Loading...
20240408 4.B Attachment UCLA Study on Homekey ProgramUCLA Reports Title Lessons from California's Homekey Program: Adding Affordable Housing by Buying Market- Rate Apartment Buildings Permalink https://escholarship.org/uc/item/595296xr Author Phillips, Shane Publication Date 2024-02-29 eScholarship.org Powered by the California Digital Library University of California ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 1February 2024 LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM Adding Affordable Housing by Buying Market-Rate Apartment Buildings LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 2 ///////////////////////////// About The UCLA Lewis Center For Regional Policy Studies The Ralph & Goldy Lewis Center for Regional Policy Studies is a research center in the UCLA Luskin School of Public Affairs. The Lewis Center advances research on how people live, move, and work in the Los Angeles region, with a focus on policies and interventions that provide paths out of poverty. Since 1989, Lewis Center scholars and staff have produced high-quality research on transportation access, housing affordability, labor, immigration, and many other topics, with a focus on the policy impact on vulnerable populations. lewis.ucla.edu Acknowledgments As a land grant institution, the Lewis Center for Regional Policy Studies at UCLA acknowledges the Gabrielino/Tongva peoples as the traditional land caretakers of Tovaangar (Los Angeles basin, So. Channel Islands). Thank you to Carolina Reid of the UC Berkeley Terner Center for advice on identifying and interpreting TCAC project cost data, Christopher Yenson of the Housing Authority of the City of Los Angeles for providing additional context and data regarding Homekey acquisitions, and to Carolina, Christopher, Michael Manville, and Michael Lens for feedback on drafts of this analysis. Report Authors Shane Phillips, Housing Initiative Project Manager, Lewis Center for Regional Policy Studies Photo Credit Homekey acquisition at 21121 Vanowen St/Google Street View (cover) Sundry Photography/iStock ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 3 Upstream Determinants of Homelessness Research Agenda This publication is part of a Lewis Center research agenda, funded by a generous grant from the Conrad N. Hilton Foundation, on the upstream factors in housing and land use policy that increase the systemic risk of homelessness in Los Angeles County. At any given moment, thousands of people in L.A. County experience homelessness, but many thousands more teeter on its brink, living precariously in the region’s unforgiving housing market. Despite considerable public investments in supportive housing and homeless services, the county has thus far failed to reduce homelessness. This lack of progress can partly be attributed to inattention to the upstream determinants of homelessness. A combination of local, state, and federal efforts have helped a growing number of unhoused people return to stable housing, but we have made little headway combatting the conditions that put people at higher risk of homelessness in the first place. We have successfully increased the outflow from homelessness, but we have failed — so far — to reduce the inflow into homelessness. Twenty years ago, Brendan O’Flaherty wrote that homelessness is a problem of “wrong person and wrong place”: People become homeless at the intersection of individual and structural risk factors. Thus, while programs helping people back into stable housing are essential, policy must also change the housing market conditions that increase the systemic risk of homelessness. At the city and county level, high rents and low rental vacancies are the strongest predictors of high rates of homelessness, and so our focus is on the policies that make housing scarce and unaffordable. We place a particular emphasis on the production and preservation of housing affordable to lower-income households, people from disadvantaged racial and ethnic groups, and people experiencing homelessness, and on increasing housing opportunities in historically exclusionary communities accessible to good jobs, schools, parks, and other amenities. We thank the Conrad N. Hilton Foundation for its support. We are grateful to be part of a community committed to ending homelessness in L.A. County, including philanthropies, scholars, public servants, private institutions and businesses, community-based organizations, local advocates, and people with lived experiences of housing instability and homelessness. