literature review
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rhpd20 Download by: [University of Utah] Date: 21 July 2017, At: 14:23 Housing Policy Debate ISSN: 1051-1482 (Print) 2152-050X (Online) Journal homepage: http://www.tandfonline.com/loi/rhpd20 Future Prospects for Public Housing in the United States: Lessons From the Rental Assistance Demonstration Program Alex Schwartz To cite this article: Alex Schwartz (2017): Future Prospects for Public Housing in the United States: Lessons From the Rental Assistance Demonstration Program, Housing Policy Debate, DOI:
10.1080/10511482.2017.1287113 To link to this article: http://dx.doi.org/10.1080/10511482.2017.1287113 Published online: 31 Mar 2017.Submit your article to this journal Article views: 127View related articles View Crossmark data Housing Policy Debate, 2017 http://dx.doi.org/10.1080/10511482.2017.1287113 Future Prospects for Public Housing in the United States: Lessons From the Rental Assistance Demonstration Program Alex Schwartz t he new s chool, new york, usa ABSTRACTThis article provides an overview of the Rental Assistance Demonstration (RAD) program in the United States and examines its early implementation from its start in 2013 through April 6, 2016. RAD was devised to address the physical deterioration of public housing and secure a more stable funding stream. It requires public housing authorities to shift properties out of the public housing program into a different subsidy program (project-based Section 8) which enables them to obtain mortgages on more favorable terms and to secure tax-credit investment. The program is currently limited to 185,000 housing units. As of April 6th, the program was fully subscribed, and had generated more than $2 billion in new investment. Extrapolating from the early results, RAD has the potential to yield more than $15 billion for fund the redevelopment and renovation of public housing. In 2013 the U.S. Department of Housing and Urban Development (HUD) launched the Rental Assistance Demonstration (RAD) program for the preservation of public housing. Originally limited to 65,000 units, Congress subsequently expanded the program in 2014 to cover 185,000 public housing units. Devised to reverse the physical deterioration of the public housing stock and address the chronic failure of federal funding to meet the capital needs of public housing, RAD in legislative principle is intended to enable public housing agencies to access new sources of funding to finance essential rehabilitation needs far more quickly than would otherwise be possible.
The purpose of this article is to provide an overview of the RAD program and examine its early implementation from its start in 2013 through April 6, 2016, focusing on public housing. 1 Besides explaining the complex structure of the RAD program, the article examines the extent to which housing authorities of varying size are participating in the program, and the cost and financing of properties that have already completed the program. A key question for the article is to assess how well the pro - gram has enabled housing authorities to access debt and other new funding sources to rehabilitate or redevelop public housing. The article is based on a combination of administrative data, interviews with program administrators and public housing authority (PHA) officials, program documentation, and a review of the nascent literature on the program. HUD’s Office of Recapitalization provided data on individual RAD projects, including their location and size, and, for properties that have completed the conversion process, their cost and sources and uses of funding. 2 Interviewees included program administrators at HUD’s Office of Recapitalization, which oversees the RAD program, and officials at several large PHAs that are par - ticipating most actively in the program. ARTICLE HISTORYReceived 13 o ctober 2016 a ccepted 23 January 2017 KEYWORDSunited s tates; public housing; housing policy © 2017 Virginia Polytechnic i nstitute and state university CONTACT alex schwartz [email protected] 2 A. SCHWARTZ The article begins with basic background on the public housing program and the need for reha- bilitation. Next, it describes the design of the RAD program and its key elements. Third, drawing from unpublished project-level data provided by HUD, the article examines the projects that have received preliminary approval from HUD and the extent to which PHAs have participated in the program.
Then I examine the cost and financing of RAD projects that have completed the process of converting from public housing to project-based Section 8—the mechanism by which RAD enables properties to become eligible for additional funding. Finally, some conclusions are drawn about the program and identifies areas for further research. Overview of RAD Program Context RAD is the latest federal response to the dire needs of public housing. The nation’s oldest housing subsidy program for low-income renters, public housing has been starved for resources for many years (Goetz, 2013; Schwartz, 2014). As of 2013, 44% of the public housing stock had been constructed before 1970; only 9% had been built since 1990 (Schwartz, 2014, p. 165). The most recent assessment of the public housing stock’s capital needs, released in 2010, found a backlog of nearly $25 billion, or $23,593 per unit.
In addition, public housing accrues an additional $3.4 billion in new capital needs every year (Finkel et al., 2010). More recently, it was estimated that the New York City Housing Authority, the largest in the nation, faced a backlog of $17 billion for its public housing (New York City Housing Authority, 2015). Partly as a result of the declining quality of public housing, PHAs tore down or otherwise eliminated more than 250,000 public housing units between 1994 and 2012 (Schwartz, 2014, p. 164). On average, about 10,000 public housing units are lost on net annually through demolition or dispositions (HUD, 2014, p. C-2). One reason for the poor physical condition of public housing is the insufficiency of federal funding to address the stock’s rehabilitation needs, coupled with the limited funding available for the initial construction of projects (Goetz, 2013). Most funding for the renovation of public housing comes from the federal Capital Fund, which is part of HUD’s budget. In 2000, Congress appropriated about $2.9 billion for the Capital Fund; in 2015, this figure was less than $1.9 billion—a decline of 35% without taking inflation into account. Adjusted for inflation (using 2013 dollars), appropriations for the Capital Fund decreased by more than 50% during this period. Moreover, the federal government does not obligate PHAs to use all of their Capital Fund allocations for renovations and other capital improvements. HUD has authorized them to use up to 20% of their Capital Fund allocations to help subsidize operating costs (Schwartz, 2014). In sum, federal support for public housing fails to cover the annual accrual of new capital needs, much less the accumulated backlog.
Program Design The RAD program was designed to enable public housing agencies to access funds that were previously unavailable for capital improvements and other purposes. Although PHAs have been permitted to bor - row against their Capital Fund allocations since the late 1990s (Solomon, 2005), RAD greatly increases their ability to utilize debt. 3 In addition, RAD made public housing eligible, for the first time, for Low- Income Housing Tax Credits (LIHTC), an essential means of attracting equity investment for low-income rental housing (Schwartz, 2014). 4 To receive these funds, however, public housing developments must leave the public housing program and convert to project-based Section 8. 5 The local housing authority must enter into a multi-year Housing Assistance Payment (HAP) contract with HUD, in which HUD commits to provide the development with funds equivalent to the Capital and Operating Funds previously received by the development. For example, if a development received $2 million in Operating and Capital Funds, HUD would commit this amount to its Section 8 contract, and subtract it from its annual Capital and Operating Fund allocation to the PHA. The former public housing HOUSING POLICY DEBATE 3 development, now converted to project-based Section 8, can still be owned by the PHA or by a nonprofit organization. As a project-based Section 8 development, it is now eligible for mortgage financing, tax credits, and other funding to cover essential rehabilitation and development costs.
The following hypothetical example illustrates how the RAD program functions. This 100-unit property, in need of extensive capital improvements, including windows, roofs, and boilers, received $173,000 in Capital Funds and $346,000 annually in Operating Funds and $360,000 in rental income.
