administration, actions needed on the supply-side of housing development (developers, HDA, municipalities, service providers, NT/CSP), and actions to expand lending to households in the affordable segment (NT and CSP, Reserve Bank, BASA, financial institutions). This will be done during a workshop with stakeholders during the next mission based on the current mission report. The second step is to identify leading institutions and individuals to spearhead the actions in each main area. Administrative reforms including how municipalities can play a role in implementing demand-side subsidy programs, real supply-side issues at the municipal and national level and national financial sector reforms should be worked out. Third is to take a decision on the precise focus of the next phase of the CSP demand-side subsidy project. There are three main options: 1. Provide support to improvements of the national demand-side programs in collaboration with NDHS that includes comprehensive procedural and operational reforms and addressing identified supply bottlenecks. 2. The design of stand-alone CSP pilot programs in select municipalities/metros that can test the validity of the assumption that sustained municipal involvement improves both the uptake and operation of demand-side subsidies, and facilitates developer/realtor interest. 3. The design of CSP pilots in selected municipalities that focus on local level improvements in the supply-chain and link these to an improved and revamped set of national demand-side programs (for rental and ownership housing, new and existing housing, finished units and home- improvement). 26 Options Description Inputs and Outcomes Comprehensive Providing support to improve the national • Will result in structural redesign of demand-side programs that include i) improvement, efficiency national comprehensive legal, procedural and and expansion of the demand-side operational reforms and specification of affordable housing market programs different applications of demand-side subsidies and greater efficiency of the (see above), ii) addressing identified real subsidy program. supply bottlenecks with NDHS, NHBC, NT, • Requires collaboration of municipalities; iii) addressing lending NDHS and several national constraints with NT, RBSA, FIs, iv) developing and local parties. consumer debt repair & savings programs and • May require external TA other household support and public outreach and possible funding for programs with NDHS, NT, Municipalities, development of NGOs. administrative and operational systems. City pilot w/o The design of stand-alone CSP pilot programs • Requires ring-fenced national in select municipalities/metros that can test the demand-side subsidy redesign of validity of the assumption that sustained amount (e.g., for 40,000 current municipal involvement can make a considerable subsidies); and negotiation demand-side difference in demand-side programs’ successful with NDHS on possible program implementation through i) addressing real program adjustments. supply-side bottlenecks at the local level for • Requires a dedicated new affordable housing developments and the planning entity to be resale market, ii) establishing municipal level established for affordable consumer information, subsidy qualification and and low-income housing at application and disbursement systems for each participating demand-side subsidy, including real time links municipality with sustained with financial institutions, iii) NT/CSP at the funding from NT/CSP. national level could, as much as feasible, • W/o national administrative support reforms in HS grant systems and and operational systems in financial sector. place, CSP/municipalities have to build and fund municipal capacities and admin systems to efficiently implement various demand- side subsidy programs. Costs will depend on sophistication of systems. • Local and small scale focus hinder substantial national level legal and financial sector reforms. City pilots The design of CSP pilots in selected • Will provide important linked to municipalities that focus on local level lessons for the role of national improvements in the supply-chain and link these municipalities in the redesign of to an improved and revamped set of national successful implementation demand-side programs (see above), including of a diverse set of demand- 27 demand-side national level changes to expand mortgage side subsidy programs, programs lending and borrowing for the target groups. while avoiding the cost of stand-alone pilots. • Option depends on the appetite for reforms at the national level. The decision on which option would be the best way forward will depend on discussions between NDHS and NT/CSP. From the recently released Draft White Paper “Towards a Policy Foundation for Development of Human Settlements Legislation�, November 2015, it is clear that NDHS is interested in improving demand-side approaches to subsidizing housing and that it needs to rethink the finance component of demand-side subsidies. There could therefore be important synergies between NT/CSP and a national reform effort of the current demand-side subsidy programs. Moreover, if there is agreement on the need to make changes to the current national program, this would be much more cost-efficient. The mission, therefore, favors the first or the third option. Based on these national level decisions the scope of the current project will become clear, as will the ultimate leadership of the project. No matter, the multi-dimensional nature of developing workable finance-linked demand-side housing subsidy programs requires strong leadership -a champion-, which has the capacity, funding, and authority to drive change at multiple levels. And depending on the decision on the overall scope of the project, the leadership may be with NT/CSP or, for example, with a taskforce composed of a small number of representatives from two National Departments. 28 Annex 1 Role of Pilot Projects at Municipal and Metro Level Only after these main issues related to the current FLISP have been decided, and the role of national ministries and other national parties has come into focus, should a decision be made about the next phase of this project and the usefulness of pilot projects to pioneer a reformed demand-side subsidy program. If the decision is made to design pilot projects either under the scenario of a reformed national program or as stand-alone experimental project, the following issues have to be considered: 1. What is to be measured? • The feasibility of mortgage linked subsidies to deepen mortgage market for new and secondary affordable market housing? • The effectiveness of the application and outreach process at municipal level? • The type of support needed at the municipal level to improve planning and building procedures; alleviate supply bottlenecks? NB: a possible pilot would be designed in such a way that minimize changes would be introduced to the national FLISP program. 2. Selection of Participating Municipalities and Metros: • Selected municipalities should have effective demand for new affordable housing. • Municipalities should have a sizable secondary market, i.e., RDP/BNG housing with titles; former township housing with titles. • Decision-makers should have a keen interest in expansion of FLISP-like projects; o Political interest to work with the national program o Housing and planning department must have champions for the project • Municipalities must already follow NHFC / Provincial protocol. • Pilot characteristics must be able to be applied to national level (e.g., income level, land market, political issues). 3. Pilot Requirements at CSP level: • A dedicated CSP budget allocation for support to municipalities (differs for different municipalities/metros) and time-frame. • Possible additional funding for subsidy program (depends on current provincial allocation). • Dedicated CSP staff to support pilot and coordinate inputs (project management). • Good working relationship between CSP and NHFC/Housing Development Agency/NDHS • A Monitoring and Evaluation system must be in place. Possible Actions by the City Support Program and other Treasury Departments Irrespective of whether CSP will implement a pilot, the National Treasury can explore the following actions to improve financial sector competitiveness, consumer education and credit repair programs and savings programs in order to assist the implementation of demand-side finance-linked subsidy programs. 29 1. General • Find the right partners and coordinate actions to establish a savings program, revamping financial education on mortgage lending and debt counselling, and improving credit repair programs • Qualify other financial institutions than the Big Four banks for participation in demand-side programs (increase competition and innovation) • Consider other lending products that might qualify under a demand-side finance-linked subsidy program, e.g., trust-deed/capital lease, pension-backed loans, owner builder loan • Improve legal framework for securitization • Lead efforts to improve understanding of public investors on mortgage related debt or equity instruments • Revamp efforts to establish a PPP national mortgage insurance scheme • Support industry efforts to reflect mortgage insurance in the reserve treatment of insured mortgages and other Basel III related mortgage issues (e.g., available stable funding ratio) • Lead the efforts to make changes in the Housing Act on the pre-emptive clause • Align FSC and FLISP income brackets • Start a process of structured discussions among the banking sector, NT and RBSA on the various national actions to stimulate mortgage lending mentioned above 2. Supply side support to municipalities • Education in housing market functioning, filtering process –supported by CAHF • Review and align current grant funding (MHSCG, USDG) to municipalities and allow grants to be applied to subsidy administration • Assistance to establish fast-tracked project approval • Support to analyse the entire cost/subsidy/taxation/package • Support to differentiate RDP/BNG from affordable market product • Support and motivate expansion of rating and titling • Champion regional/urban variations in subsidy amount 30