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 4 ///////////////////////////// Summary of Findings »While best known for helping local governments purchase and convert underused motels into supportive housing, California’s Homekey program has also been used by the Housing Authority of the City of Los Angeles to acquire nine new market-rate buildings to lease at affordable rents to low-income tenants. »When compared to a random sample of 24 similar subsidized ground-up developments (projects that receive subsidies before starting construction), the cost of market-rate Homekey acquisitions was 20% less per unit ($556,100 vs. $446,800). »Potential benefits from the acquisition and conversion of new market-rate housing include: faster delivery of affordable units, reduced risk of overspending on exceptionally costly projects, and permanent public ownership and affordability of acquired properties. »The lower cost of Homekey projects cannot be explained by differences in size, location, or design, nor by capitalizing future social service expenses into development costs. Instead, the lower cost of Homekey projects can likely be attributed to simpler financing mechanisms and fewer non-essential project requirements such as local hire and community spaces. While these requirements can provide important benefits, they often come at the expense of building fewer affordable homes. »Acquisition of new market-rate housing is not a substitute for increasing overall housing production in expensive cities and regions, nor for building housing that meets the needs of specific marginalized communities. »Homekey illustrates how subsidies can be used to lower the price of a housing unit at any point in its lifecycle, regardless of who builds it or to whom the builder intends to rent or sell. Compared to purpose-built affordable housing, the rent paid by tenants of market-rate housing is higher, but the cost to build and acquire it is often lower — and cost ultimately determines how much housing can be purchased with limited subsidies. »State and federal policies and funding programs are not currently structured to encourage acquisition of new market-rate housing. With this in mind, public officials should explore reforms that can better level the playing field between acquisition and subsidized development. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 5 Rising development costs motivate new approaches to providing affordable housing The high cost of development is a major obstacle to increasing the supply of affordable housing in California. This is especially true for publicly-subsidized projects in the state. Between 2008 and 2019, the inflation-adjusted cost of projects funded by the Low-Income Housing Tax Credit (LIHTC) rose nearly 17% — and costs per square foot ballooned by 55% (Reid, 2020). Since 2019, costs have only increased further. In Los Angeles, the average per-unit cost of supportive housing reached nearly $600,000 in 2021 (Galperin, 2022), and a growing number of projects now cost $1 million per unit or more (Dillon and Poston, 2022). These rising costs translate directly into fewer affordable homes for low-income Californians, including many chronically unhoused people. Rising costs have not been limited to subsidized housing; development costs for unsubsidized projects have also increased. Higher costs are driven by a range of factors including impact fees, material and labor costs, land values, and more recently, high interest rates. Still, subsidized developments frequently face unique burdens in regard to costly labor standards, complex funding sources, and local opposition to the presence of low-income housing (and people), among other challenges. In California, specifically, it is widely understood that building subsidized below-market housing is more expensive than building unsubsidized market-rate homes (Raetz et al., 2020). Amidst these challenges, Homekey was established during the COVID-19 pandemic to rapidly expand the supply of affordable housing for people experiencing or at risk of homelessness. By acquiring or converting existing buildings, it has delivered units faster and at lower costs than traditional affordable housing development (Reid et al., 2022). Since 2020, the program has awarded a total of nearly $2 billion to 116 Homekey projects across the state, funding the acquisition or rehabilitation of approximately 6,900 homes. 1 Perhaps best known for helping local governments purchase underutilized motels and convert them into housing, Homekey also funds the acquisition of new and never-before-occupied market-rate buildings. These buildings are then rented at below-market rates to low-income 1 Data is from the California HCD Awards Dashboard, last updated 11/30/2022 and retrieved 8/22/2023: https://www.hcd.ca.gov/grants-and-funding/homekey/awards-dashboard LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 6 ///////////////////////////// tenants. The Housing Authority of the City of Los Angeles (HACLA) has led this effort, purchasing nine market-rate projects in the first two Homekey funding rounds. Similar to other regulated affordable housing programs in California, permanent housing projects funded by Homekey are restricted to low-income households at affordable prices for a term of at least 55 years.2 Compared to converted hotels and motels, market-rate apartments are very similar to affordable housing that is purpose-built for low-income and formerly unhoused residents, including developments funded by the federal government’s LIHTC program, the largest source of affordable housing funding in the country. In this white paper, I compare the costs of acquiring and converting market-rate housing through Homekey to a random sample of 24 similar purpose-built affordable housing developments, financed in part by the Los Angeles Housing Department (LAHD).3 I find that acquisitions are considerably less expensive. Despite similar dwelling sizes, Homekey acquisitions cost an average of 20% less per unit than non-Homekey subsidized