The housing authority and HUD signed a Section 8 HAP contract obligating the federal government to provide $519,000 annually for the project (the sum of the Capital and Operating Fund allocations) to the RAD development. (HUD also reduces its future Capital and Operating Fund allocations to the PHA by this amount.) The project-based Section 8 allocation, combined with the tenant rents, is termed the project’s contract rent, which the owner can borrow against. Assuming that the project’s annual operating costs will total $620,000 after renovations are completed, its net operating income would total $259,000.
Assuming a debt coverage ratio of 1.2, the development would be able to afford $215,833 for debt service. With this income, the owner could take out a 40-year mortgage at 6% interest in the amount of $3.27 million. In other words, the property could undertake more than $3.2 million in repairs and capital improvements almost immediately. Had the project remained in public housing, the most that could have been borrowed would have been $753,155. 6 Without borrowing, the project would have continued to receive $173,000 annually in Capital Funds. With the conversion to project-based Section 8, the property’s owner enters into a 15- or 20-year Rental Assistance Payment (HAP) contract with HUD, which must be renewed on expiration, as must all subsequent HAP contracts (HUD, 2015a). The owner must decide on one of two types of assistance:
project-based vouchers (PBV ) or project-based rental assistance (PBRA). These two options are very similar. In both cases tenants contribute 30% of their adjusted income toward rent, with the govern- ment covering the rest. The main difference between the two options is that that PBV HAP contracts are usually administered by the housing authority that converted the property to RAD, whereas PBRA contracts are generally administered by HUD’s Office of Housing. The initial length of PBV contacts must be 15 to 20 years, whereas PBRA contracts are set at 20 years. The initial rents and maximum rent increases allowed under the two types of contracts also differ slightly (see HUD, 2015a for details regarding the two types of contract). RAD differs in important respects from previous programs aimed at preserving or redeveloping public housing, including the HOPE VI program. HOPE VI, which redeveloped 262 public housing devel - opments from 1993 to 2010, demolished far more public housing units than it replaced, and tenant selection criteria for the redeveloped public housing often deemed many original residents ineligible to live there (Schwartz, 2014; Vale & Shamsuddin, 2017). For example, the Chicago Housing Authority required, among other things, that all tenants of working age in the redeveloped public housing be employed or in school, and that they not have criminal records (Popkin, 2016; Schwartz, 2014).The RAD program, in contrast, requires one-for-one replacement of nearly all public housing units; conversion to RAD must not involve any loss of public housing units. 7 In addition, with very few excep - tions, all current residents are eligible to reside in the property after its conversion to RAD. HUD’s Final Implementation Notice states that: any resident that may need to temporarily be relocated to facilitate rehabilitation or construction has a right to return to an assigned unit at the Covered Project once rehabilitation or construction is completed. Permanent involuntary displacement of residents may not occur as a result of a project’s conversion of assistance [from public housing to project-based Section 8], including but not limited to, as a result of a change in bedroom distribution, a de minimis reduction of units, the reconfiguration of efficiency apartments, or the repurposing of dwelling units in order to facilitate social service delivery. (HUD, 2015a, p. 27) Another distinction is RAD’s tenant mobility option. Tenants in RAD properties may trade their unit for a rental voucher. In other words, residents who chose to relocate do not forfeit their rental subsidy.
The terms of the tenant mobility option differ somewhat between the two types of project-based assistance contracts (PBV and PBRA; see HUD, 2015a for details).
4 A. SCHWARTZ Finally, the RAD program can be applied for the rehabilitation of existing buildings or the construc - tion of new housing. PHAs may combine two or more public housing developments into a single RAD project; conversely, they can also convert a portion of a public housing development to RAD and leave the rest in public housing. RAD projects need not consist entirely of former public housing units. PHAs have the option of using the program to develop mixed-income developments that combine heavily subsidized units for former public housing residents with market-rate housing and/or less deeply subsi - dized rental housing (e.g., with LIHTC). PHAs often seek to convert individual public housing properties to project-based Section 8, but they can also apply to convert multiple properties or entire portfolios.
Multiphase projects are also eligible (HUD, 2015a).
Key Assumptions RAD is premised on several key assumptions. One is that the transfer of properties from the public housing program to project-based Section 8 enables them to access bank loans, bond proceeds, and tax-credit equity that would otherwise not be available had the housing remained public housing. Second, as a project-based Section 8 property, its federal funding would be more stable and secure. As noted above, Congressional appropriations for the public housing Capital Fund have declined sharply since 2000, even without considering inflation. Appropriations for PBRA, the category that includes project-based Section 8, have increased over the years, in both nominal and real terms. Whereas appro - priations for the Capital Fund decreased by $704 million, or 27% from fiscal year 2005 to 2015, appro - priations for PBRA increased by $2 billion, or 82%. In inflation-adjusted 2013 dollars, funding for the Capital Fund decreased by 43% whereas funding for PBRA increased by 47% (see Figure 1). Historically, because PBRA is largely used to support low-income properties with subsidy contracts involving private owners, Congress has been reluctant to undermine these contracts by failing to appro - priate adequate sums for the program. If appropriations for PBRA fall short of the need required by the subsidy contracts, the properties would be at risk of foreclosure. At times Congress has delayed its appropriations for this program, and sometimes it has provided funding for less than a full year, but it has seldom cut back support for PBRA by a substantial amount (Schwartz, 2014).
Criticisms and Responses The RAD program is not without criticism. Tenant advocates and others have expressed concern about the privatization of public housing, and especially the possibility of foreclosure (Lee, 2015; Smetak, 2014). Public housing is a public asset, owned and operated by PHAs, a unit of government. Under the project-based Section 8 program the property can be owned by nongovernmental entities and, more importantly, could face mortgage foreclosure if the government curtails or reduces its subsidy payments. In the event of foreclosure, the mortgage lender could take possession of the property and possibly displace the residents. Current residents may receive tenant-based vouchers, but the prop - erty itself would be lost as a resource for affordable housing. Displacement might also occur for other Figure 1. budget authority in billions of constant 2013 dollars for project-based rental assistance, public housing operating, and capital funds.
s ource: HuD, 2015d and s chwartz, 2014. HOUSING POLICY DEBATE 5 reasons, such as the bankruptcy of the property owner or the property’s failure to meet HUD’s housing quality standards. Another set of concerns has to do with the ability of the government to enforce and implement the various tenant-protection safeguards in the RAD program (Smetak, 2014 provides an excellent discussion of these concerns).
RAD includes a variety of provisions that HUD intends to help prevent foreclosure and to protect tenants from displacement should foreclosure occur. It is a matter for future research to learn how successful these treatments have been. Below are some of the key HUD provisions:
1. Public or nonprofit ownership. HUD requires RAD properties to be owned by a “public or non- profit entity” (HUD, 2015a, p. 30). HUD may allow ownership to be transferred to a for-profit entity “to facilitate use of tax credits” but only if the PHA preserves its interest in the property by acting as the sole general partner of the limited partnership, by retaining ownership or con - trol of the land, and other means (HUD, 2015a, p. 30). In the event of foreclosure, bankruptcy, or major violations of the HAP contract, HUD can require that ownership of the property be transferred to a “capable” public (preferred) or nonprofit entity (HUD, 2015a, pp. 30–31).
2. Federal Housing Administration (FHA) mortgage insurance. Many RAD projects are eligible for FHA insurance, and HUD encourages owners to obtain it. With FHA insurance, property ownership would revert to the federal government in the event of foreclosure. Among other things, this could incentivize Congress to fully fund PBRA (Smetak, 2014, p. 56).
3. Use Agreement. Every public housing conversion is subject to a Use Agreement, coterminous with the subsidy contract, including renewals, and “recorded in a superior position to any new or existing financing on a covered project” (HUD, 2015a, p. 15). The Use Agreement requires that if the HAP contract is removed (e.g., in the event of foreclosure), all new tenants must have incomes at or below 80% of area median family income and that rents must not exceed 30% of 80% of area median family income (HUD, 2015a, p. 47). Although, as Smetak (2014, p. 55) points out, the incomes of the vast majority of public housing tenants are far below 80% of the area median, this provision limits the ability of owners to convert properties to market-rate rents. Moreover, existing residents would most likely receive tenant protection vouchers that would enable them to remain in place. Program Scale and Growth When Congress first authorized the RAD program, it capped the total number of units that could be covered at 65,000. In 2014, as part of the 2015 Appropriations Act, Congress increased the cap to 185,000 units, although the Obama Administration requested that there be no limit to the number of allowable units. The Obama Administration also requested Congress to appropriate $10 million for the RAD pro - gram to help fund capital improvements for “public housing properties that cannot feasibly convert to long-term Section 8 rental assistance contracts at existing funding levels, specifically those located in high-poverty neighborhoods” (HUD, 2014) that could not be covered by debt alone. Congress, however, refused to allocate any additional funds to the RAD program, insisting that it be revenue neutral. As a result, all capital improvements must be financed by debt, tax credits, or other existing funding sources.
Table 1. summary of all Rental assistance Demonstration program projects with c ommitments to enter Housing assistance Payment c ontracts or that have closed.
Note. cHaP = c ommittment to enter Housing assistance Payment c ontract.
s ource: unpublished data provided by u .s. Department of Housing and urban Development, o ffice of Recapitalization.
Total projectsTotal unitsMin Max Mean Median closed conversion 283 29,805 4 416 105 77 a ctive cHaPs 877 103,098 1 1,359 118 84 t otal 1,160 132,9031 1,359 115 84 6 A. SCHWARTZ As of April 1, 2016, the 185,000-unit cap had been reached, and HUD had formed a waiting list of additional applicants (HUD, 2016b ). Conversions had been completed for 29,805 units, and prelimi- nary approvals (Commitments to enter Housing Assistance Payment Contracts, or CHAP) were made for an additional 103,098 units. HUD was in the process of finalizing CHAPs for the remaining 51,960 units—which were reserved for PHAs that HUD had approved for portfolio conversions or multiphase projects. 8 Initially, HUD awarded CHAPs on a competitive basis to ensure that “awards were made to PHAs of various sizes and across geographies” (HUD, 2015a, p. 86). From October 25, 2012, to July 2015, HUD granted awards on a first-come, first-served basis to PHAs with complete applications that met HUD’s eligibility criteria. Starting in July 2015, HUD insitituted new selection criteria that favored projects that met one or more of six Priority Categories. These categories include projects that involve rehabilitation or redevelopment of physically or financially obsolete housing, and applications that are part of a comprehensive neighborhood revitalization plan (HUD, 2015a, p. 87). 9 All applications, regardless of these changes in selection criteria, were required to include, among many other things, details on capital needs and project financing. Overview of RAD Projects As of April 6, 2016, HUD had issued CHAPs to 360 PHAs for 1,160 RAD projects involving 132,903 units.
Of these projects, 283 with 29,805 units had closed—that is, they had completed the conversion process from public housing to project-based Section 8 (see Table 1). The mean project, including all with active CHAPs and those that had closed, contained 115 units, with a median of 84 (see Table 1). Slightly more than half of all RAD projects are slated for PBV and slightly less than half for PBRA. However, PBRA projects tend to be larger, with a mean of 127 units compared with 105 for PBV projects. As a result, PBRA and PBV projects each account for about 50% of all RAD units (see Table 2).
Table 2. subsidy contract type.
Note. the database does not indicate the subsidy contract type for one Rental assistance Demonstration program project, a 24-unit development in c onnecticut.
Total projects Total projectsShare (%)Total units Share (%)Mean size (units) Median size (units) Project-based rental assistance 52245.0 66,152 49.8127 99 Project-based vouchers 63754.966,727 50.2105 67 t otal 1,160100.0 132,903 100.0115 84 Table 3. summary of public housing authority (PHa ) participation.
Total projects 1,103 t otal units 132,903 t otal PHas 360 Projects per PHa Mean 3.2 Median 1 Minimum 1 Maximum 46 units per PHa Mean 369 Median 195 Minimum 8 Maximum 6,189 HOUSING POLICY DEBATE 7 Overview of Participating PHAs Number of RAD Projects and Units per PHA Three hundred sixty PHAs had at least one CHAP by April 2016—or had already completed a RAD con- version. The mean number of RAD units per PHA was 369 with a median of 195. The largest RAD port - folio included more than 6,000 housing units, whereas the smallest contained only eight (see Table 3). Half of the participating PHAs had just one RAD project, and almost 19% had two. At the other extreme, 17% of the PHAs had five or more RAD projects, including 11% with seven or more (see Table 4 ). In terms of units, 28% of all participating PHAs had fewer than 100 units in the RAD program, and PHAs with 100 to 499 RAD units accounted for 55%. At the other extreme, PHAs with 1,000 or more RAD units comprised 7%. These PHAs with the largest portfolios of RAD units accounted for 41% of all units in the program. PHAs with 750 to 999 units comprised an additional 8%. PHAs with 100 or fewer units accounted for just 4% of all units in the program (see Table 4). Size and Location of Participating PHAs Although most of the PHAs that have participated to date in the RAD program are small, owning no more than 500 public housing units, large PHAs are overrepresented in the RAD program compared with their profile in the public housing program as a whole. For example, whereas PHAs with more than 3,000 public housing units make up 7% of all PHAs in the RAD program, they constitute only 1% of all PHA. Conversely, PHAs with up to 100 units account for 49% of all PHAs but only 19% of those in the RAD program. Put differently, a much larger proportion of large PHAs are participating in the program than are small PHAs (see Table 5). Regionally, PHAs from the South are overrepresented in the RAD program. Southern PHAs account for 36% of all public housing in the United States, but 55% of all RAD units (with CHAPs or already converted), and 72% of all units that have completed the conversion process. Northeastern PHAs, on contrast, are underrepresented. They make up 35% of all public housing but 19% of all RAD housing and 7% of the all converted units. The proportion of RAD units in the West and Midwest is roughly pro - portional to these regions’ share of all public housing. One reason for the prevalence of southern PHAs in the RAD program is that they tend to be small, with more than 90% holding fewer than 500 units, and nearly half owning 100 or fewer units. On the other hand, Northeastern PHAs tend to be larger. RAD units vary widely as a percentage of the total public housing stock of the participating PHAs.
The mean RAD share is 72%, with a median of 96%. The minimum share is less than 1%, whereas the maximum is 100%. Not surprisingly, the RAD program accounts for the highest proportion of public Table 4. Distribution of projects and units per public housing authority (PHa ).