developments. Average costs per square foot were 21% to 28% lower. Given these costs, $1 billion of public funding could support the acquisition of 2,238 market-rate units, 440 more units than could be financed through publicly-subsidized development. My research shows that household incomes and rents were similar in neighborhoods where Homekey acquisitions and LAHD-financed development took place. Additionally, the share of residents who are non-Hispanic white and who own their homes were similar or higher in Homekey neighborhoods, but were low in absolute terms across both project types. These results indicate that the cost savings on Homekey projects do not come from acquiring buildings in less desirable neighborhoods or neighborhoods that are less likely to oppose new housing. Cost differences are also not explained by capitalized service costs or differences in the building amenities sometimes included in subsidized developments. Together, these findings suggest that the acquisition of market-rate housing may be an effective strategy for increasing the stock of affordable housing in California, and potentially other markets as well. At the same time, subsidized development often advances priorities beyond creating 2 Homekey Notice of Funding Availability, Round 2, Section 208. Affordability Term, pg. 15: https://www.hcd.ca.gov/sites/default/files/docs/grants-and-funding/homekey/Homekey- 2021Guidelines-NOFA-Amended-05-05-22.pdf 3 This includes projects supported by programs including the Affordable Housing Managed Pipeline, Supportive Housing Program, Affordable Housing Bond Program, and the Proposition HHH Supportive Housing Loan Program. A full list of LAHD-financed projects from 2003 to present can be found here: https://data.lacity.org/Housing-and-Real-Estate/LAHD-Affordable-Housing- Projects-List-2003-to-Pres/mymu-zi3s/about_data ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 7 affordable units, and may include services or design features that better meet the needs of specific populations. I therefore also ask: what might be lost from a greater emphasis on acquisition relative to subsidized development, and what is the right balance between the two? I structure the white paper as follows: in the next section, I provide an overview of data sources and project characteristics for Homekey acquisitions and the comparison group of LAHD- financed developments. A cost analysis for both sets of projects follows. Next, I discuss other possible benefits of market-rate housing acquisition, then I address potential drawbacks and study limitations, and conclude with key takeaways and recommendations for increasing acquisitions as part of a more holistic affordable housing strategy. I began by identifying all Homekey market-rate apartment acquisitions awarded as part of the second round of Homekey applications in a Housing Authority of the City of Los Angeles memo dated January 27, 2022. I then accessed the building permits issued for each project through the Los Angeles Department of Building and Safety’s online building records system. Building permits provide project-level data on the number of dwelling units and parking spaces, total square footage, and square feet by occupancy group.4 The nine Homekey acquisition projects ranged from 27 to 128 units, with an average of 63 units. I drew the sample of non-Homekey affordable housing developments from the LAHD Affordable Housing Project List (2003 to Present), published by the city. This list provides development costs and other details for each project. I filtered the list to include only new construction funded from 2020 through 2022, excluding outlier projects that reported development costs below $300,000 per unit.5 I also excluded projects in which fewer than half of all units were supportive housing, reflecting the intended use of most Homekey units as supportive housing for individuals who are homeless or at risk of homelessness. I then randomly sorted the filtered sample and documented building permit data for the first 24 projects in the list, including the same project-level data as the Homekey buildings on square footage by occupancy group and parking spaces. These projects ranged from 25 to 98 units, with an average of 56 units. 4 The building permit for a Homekey project is linked here for illustrative purposes: https:// drive.google.com/file/d/1rauYSFGSVYP59GAZUVO666fEp20KXY57/view?usp=sharing 5 Nineteen of these projects were identified in the LAHD affordable housing project list, with reported development costs ranging from $0 to $204,156 per unit. Later investigation of the 10 highest-cost projects revealed actual development costs ranging from $497,000 to $678,000 per unit, and a weighted average of $605,000 per unit. Data: project and neighborhood characteristics LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 8 ///////////////////////////// Table 1 (above) shows project characteristics for the full list of nine Homekey acquisitions and the random sample of 24 LAHD-financed affordable housing developments. Projects acquired through Homekey and developments financed by LAHD were broadly similar with respect to the number of units and average dwelling unit size. On average, Homekey projects have slightly more units and nearly four times as much vehicle parking per unit. Dividing total building square footage (using the zoning code definition6) by dwelling units yields similar results for both groups. However, actual dwelling unit sizes (based on square footage in the R2 occupancy group7) are approximately 50 square feet larger on average for Homekey projects. This difference can likely be explained by the fact that subsidized development projects devote a larger share of 6 The zoning code definition of square footage includes dwelling units and common areas such as hallways and stairwells, but generally excludes utility rooms, car and bicycle parking, service areas, and storage areas, and balconies in some cases. 