Total projectsPHAShare (%) 1 18150 2 6719 3 349 4 185 5 113 6 82 7 72 8+ 349 t otal 360100 t otal units PHashare (%) total units in category share (%) < 99 101 28 5,653 4 100–499 19955 47,941 36 500–749 236 14,057 11 750–999 134 11,074 8 1,000+ 247 54,178 41 t otal 360100 132,903 100 8 A. SCHWARTZ housing units in the smallest PHAs, and smaller shares in the larger ones. Virtually all participating PHAs with fewer than 100 units are in the process of converting all of them to the RAD program. Only a handful of participating PHAs with 101 to 500 public housing units are not converting them all as well. However, as shown in Table 6, a sizeable share of larger PHAs are also converting major portions of their public housing to RAD. For example, on average, participating PHAs with 1,001 to 3,000 public housing units are converting more than 30% of their portfolios to RAD. 10 The largest PHAs that are converting large amounts of public housing to RAD are shown in Table 7.
Of the 15 PHAs included, each with public housing stocks of 2,000 units or more, nine are converting at least half of their housing to RAD. The only PHA that is converting less than 15% is the New York City Housing Authority, the nation’s largest, which is converting just one development—albeit with 1,359 units.
11 In fact, the list understates the extent to which these large PHAs are converting their public housing under RAD. It summarizes the number of units for which the housing authorities have received CHAPs; it does not include units that are reserved for portfolio conversions. Several large PHAs, including Birmingham, Alabama; Chicago, Illinois; Greensboro and Charlotte, North Carolina; and San Francisco, California, have permission from HUD to convert all of their public housing to RAD; properties that do not yet have CHAPs will receive them later (see Note 12).
Analysis of RAD Projects That Have Completed the Conversion Process as of April 6, 2016 A total of 283 developments had completed the RAD conversion process by April 6, 2016. This analysis examines 206 of these projects. Sixty-seven projects were omitted from analysis because they combined public housing with other types of housing (subsidized for somewhat higher income households or Table 6. Rental assistance Demonstration program units as a percentage of public housing authority total public housing.
PHa = public housing authority.
s ource: unpublished data provided by u .s. Department of Housing and urban Development, o ffice of Recapitalization, u.s. Depart- ment of Housing and urban Development, Picture of subsidized Households, and affordablehousingonline.com.
MeanMedian MinMax all PHas 71.796.31.0100 by PHa size (total units) <100 95.6100.0 101–500 84.299.7 501–1,000 55.353.4 1,001–3,000 37.426.3 3,001–7,500 3227 7,501+ 18.321.8 Table 5. Distribution of participating public housing authorities (PHa ) by total units in inventory, compared with distribution of all public housing authorities with public housing.
Note. R aD = Rental assistance Demonstration program.
s ource: unpublished data provided by o ffice of Recapitalization, u.s. Department of Housing and urban Development, Picture of subsidized Households, and affordablehousingonline.com.
PHA public housing stock Participating PHAs All PHAs RAD as % of total Total %Total % <100 6719 1,508 49.2 4 101–500 16345 1,210 39.5 13 501–1,000 5816 178 5.833 1,001–3,000 4914 125 4.139 3,001–7,500 175 31 1.0 55 7,501+ 62 10 0.3 60 t otal 360100 3,062 100.0 12 HOUSING POLICY DEBATE 9 unsubsidized market-rate) or commercial space. These projects were eliminated from consideration because the data set included total costs and financing sources for the projects as a whole, but did not indicate the number of non-RAD units (e.g., market rate) that are involved. As a result, it was not possible to determine the cost per RAD unit or to differentiate between the funding sources for the RAD (former public housing) units as opposed to the other units in the development. In addition, 10 other projects were excluded from analysis because of missing cost and/or finance source data. 12 In the following analysis we will look at the total cost per RAD unit of the projects, and the funding sources for these projects. Costs The 206 RAD projects that completed the conversion process and for which data are available cost a total of $2.2 billion. Controlling for differences in project size, the cost per unit ranged from $670 to more than $600,000. The mean cost per unit amounted to $100,663 whereas the median was much less at $42,527, reflecting a strong skew in the cost distribution (see Table 8). One reason for the wide range in costs is that some projects involve new construction, and others varying levels of physical rehabilitation. Unfortunately, the data set does not currently include reliable information on whether projects involve new construction or rehabilitation.
Funding Sources The average RAD project received funds from about four separate sources (mean 4.0; median 3.0). Only 22 (11%) of the projects that closed (and for which data were available) were funded by a single source.
Almost 25%, however, required funds from just two sources. At the other extreme, 18% of the projects received funding from seven or more sources (see Table 8). Projects with one to three funding sources tend to cost much less than projects involving more sources of funding. Table 9 shows that the mean cost per unit for projects with more than three funding sources exceeds $100,000 whereas that for projects with fewer funding sources is less than $50,000 (and $7,684 for projects with just one source). Surprisingly, given that a key objective of the RAD program is to enable public housing projects to borrow funds to finance rehabilitation costs, nearly half of the RAD projects that have closed by April 6 do not use debt. Only 51% of all projects (106) took on debt, and 49% (101) did not (see Table 10).
However, 58% of all RAD units converted through April 6th were in projects that involved debt. As Table 7. large public housing authorities (PHa ) with largest shares of Rental assistance Demonstration program (R aD) units.
Note: table refers to public housing developments with cHaPs or that have already converted to project-based s ection 8. it does not cover public housing with pending cHaPs that are part of portfolio or multi-phase R aD conversions (see note 10).
PHATotal public housing unitsTotal RAD unitsRAD (%) Housing a uthority of the birmingham, al District 5,136 4,760 93 Housing a uthority of the city of greensboro, nc 2,401 2,202 92 c ambridge Housing a uthority 2,434 2,130 88 Housing a uthority of the city of charlotte, nc 3,399 2,922 86 Housing a uthority of the c ounty of cook, il 2,026 1,219 60 Housing a uthority of the city of el Paso, tX 6,103 3,295 54 Mobile Housing b oard 3,409 1,850 54 Metropolitan Development & Housing a gency (nashville, tn) 5,399 2,806 52 Housing a uthority of the city & c ounty of san Francisco, ca 6,103 3,061 50 t ampa Housing a uthority 3,037 1,314 43 Housing a uthority of baltimore city, MD 11,259 4,061 36 chicago Housing a uthority 21,222 6,189 29 Housing a uthority of Philadelphia, P a 14,895 3,623 24 c uyahoga Metropolitan Housing a uthority 10,083 1,925 19 new york city Housing a uthority 175,457 1,359 1 10 A. SCHWARTZ Table 8. summary of completed Rental assistance Demonstration program (R aD) conversions.
Note. all projects includes mixed-income, mixed-use. R aD-only projects indicates no additional market-rate or affordable housing; no commercial space; excludes projects with missing data on total costs.
RAD projects all projects t otal projects 282 t otal units 29,780 Mean units 105.32 Median units 77 Minimum units 4 Maximum units 416 R aD-only projects t otal projects 206 t otal units 24,682 Mean units 119 Median units 100 Minimum units 4 Maximum units m t otal cost per project ($) Mean 10,544,122 Median 3,693,695 Minimum 16,107 Maximum 118,087,263 sum 2,172,089,206 t otal cost per unit ($) Mean 100,663 Median 42,527 Minimum 670 Maximum 637,782 Table 9. total finance sources: completed Rental assistance Demonstration program conversions.
Number of sources Total projectsShare (%)Total units Share (%)Mean cost per unit ($) 1 2210.7 2,236 9.2 7,684 2 5024.4 6,238 25.5 28,793 3 3316.1 3,464 14.2 41,231 4 188.8 2,391 9.8 99,467 5 2210.7 2,746 11.2 198,005 6 2311.2 2,984 12.2 211,422 7 2713.2 2,236 11.9 153,382 8 52.42,898 2.9 312,789 9 42.0701 2.6 183,914 10 10.5643 0.5 195,324 t otal 205100.024,415 100.0 100,560 Table 10. Frequency and amount of selected funding sources.
Note. liHtc = low-income Housing tax credit.FHa = Federal Housing a dministration.
Source Projects (%)Units (%)Mean amount per unit ($) Median amount per unit ($) Mortgage (debt) 5158.0 40,728 25,953 PHa c apital Fund and operating reserves 74.6 74.9 11,166 7,078 4% liHtc 24.129.1 64,491 49,783 9% liHtc 17.113.7 122,217 118,891 Deferred development fees 26.927.9 5,213 4,031 t akeback financing 17.113.7 64,952 45,513 FHa Mortgage insurance 33.342.8 na na (as percentage of properties with mortgages) HOUSING POLICY DEBATE 11 shown below, debt-free properties tend to be smaller than those with debt. (The incidence of debt-free projects is discussed in further detail below.) The most prevalent funding source for RAD projects was PHAs Capital Funds, and capital and oper - ating reserves of the participating PHAs. Although the RAD program requires PHAs to shift a portion of their Capital and Operating funds to designated projects to augment their rental income on a per - manent basis, PHAs can also invest unencumbered Capital Funds and operating reserves to help pay for rehabilitation and other development costs. A total of 144 projects (75%) were funded in this way. Deferred development fees were utilized in 27% of the RAD projects that have closed. Developers postpone some or all of their development fees, thus reducing the amount of money necessary to complete the conversion to RAD and the corresponding rehabilitation or development. 13 Takeback financing (sometimes called seller financing, or similar terms) was used in 20% of the projects. Here, the PHA transfers the current value of the property to the new owner (sometimes an affiliate of the PHA, sometimes a separate nonprofit organization). While takeback financing does not usually involve the transfer of funds, it makes the property eligible for an increased amount of LIHTC by increasing the property’s qualified basis. The LIHTC is another key source of funding. The 4% credit, typically used in conjunction with tax-ex - empt bonds, was utilized by 24% of the projects, and the larger 9% credit was used by 17%.
Other funding sources included local and state governments, often with support from the federal Community Development Block Grant and HOME block grant programs, the Federal Home Loan Bank Affordable Housing Program general partner contributions, housing trust funds, energy conservation grants, disaster recovery grants, and many other sources. Most RAD projects that do not carry debt also do not have LIHTC. Of the 101 properties without debt, 93 also did not receive equity investment in exchange for tax credits. The debt- and LIHTC-free properties account for nearly half of the 206 RAD projects that closed by April 6, 2016, and 40% of all units (see Table 11).
Looking now at the average amounts provided through these funding sources (see Table 10), the LIHTC usually generates the most per unit. On average, the 9% credit yields $122,217 per unit (median $118,891) and the 4% credit $64,493 (median $49,784. Takeback financing also generates substantial sums, with a mean of nearly $65,000 per unit. However, unlike tax credits, this source does not directly generate cash for rehabilitation or other needs; instead, it enables the new owner of the property to qualify for more tax credits than would otherwise be possible. Mortgage debt, when used, generates $40,728 per unit on average (median $25,953). The mean interest rate for RAD mortgages amounted to 4.64% (median 4%), with an average term of 30 years (median 30 years).
Comparison of Projects With and Without Debt As noted above, only about half of the RAD projects that closed through April 6, 2016 relied on debt to finance their rehabilitation or development costs. However, and not surprisingly, debt is an impor - tant resource for the more expensive projects (see Table 11). The average development cost per unit for projects with debt is $162,733 (median, $134,907), compared with just $36,134 (median $9,137) Table 11. Distribution and cost per unit of Rental assistance Demonstration program (R aD) projects with and without mortgages and tax credits.
Total projects Share (%) Total units Share (%) Mean cost per unit Median cost per unit Mortgage and tax credits 8139 10,357 42194,829 149,625 Mortgage, no tax credits 2412 3,828 16 54,411 17,402 no mortgage, tax credits 84 526 2248,228 228,996 no mortgage, no tax credits 9345 9,910 40 17,889 8,036 t otal 206100 24,621 100 100,663 42,527 12 A. SCHWARTZ for projects without debt. In other words, projects with debt are more than 4 times as costly as those without. Projects with mortgage financing also tend to be larger than those without, averaging 135 units as compared with 103.
About one third of the 105 RAD projects that were debt financed, accounting for 43% of all debt-fi- nanced units, carry mortgage insurance from the FHA. FHA mortgage insurance is an obvious way of mitigating the risk of foreclosure—and subsequent displacement of residents (Smetak, 2014). It is unclear why more properties with debt financing have not obtained FHA insurance. Projects with tax-credit equity were much more costly than those without. The average LIHTC prop - erty cost $199,629 per unit (median $162,770), whereas the average property without LIHTC cost $25,381 per unit (median $9,388). Table 12 compares the distribution of RAD projects with and without mortgages across PHAs of varying size, and Table 13 does so across PHAs that are converting varying proportions of their public housing stock to project-based Section 8 under the RAD program. Table 12 shows that smaller PHAs are especially disinclined to borrow funds to support their RAD projects. For example, 86% of the RAD projects owned by PHAs with 100 or fewer public housing units do not involve debt, nor do 52% of the projects owned by PHAs with 100 to 500 units. The table shows that all of the properties owned by PHAs with 3,000 or more public housing units include mortgage financing, but it also shows that more than half of the properties in PHAs with public housing portfolios of 1,000 to 3,000 units do not use debt to cover their rehabilitation or redevelopment costs. Table 13 shows that the extent to which PHAs are converting their public housing to RAD does not correlate closely with their proclivity to use debt. It does indicate that about 70% of the properties and units without debt are owned by PHAs that are departing altogether from the public housing program by means of RAD; with the closing of their RAD projects they no longer own any public housing. Most of these PHAs operated only one public housing development, so that participation in RAD necessitates departure from public housing, but some owned several public housing properties (see Table 12). That about half of all projects that have completed the RAD conversion process through April 6, 2016, did so without debt or LIHTC would seem to run counter to the main purpose of RAD: providing properties with the opportunity to access mortgage credit and equity investments to finance essential Table 12. Public housing authority (PHa ) size and use of mortgages.