7 The R2 occupancy group corresponds to the space enclosed by the exterior walls of dwelling units. In other words, it excludes everything except dwelling units. Homekey Acquisitions LAHD-Financed Developments Project count 9 24 Total dwelling units 568 1,337 Average units per project 63 56 Average building square feet per unit (zoning code*)739 751 Average dwelling unit size (R2 occ. group**)724 675 Average parking spaces per unit 0.83 0.22 Table 1. Project characteristics of Homekey acquisitions and LAHD-financed developments. * See footnote 6 ** See footnote 7 ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 9 their buildings to non-residential uses, such as commercial space, community rooms, or social service offices. I also investigated neighborhood characteristics with data from the 2017–2021 American Community Survey to confirm that differences in project costs could not be explained by differences in neighborhood socioeconomic or demographic profiles. For example, if LAHD- financed developments are systematically built in higher-income neighborhoods than Homekey acquisitions, then higher LAHD development costs could be explained by higher land values. Similarly, communities where a larger share of residents are homeowners or non-Hispanic white might organize more effectively against development in their neighborhoods (Been et al., 2014; Gabbe, 2018). Table 2 shows that neighborhoods with market-rate Homekey acquisitions have similar characteristics to those where LAHD-financed projects were developed. Neighborhood median and average household incomes are similar for both groups,8 as are median and average gross rents. The median share of non-Hispanic white households is higher in Homekey neighborhoods than LAHD development neighborhoods, but the average across both remains the same. The share of households who own their homes is higher in Homekey neighborhoods. That said, the non-Hispanic white and homeowner population shares are very low for both groups of neighborhoods. 8 Median values are determined by taking the median of census tract-level medians for each neighborhood characteristic. Averages are also calculated because these values do not represent the true median of all households for either group; they are an approximation. Homekey Acquisitions LAHD-Financed Developments Median Average Median Average Household income $53,523 $53,137 $56,130 $53,861 Gross rent $1,437 $1,415 $1,377 $1,387 Share non-Hispanic white 19%22%13%22% Share homeowners 12%18%8%12% Table 2. Neighborhood characteristics of Homekey acquisitions and LAHD-financed developments. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 10 ///////////////////////////// Cost comparison and analysis: Homekey acquisitions and LAHD-financed developments In this section, I compare the costs of acquiring market-rate developments with funding from Homekey to the cost of developing purpose-built affordable housing through more traditional public subsidies. I also explore possible explanations for the cost disparity between these two types of affordable housing. Purchase costs for Homekey acquisitions come from the previously mentioned January 27, 2022 HACLA memo. Initial project costs for LAHD-financed developments came from the LAHD Affordable Housing Project List. Detailed cost data was taken from LIHTC applications submitted to the California Tax Credit Allocation Committee (TCAC);9 23 of 24 projects in the sample sought LIHTC funding. Where reported costs differed between the LAHD project list and TCAC applications — as they often did, with LAHD-reported costs slightly higher on average — I used TCAC costs. Market-rate developers provide amenities like gyms, pools, and gathering spaces with the expectation that tenants will pay more to live in a building with them. Ideally, the additional rent that these facilities bring in will offset the initial cost of building and operating them. In contrast, rents for affordable units are regulated without regard for on-site amenities. This means that affordable housing developers cannot increase rents to pay for things like tenant services and common area maintenance. To solve this problem, they sometimes capitalize these operating costs into their development budgets. If the LAHD-financed projects include operations costs in their development budgets, but Homekey projects do not, this would improperly bias the results in favor of lower Homekey acquisition costs. The reverse is also true if Homekey acquisition budgets incorporate operating costs but LAHD projects do not. To identify possible bias in reported development costs due to capitalized operations costs, I inspected the Sources and Uses budget of each LAHD-financed project’s TCAC application. Specifically, I looked for budget allocations to “Other” reserves line-items such as “operating deficits,” “replacement reserves,” or “transition reserves.” Although not all of these categories represent services or operating reserves, I removed this line-item from the total development cost (TDC) of each project, resulting in a slightly conservative estimate of project costs. Among 9 Applications are distributed across multiple pages on the TCAC website; e.g., each round of 2021 applications can be found at this url: https://www.treasurer.ca.gov/ctcac/2021/application.asp ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 11 the 23 projects that received LIHTC funding, the largest reserve was $773,000, representing 3% of TDC. The development cost reported by LAHD was used for the one non-LIHTC project. In email communications, HACLA staff shared the portion of Homekey acquisition budgets allocated to operating and lease-up reserves, with total reserve costs ranging from $303,000 to $563,000. I did not deduct these reserve costs from the Homekey budgets for the purposes of this analysis, in part because it was not clear how these costs align with the budget categories in TCAC applications. Overall, per-unit costs are much lower for Homekey acquisitions than LAHD-financed developments (Table 3). On average, Homekey projects cost $446,800 per unit, compared to $556,100 per unit for subsidized developments, a 20% difference. At these costs, $1 billion of funding can support the acquisition of either 2,238 affordable units or the ground-up development of 1,798 affordable units. In addition to costs per dwelling unit, I also evaluated costs per square foot to confirm that differences in building or unit sizes cannot explain the lower cost of Homekey acquisitions. Average cost savings for Homekey projects are larger per square foot (psf) than per dwelling unit. Using the more expansive zoning code-defined measure of building size (see footnote 6), Homekey projects cost 21% less per square foot on average compared to LAHD-financed projects. A smaller share of building space is devoted to dwelling units in LAHD-financed developments, and thus Homekey performance improves when costs are divided by dwelling unit square footage (i.e., R2 occupancy group; see footnote 7), saving 28% compared to LAHD-financed developments. Homekey Acquisitions LAHD-Financed Developments Homekey Cost Savings Average cost per dwelling unit $446,800 $556,100 20% Minimum cost per unit $386,719 $376,135 (3%) Maximum cost per unit $546,931 $697,010 22% Average cost per square foot (ZC)$604 $768 21% Minimum cost per square foot (ZC)$391 $418 6% Maximum cost per square foot (ZC)$975 $1,111 12% Average cost per square foot (R2)$595 $824 28% Table 3. Cost comparison of Homekey acquisitions and LAHD-financed developments. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 12 ///////////////////////////// Homekey acquisition costs were also more consistent — that is, within a narrower range — than LAHD-financed development costs. The lowest per-unit cost among Homekey projects was $386,719, just 3% more expensive than the lowest-cost LAHD project. The most expensive LAHD project cost 22% more per unit than the most expensive Homekey acquisition. Put another way, the maximum cost per unit for Homekey acquisitions was lower than the average cost per unit of LAHD-financed projects in the sample. The lower and less variable costs of Homekey projects could be partly due to selection bias: Developers may be more likely to sell their projects if they are distressed or facing uncertain market conditions (such as those related to the COVID-19 pandemic and its aftermath). Meanwhile, public agencies may be more likely to buy steeply discounted developments than more expensive ones. At a more typical time in the region and nation’s economy, finding similar deals might be more challenging, with greater competition from private investors. Regardless, the ability to selectively acquire projects should encourage cost-effectiveness in any economic climate. While this analysis cannot confirm the causes of higher and more variable costs for LAHD- financed developments, they are likely driven by a range of factors identified in previous studies of affordable housing development costs. These include: the complexity and multiplicity of funding sources, prevailing wage and labor-related requirements, and other mandates that may not be required of unsubsidized developments, such as sustainable design features (Kneebone and Reid, 2021; Reid, 2020). The acquisition approach used by Homekey addresses these challenges to varying degrees. While it does not necessarily eliminate the need for multiple funding sources, acquiring projects during or after construction means that financing can be arranged without delaying construction and thus increasing costs. Acquisition saves on labor costs because most unsubsidized development projects are not subject to prevailing wage requirements. Market-rate projects are also sometimes The maximum cost per unit for Homekey acquisitions was lower than the average cost per unit of LAHD- financed projects in the sample. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 13 exempt from or subject to less stringent requirements than those imposed on publicly-funded developments, such as LEED sustainability certifications. They may also face less delay or require less bargaining with local power brokers if market-rate or mixed-income projects are generally subject to less neighborhood opposition than 100% affordable housing projects. Alternatively, policies in the City of Los Angeles favoring 100% affordable projects, including the Transit Oriented Communities incentive program and Mayor Karen Bass’s Executive Directive 1, may give subsidized developments a competitive edge in entitlement and permitting processes. One factor that does not appear to explain the disparity in project costs is the fact that subsidized developments often include non-revenue-generating, community-serving spaces such as computer labs or office space for on-site services. Approximately 4% of total building area in LAHD-financed developments was dedicated to B and A-3 occupancy group uses, which may be used for offices and community rooms, respectively. While this is significantly more than the 0.2% of building area linked to these uses in the Homekey projects, it is not enough to account for a cost differential of 20% per dwelling unit. Furthermore, only 14 of 24 LAHD-financed projects included any B or A-3 occupancy group space, and the average per-unit cost of these 14 projects was nearly equal to the full sample of LAHD projects ($558,800 vs. $556,100). HACLA should not need to invest heavily into adapting its acquisitions if these spaces are not essential in all purpose- built supportive and affordable housing developments. Acquisitions and ground-up subsidized developments may also simply be used to meet different community needs. One area where acquisition costs may be higher is fees. Purpose-built affordable housing projects have some fees reduced or waived (such as park fees, which cost upward of $8,200 per unit in 2023). Because development fees are usually paid before construction begins, such waivers may be unavailable to market-rate or mixed-income developments, even if they are eventually acquired by a public agency and used as low-income housing. Further research is needed to better understand how acquisitions reduce costs relative to subsidized developments and what trade-offs may be involved. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 14 ///////////////////////////// Other benefits of acquisition Arguably the most important benefit of market-rate housing acquisition, as a complement to subsidized development, is that lower costs mean more households can access affordable housing. But acquisition has several other potential advantages. First, it can deliver affordable housing more quickly than ground-up subsidized development. While it takes 3 to 5 years or more for a project to move from site acquisition to entitlement and permitting, and finally to construction and lease-up, acquisition allows a public agency to engage much later in the process, when construction is already complete or well underway. This can shorten the timeline from funding allocation to occupied affordable housing to a total of 1 to 2 years or less. Second, acquisitions transfer risk more effectively than traditional subsidized development. The latter approach relies on publicly-subsidized private sector (for-profit or non-profit) developers, with high costs and contingencies that are generally borne by taxpayers. Instead of paying what it costs to build a housing development — regardless of its value to future tenants or on the open market — acquisition allows public agencies to pay what projects are actually worth. Among the least expensive projects, Homekey acquisitions and LAHD-financed developments have similar costs. However, LAHD developments have a much wider upper range, whereas Homekey acquisitions demonstrate restraint. If a market-rate development is exceptionally expensive or comes in over budget, the public agency can simply choose not to purchase that project. In any given year, market-rate projects significantly outnumber subsidized developments. Thus, acquisition may also expand affordable housing opportunities to include a wider range of project amenities and locations. While the location of Homekey projects was similar to LAHD-financed developments within this sample, public agencies could consider prioritizing acquisitions in more geographically diverse or higher-resourced neighborhoods, and they would have many options to choose from. The market-rate projects included in this analysis were purchased by the Housing Authority of the City of Los Angeles, opening the door to permanent public ownership and permanent affordability.10 Most publicly-funded affordable housing is built and owned by for-profit or non- profit corporations, with affordability restrictions that expire after 55 years in California, and only 15 or 30 years in many other states (Phillips, 2020). Upon expiration, units typically revert to market- 10 According to communications with HACLA, four of the properties are being transferred to owner-operators selected by the City of Los Angeles Housing Department. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 15 rate rents or require additional public funding to maintain below-market rents. As a result, existing affordable units are lost or new units are foregone. In contrast, a public agency can maintain affordability indefinitely, not giving up ownership of valuable land to private entities. Public ownership doesn’t mean that preserving affordability is free of cost — buildings may still require subsidies to fund rehabilitation in later years, for example — but it does make it less likely that the affordability provisions will fully and permanently expire. More abstractly, but importantly, these acquisitions draw attention to the fuzzy line separating “luxury” market-rate housing from affordable below-market housing. While some below-market- rate units are produced without public funding — e.g., via density bonus incentives — most are built using government subsidies. Building homes is expensive, and producing enough affordable housing to meet the needs of unhoused and low income people requires vast public subsidies. This remains true whether subsidies are introduced before or after construction, and whether developers operate as profit-seeking or non-for-profit entities. Evidence shows that increased market-rate production improves affordability for all income levels (Asquith et al., 2023; Been et al., 2023; Saks, 2008). Despite this, some view new market-rate homes as a missed opportunity for affordable units, at best, or as a driver of higher rents and displacement at worst. The relatively high rents of market-rate units are viewed as an immutable consequence of being built for profit, or for higher-income residents. This analysis shows that acquiring market- rate housing is not inherently more expensive than building below-market homes. If anything, the opposite is true. From an affordability perspective, the most salient difference between the two is that below-market homes are usually publicly subsidized and market-rate homes are not. Homekey illustrates how subsidies can be used to lower the price of housing at any time in its lifecycle. Compared to below-market housing, the rent paid by tenants of market-rate housing Instead of paying what it costs to build a housing development, regardless of its value, acquisition allows public agencies to pay what projects are actually worth. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 16 ///////////////////////////// is higher, but the cost to build and acquire it is often lower — and costs ultimately determine how much housing can be purchased with limited subsidies. Limitations and potential drawbacks Homekey is a new program and market-rate acquisitions account for only a small share of Homekey-funded projects. The Housing Authority of the City of Los Angeles acquired nine such properties with Homekey Round 2 funding, although it plans to acquire three more with funding from Round 3. While the cost savings revealed by this analysis are promising, they should be interpreted cautiously. One remaining question is whether a scaled-up acquisition program could maintain the cost savings highlighted throughout this brief. If some or all of the nine acquisitions were inexpensively priced only due to exceptional circumstances — motivated by COVID-19-related difficulties or high interest rates, for example — then it would be difficult to replicate or expand cost savings within the context of a different market environment. But while there are always challenges with scaling up successful pilot programs, this concern may be unfounded. Since 2019, more than 80% of the approximately 89,000 housing units entitled by the city have been in unsubsidized projects. The 568 units acquired through Homekey represent less than 1% of this total.11 As noted above, the maximum cost per unit for Homekey acquisitions was lower than the average per-unit cost of LAHD-financed projects, indicating considerable latitude for further acquisitions at reasonable prices. Acquisitions also depend upon willing sellers, which are not always guaranteed. It is not known exactly how frequently developers sell their properties shortly after construction, but evidence from Los Angeles shows that most buildings do not change ownership within five to seven years of opening (Phillips and Ofek, 2022). Prior to the introduction of Homekey, there were fewer opportunities for homebuilders to sell their projects before completion — UCLA’s acquisition of five market-rate projects for student housing is a notable exception — so it is difficult to predict how developers might respond if funding for such acquisitions were increased. Public agencies could also expand their acquisition opportunities by purchasing newer buildings that were already partially or fully occupied, similar to recent efforts by the California Statewide Communities Development Authority (Khouri, 2021), or they could purchase older buildings. 