PHA size (total units)No mortgage MortgageNo mortgage as percentage of total Projects UnitsProjects Units Projects Units <100 12 776 2 126 86 86 101–500 32 3,475 29 2,775 52 56 501–1,000 18 1,254 24 3,533 43 26 1,001–3,000 35 4,562 26 4,285 57 52 3,001–7,500 4369 14 1,667 22 18 7,501+ 0010 1,799 0 0 t otal 101 10,436 105 14,185 49 42 Table 13. Rental assistance Demonstration program (R aD) share of public housing authority public housing stock and use of mortgages.
RAD share (%) No mortgage MortgageNo mortgage as percentage of total Projects UnitsProjects Units Projects Units <15 2 61 4304 33 17 15.1–35 4 529 121,743 25 23 35.1–50 6 473 203,165 23 13 50.1–75 11 981 243,346 31 23 75.1–95 21 1,540 212,842 50 35 95.1+ 57 6,852 242,785 70 71 t otal 101 10,436 10514,185 49 42 HOUSING POLICY DEBATE 13 renovation and development costs, funding sources that are largely unavailable to public housing. This finding raises the question of what motivated the PHAs to move these developments into the RAD pro - gram in the first place. Was it to access other funding sources? To secure a more stable revenue stream?
To take out debt and other capital resources at a later date? To exit public housing for other reasons?
The properties that closed through the first week of April 2016 represent only about 24% of all the properties (and 22% of the units) that HUD has approved for conversion (i.e., issued CHAPs). It is possible that the properties still in the conversion pipeline will be more inclined to use debt and make use of tax credits. Perhaps debt-free properties are so predominant among the first properties to complete the conversion process because the absence of mortgages or tax credits expedited project underwriting.
The process of applying for mortgages and tax credits, in other words, may prolong the time it takes to complete the financing; for example, there is no need to apply for mortgages, negotiate their terms, and wait for their approval and closing. The timing of RAD conversions to date supports this interpretation. Although 70% of the properties that converted from the 4th quarter of 2013 through the 4th quarter of 2014 did not involve any debt, only 30% of the projects that completed the conversion process in the subsequent six quarters were debt free. Similarly, 69% of the units that converted through the end of 2014 were in debt-free pro - jects, compared with only 21% of the units that converted afterward (see Table 14). Likewise, projects that converted in 2015 and 2016 were far more likely to involve tax-credit investment than the earlier cohorts of conversions. For example, whereas 72% of the projects that closed from the 4th quarter of 2013 through the last quarter of 2014 did not involve LIHTC, the same was true for just 46% of the properties that converted in the subsequent six quarters. 14 Nevertheless, even if the trend line points to increased use of debt and tax credits over time, a substantial number of properties continue to convert without debt or tax credits, especially among the smaller PHAs. These findings, supported by interviews with program administrators and selected PHA officials, suggest that some PHAs see RAD primarily as a way to exit the public housing program and secure a more steady flow of subsidy dollars from the Section 8 program. It is also possible that PHAs selected properties that need relatively little renovation and therefore could be funded without debt or tax credits, or they may have decided to convert to project-based Section 8 as expeditiously as possible, and take out debt at a later time, post-RAD conversion, to invest in rehabilitation. 15 In any case, the question remains why PHAs have not selected properties with more extensive needs, and Table 14. Percentage of projects and units with mortgages and tax credits, by date of Rental assistance Demonstration program conversion.
Note. liHtc = l ow-income Housing tax credit, R aD = Rental assistance Demonstration program. Date of RAD conversion 4Q 2013–4Q 2014 1Q 2015–April 2016 Projects Mortgage ye s 3070 no 7030 liHtc ye s 2854 no 7246 units Mortgage ye s 3179 no 6921 liHtc ye s 2459 no 7641 14 A. SCHWARTZ if PHAs are excluding such properties because of difficulty securing sufficient funding through RAD.
These are essential questions for future research. 16 Conclusions Summary of Findings The RAD program has proven itself to be very popular among PHAs. It quickly reached its original 60,000-unit cap, and promptly reached the increased cap of 185,000 units. As of April 6, 2016, 360 PHAs had received commitments from HUD to enter into a CHAP contract for at least one property, involving 1,160 properties with nearly 133,000 units. A total of 283 properties with nearly 30,000 units had closed, completing the conversion process from public housing to project-based Section 8. An additional 45,691 units are reserved for PHAs for which HUD has approved portfolio conversions or multiphase conversion plans. As of May 1, 2016, RAD has a waiting list of 161 projects with 13,653 units (HUD, 2016c). In little more than 2 years, the RAD program eclipsed HOPE VI as the largest program to reposition public housing. The latter program redeveloped 262 public housing projects over more than 15 years (Schwartz, 2014; Vale & Shamsuddin, 2017).
The properties that have completed the RAD conversion process received nearly $2.2 billion (more, when mixed-income and mixed-use projects are included) to cover rehabilitation and redevelopment costs—orders of magnitude more than what they would have received from federal public housing Capital Fund appropriations. If the properties that have closed resemble the much larger number of properties that await conversion, the entire set of properties in the RAD pipeline (i.e., with CHAPs or portfolio or multiphase reservations) will ultimately receive more than $15.6 billion for capital invest - ments. 17 This amount could well be higher if the proportion of properties in the pipeline involve debt and tax credits to a greater extent than those that have already completed the RAD conversion process.
Other Key Findings • PHAs of all sizes have participated in the RAD program, although larger PHAs are more likely to participate than smaller ones; • Many PHAs are using RAD to leave the public housing program altogether. This is particularly true of smaller PHAs, but several large ones are also converting their entire public housing stock to project-based Section 8 under RAD; • Properties that have completed the RAD conversion process cost $100,663 per unit on average (although the median was less than half of that, at $42,527). On average, these properties are funded by four separate sources, the most frequent of which were PHA Capital Funds and operating reserves, debt, LIHTC, deferred development fees, and takeback financing.
• About half of all the properties to complete the RAD conversion process through April 6, 2016, involved debt financing. However, most debt-free properties converted during the first year of the program’s operation; less than one third of all properties to close after the fourth quarter of 2014 involve no debt, compared with 70% of all properties that closed earlier. Larger PHAs are more likely than smaller ones to use debt. Research Priorities Before concluding with some comments on the broader implications of RAD for public housing and its residents, I offer several questions and topics for additional research on RAD: 1. The objectives of participating PHAs. This article is based largely on administrative data. It pro - vides insight into the number and basic characteristics of RAD projects. These data should be complemented with more qualitative information on the criteria PHA use to decide which prop - erties to submit for RAD conversion. While many PHAs, small ones especially, are converting HOUSING POLICY DEBATE 15 their entire portfolios, others are not. It is important to understand what factors PHAs consider in deciding whether RAD is suitable or appropriate for particular public housing developments.
It is also important to understand in more detail why PHAs choose to convert properties that require little capital investment. Are these PHAs primarily motivated to remove themselves from the uncertain funding and rigid bureaucratic requirements of public housing? Do they intend to invest in capital improvements at a later date? It is also important to understand why some PHAs have opted for portfolio conversions whereas others have sought RAD conversion for only a portion of their public housing. 2. Why do only some large PHAs participate in RAD? Although large PHAs are overrepresented in RAD relative to their position in the public housing program as a whole, a substantial propor - tion are not in the program. 18 (see Table 5). It is important to understand why these PHA have not participated in the program. Have their applications been rejected? Why? If they haven’t applied, why? If the program is expanded, will they apply?
3. A third key area of research is to monitor the experience of properties and their residents after they have converted to project-based Section 8. Will the safeguards built into the pro - gram be sufficient to prevent displacement in the event of mortgage foreclosure? Will the Use Agreements hold up in court if challenged by lenders? Will federal funding for project-based Section 8 remain sufficient to support the operations of the converted properties? Will the properties remain physically and financially viable without additional federal subsidy? Will Congressional appropriations for project-based Section 8 keep pace with the growth of the project-based Section 8 program under RAD? Implications for the Future of Public Housing RAD is not without risk. Shortfalls in Congressional appropriations and other factors could trigger foreclosures, and perhaps displacement of residents. Compared to public housing, appropriations for project-based Section 8 have historically been more stable and have seen real increases over time.
However, there is no guarantee Congress will not cut back on Section 8 funding in the future. That said, the program also includes various provisions aimed at reducing the likelihood of foreclosure, and mitigating its impact should foreclosure occur. Moreover, unlike Hope VI and other public housing redevelopment programs, RAD mandates that the number of units in the original developments must not be decreased in the RAD conversion process by more than a de minimus amount, and it mandates that the original residents be permitted to reside in the property post conversion as well; PHAs are not permitted to impose any eligibility requirements on the original residents.
Although it is impossible to say that these provisions eliminate all risk to the original residents, the program offers a means to rehabilitate or redevelop the properties far more quickly and comprehen- sively than would be possible if the properties remained in the public housing program. In other words, RAD represents a tradeoff between the seemingly remote possibility that foreclosure or other factors could cause residents to be displaced, and the property lost to the affordable housing stock, against the near certainty that if the property remained in public housing it would continue to deteriorate as inadequate federal subsidies cause capital needs to accumulate and put the property at increasing risk of demolition. What does the RAD program mean for the future of public housing? If the program does not increase in size beyond the current cap of 185,000 units, RAD will cause the public housing stock (at about 1.1 million units as of 2014) to decline by 17%, leaving about 915,000 units in the program. The housing that converts to project-based Section 8 under the RAD program is likely to undergo sizeable amounts of rehabilitation and redevelopment. The housing is likely, too, to benefit from a more stable stream of federal subsidy payments over time. The remaining public housing will continue to decline as federal Capital Fund appropriations fall far short of need, and demolitions and dispositions (whereby the PHA sells off or otherwise disposes of the public housing, usually supplying the tenants with vouchers) are likely to continue at their current pace, if not faster.
16 A. SCHWARTZ If Congress eventually increases the maximum number of units that can be part of the RAD program, or lifts the cap altogether, as the Obama Administration requested, there is little reason to doubt that the public housing stock would diminish far more. The housing that remains in public housing would probably consist heavily of properties with the most extensive rehabilitation needs, needs that would be difficult if not impossible to cover without additional subsidy; debt financing, LIHTC, and other public and private sources currently used in RAD conversions would likely be insufficient.
In conclusion, RAD represents an important but insufficient step toward the preservation of pub - lic housing. At a time when nearly half of all very low-income renters spend more than half of their income on rent, including more than 70% of all renters with incomes as low as most public housing residents (HUD, 2015b ), it is all the more important to invest in public housing. Public housing is not only affordable to the lowest income households, it also offers advantages over tenant-based vouch- ers in accommodating large families, the elderly, people with criminal records, and other people who are hard to house in the private housing market (Goetz, 2013; Popkin, 2016). Much of the worst public housing has been demolished and redeveloped under HOPE VI and other programs; the remaining stock is a vital resource that must be preserved.
Notes 1. In addition to public housing, the RAD program also applies to the Section 8 Moderate Rehab program, the Rent Supplement program, and the Rental Assistance Program. These latter legacy programs were established in the 1960s and early 1970s. They account for less than 40,000 units, but are otherwise “ineligible to renew their contracts on terms that favor modernization and long-term preservation” (HUD, 2014, p. C-2).
2. Some, but not all, of the data used for this article are available on RAD’s website: http://www.radresource.net/ firstcomponent.cfm 3. Previously, PHAs were prohibited from borrowing against individual public housing developments. They could borrow against a portion of their future Capital Fund payments, but such borrowing was subject to more stringent terms than what is possible under RAD. For example, the maximum debt service coverage (DSC) ratio for a Capital Fund loan is about 3.0, whereas DSC ratios for mortgages issued under the RAD program are generally around 1.2 or lower (Mathematica, 2016, p. xxi).
4. In addition to debt and tax credits, the RAD program allows PHAs to use a wide array of other sources to pay for rehabilition and redevelopment. The most important of these sources are discussed below.
5. The RAD program is roughly analogous to the Stock Transfer initiative in the United Kingdom during the 1980s and 1990s. Under stock transfer, ownership and management of several hundred thousand units of council housing (analogous to public housing in the United States) were transferred to nonprofit housing associations.
The housing and residents remained the same, but the ownership of the housing changed, as did some of the subsidy mechanisms (Pawson & Fancie, 2003 ). As with the RAD program, a chief motivation of stock transfer was to access private financing (i.e., bonds, bank loans) for the rehabilitation of the housing stock. More recently, several jurisdictions in Australia have also transferred public housing to nonprofit organizations (see Pawson et al., 2013). 6. A maximum of 30% of a project’s Capital Fund allocation can be used for debt service, and the maximum mortgage term is 20 years. This estimate also assumes a debt coverage ratio of 1.2 and an interest rate of 6%.
7. The program’s regulations specify that “conversions may not result in a reduction of the number of assisted units, except by a de minimus amount, defined as no more than the greater of five percent of the number of project or portfolio units under Annual Contribution Contract (ACC) immediately prior to conversion or five units” (HUD 2015a, p. 25). The de minimus amount, however, may be increased in two circumstances: (a) if units have been vacant for more than 24 months at the time of RAD application; (b) if reducing the number of units in the property will enable the PHA to “more effectively or efficiently serve assisted households through (a) reconfiguring apartments (e.g., converting efficiency units to one-bedroom units) or (b) facilitating social service delivery (e.g., converting a basement unit into a community space” (HUD 2015a, p. 25).
8. As of April 2016, 68 PHAs planned to convert their entire public housing portfolios to Section 8 through the RAD program and 42 had multiphase plans. See Note 12 for details. 9. Other priority categories include projects that HUD deems to be in imminent danger of losing financing if they are not awarded a CHAP; other applications using tax credits; applications that are part of a portfolio or multiphase award in which 50% of properties fall under any of the above categories; and all other applications, portfolio awards, and multiphase awards (HUD, 2015a, p. 87). 10. It is important to note that this analysis is based on projects that have received CHAPs or that have already converted to Section 8 under RAD; it does not cover projects that are pending (reserved) for PHAs that have been approved for portfolio conversions or multiphase developments. PHAs slated for portfolio conversions and/or multiphase HOUSING POLICY DEBATE 17 developments had yet to receive CHAPs for projects with about 50,000 units as of April 2016. When HUD awards CHAPs to these remaining properties, the number and percentage of PHAs that have converted all of their public housing to RAD will increase further. The 68 PHAs approved for portfolio conversions account for 19% of all PHAs in the RAD program, and 54% of all units. Ultimately, HUD estimates that about 58% of participating PHAs will convert all of their public housing to Section 8 through the RAD program (HUD, 2016a). Put differently, 75% of all properties in RAD involve PHAs that are converting their entire portfolios ( Will Lavy, Senior Advisor, U.S. Department of Housing and Urban Development, Office of Recapitalization, interview, November 19, 2015). 11. In June 2016 the New York City Housing Authority added 21 projects with 5,000 units to the RAD wait list. In January 2017, HUD issued CHAPs for 17 of these projects, involving 1,700 units.
12. My analysis of project costs and funding differs from Mathematica, Inc.’s recent interim evaluation of RAD (Mathematica, Inc. 2016). Whereas I focus exclusively on projects that only include converted public housing units, Mathematica also included projects that contain additional units, whether subsidized by the LIHTC and other sources, or market-rate. As a result, Mathematica’s estimates of total and average project costs are higher.
For example, Mathematica finds that the 185 projects that had closed by mid-October 2015 received an average of $129,000 in funding for construction, rehabilitation, and related costs. My analysis, covering 206 RAD-only projects that had closed by April 6, 2016, indicates that they received an average of $106,000 per unit.
13. Deferred development fees are often applied in underwriting affordable housing; rather than receive the fee at the time of the closing or at the completion of the project, the fee is paid over time from rental revenue.
14. The projects to close before the start of 2015 were also more likely to involve just one or two funding sources compared with projects that converted afterward, and tended to cost less per unit, and contain fewer units. 15. According to an administrator of the RAD program, some PHAs opt to address “the minimum need now,” financing this work with replacement reserves, and apply for tax credits in six or seven years when the property is in better shape and has a more stable funding stream” (Will Lavy, Senior Advisor, U.S. Department of Housing and Urban Development, Office of Recapitalization, interview, November 19, 2015). He also noted that it is not realistic to expect larger PHAs to obtain 9% tax credits for multiple RAD projects at the same time. Instead, they will apply for some projects now, and hold off on the others. These latter projects will receive less investment, and less in any debt financing, until the PHA obtains 9% tax credits for them (Will Lavy, Senior Advisor, U.S. Department of Housing and Urban Development, Office of Recapitalization, interview, November 19, 2015) 16. As noted earlier, HUD revised RAD selection criteria in 2015 to favor properties that are “physically or financially obsolete” or that are part of comprehensive neighborhood revitalization plans (HUD, 2015a). These new criteria will apply to properties currently on the RAD wait list or to new applications if Congress expands or lifts the cap on the number of RAD conversions.
17. This estimate was derived by multiplying the mean cost per unit for the properties that have closed by the total number of units in properties that have yet to complete the conversion process.
18. Some of the larger PHAs that have not participated in RAD include Boston, Massachusetts; Miami, Florida; San Antonio, Texas; and Detroit, Michigan Acknowledgments Special thanks to Gregory Byrne, Will Lavy, and Robert Robinson at HUD’s Office of Recapitalization for providing me with the RAD data and for patiently answering my many questions about the data and about the RAD program itself. Thanks too to Deborah VanAmerongen for tutoring me on the RAD program.
Disclosure Statement No potential conflict of interest was reported by the author.
Notes on Contributor Alex Schwartz is a professor of Urban Policy at the New School. He is the author of Housing Policy in the United States (3rd Edition) (Routledge, 2014) and the managing editor for North America for the journal Housing Studies.
References Finkel, M., Lam, H. K., Blaine, C., de la Cruz, R. J, deMarco, D., Vandawalker, M., & Woodford, M. (2010). Capital needs in the public housing program. Report prepared by Abt Assocites Inc. for the U.S. Department of Housing and Urban Development.
http://portal.hud.gov/hudportal/documents/huddoc?id=PH_Capital_Needs.pdf Goetz, Edward J. (2013). New deal ruins: Race, economic justice, and public housing policy . Ithaca, NY: Cornell University Press.
18 A. SCHWARTZ Lee, J. A. (2015). Rights at risk in privatized public housing. Tulsa Law Review, 50, 759–803.Mathematica, Inc. (2016). Interim report: Evaluation of HUD”s rental assistance demonstration. Prepared for U.S. Department of Housing and Urban Development (Sept.). https://www.huduser.gov/portal/sites/default/files/pdf/RAD-InterimRpt.pdfNew York City Housing Authority. (2015). Next generation NYCHA. http://www1.nyc.gov/assets/nycha/downloads/pdf/ nextgen-nycha-web.pdf Pawson, H., & Fancie, C. (2003). Maturing sssets the evolution of stock transfer housing associations. Bristol: Polity Press.
Pawson, H., Milligan, V., Wiesel, I., & Hulse, K. (2013, October). Public housing transfers: Past, present and prospective. Report prepared for Australian Housing and Urban Research Institute. (AHURI Final Report No. 215). https://www.ahuri.edu.
au/__data/assets/pdf_file/0015/2256/AHURI_Final_Report_No215_Public-housing-transfers-past,-present-and- prospective.pdf Popkin, S. J. (2016). No simple solutions: Transforming public housing in Chicago. Lanham, MD: Rowman and Littlefield.
Schwartz, A. (2014). Housing policy in the United States (3rd ed.). New York, NY: Routledge.
Smetak, A-M. (2014). Private funding, public housing: The devil in the details. Virginia Journal of Social Policy & the Law, 21(1), 1–62.
Solomon, R. (2005). Public housing reform and voucher success: Progress and challenges (Discussion Paper). Washington, DC: The Brookings Institution. http://www.brookings.edu/~/media/research/files/reports/2005/1/01metropolitanpo licy-solomon/20050124_solomon.pdf U.S. Department of Housing and Urban Development. (2014). FY2015 congressional justifications: Rental housing assistance demonstration. http://portal.hud.gov/hudportal/documents/huddoc?id=fy15cj_rent_assist_dem.pdfU.S. Department of Housing and Urban Development, Office of Public and Indian Housing, Office of Housing. (2015a, June 15). Rental assistance demonstration—Final implementation, revision 2. Notice PIH-2012-21 (HA), REV-2.
U.S. Department of Housing and Urban Development. (2015b , April). Worst case housing needs: A report to Congress.
www.huduser.gov/portal//Publications/pdf/WorstCaseNeeds_2015.pdf U.S. Department of Housing and Urban Development. (2015c). Congressional Justifications FY 2016. https://portal.hud. gov/hudportal/documents/huddoc?id=FY16-CJE-EntireFile.pdf U.S. Department of Housing and Urban Development, Office of Public and Indian Housing, Office of Housing. (2016a). HUD RAD Newsletter, Issue 15 (March). http://portal.hud.gov/hudportal/documents/huddoc?id=RAD_Newsltr_march2016.pdf U.S. Department of Housing and Urban Development, Office of Public and Indian Housing, Office of Housing. (2016b ).
HUD RAD Newsletter, Issue 16 (April/May).
Vale, L., & Shamsuddin, S. (2017). All mixed up: Making sense of mixed-income housing developments. Journal of the American Planning Association, 83, 56–67.