11 Data from the Department of City Planning’s housing progress dashboard, Slide 2: “Housing Approved Through Planning Entitlements,” and Slide 6: “100% Affordable Housing Approved through Planning Entitlements.” Link: https://planning.lacity.org/resources/housing-reports ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 17 A related question is whether increased spending on housing acquisitions would drive up purchase prices and erode the cost savings of this approach. On the one hand, expanding Homekey to acquire 5,000 units per year (e.g.) rather than 500 would likely increase competition, potentially raising the cost of acquisition for public and private entities alike. On the other hand, the presence of a well-funded acquisition program could drive additional production by providing a clear financial exit for efficient developers — in the case of Los Angeles, helping meet its ambitious target of 450,000 new homes by 2029. Going further, cities could guarantee acquisition of projects that meet certain criteria, lowering financing costs by reducing borrower risk. Another possible drawback is that by ceding construction to privately-financed developers, the public sector loses some influence over where exactly affordable housing is built. This concern is particularly salient as California lawmakers increase their oversight of local governments’ housing plans in efforts to affirmatively further fair housing (Monkkonen et al., 2023). But while acquisition of market-rate housing and conversion to below-market rents could reinforce patterns of segregation if done poorly, it may also break down those patterns more efficiently than affordable housing development alone. At least three characteristics give acquisition a potential advantage for affirmatively furthering fair housing: the lower cost of acquired units, the much larger number of privately-owned market-rate buildings compared to subsidized developments, and the ability to purchase older buildings where local regulations restrict future development. Finally, a major obstacle is that acquisitions — especially of new or soon-to-be-completed housing — are not as integrated into affordable housing finance policies and programs as ground-up development or rehabilitation of older buildings. These latter project types can receive funding from the LIHTC program, which allows states and local governments to leverage their own investments with matching funds from the federal government. LIHTC funds roughly 30% of total development costs for developers receiving 4% credits, or around 70% using 9% credits (which are fewer in number and much more competitive). Thus, while Homekey acquisitions may cost 80% as much per unit as LAHD-financed developments, the cost to the city of Los Angeles and the state of California will often be greater for Homekey-style acquisitions. The fact that affordable housing policy discourages cost-saving innovations like Homekey merits attention, and federal officials in particular should explore LIHTC reforms to better accommodate the acquisition of new (and newer) buildings. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 18 ///////////////////////////// Key takeaways and recommendations This section summarizes key takeaways from this analysis and recommendations for policymakers. 1. Homekey acquisitions are substantially less expensive than purpose-built affordable housing developments, with less variation in project costs. Increasing funding for acquisitions could create more affordable housing for unhoused individuals and lower-income households. On average, Homekey acquisitions cost 20% less per unit than the LAHD-financed developments examined in this report. Similarly, costs per square foot were lower and less variable for Homekey acquisitions. At these costs, $1 billion spent on acquisition and conversion of new market-rate housing could create 2,238 affordable units – 440 more than would be produced through subsidized development. 2. Acquisition and conversion of new market-rate housing has other potential benefits, including: faster delivery of affordable units, reduced risk of overspending, and permanent public ownership. From the moment that all necessary funds are secured, developing a new affordable housing project requires at least three to five years. In contrast, a market-rate building can be acquired in as little as a few months, helping to meet the immediate need for affordable housing in Los Angeles and across California. Instead of paying what it costs to build a housing development, acquisition allows public agencies to pay what projects are actually worth. The lower and more consistent cost of Homekey projects can be credited, at least in part, to HACLA’s ability to selectively acquire buildings that offer the best value to the public. Subsidized developments are typically built and owned by private entities. When a project’s affordability restrictions expire, it will typically revert to market-rate or require another infusion of government subsidies to extend its affordability. In either case, public funds must be expended to offset or forestall the loss of a below-market home. In contrast, acquisition allows for permanent public ownership (but not necessarily operation) of both land and structure. While this does not eliminate the need for later subsidies, it does place the value of the property and decisions about its future in public hands. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 19 3. Lower costs for Homekey acquisitions compared to LAHD-financed developments are not explained by social service cost capitalization, nor by systemic differences in unit size, building location, or space dedicated to non-revenue-generating uses. Detailed investigation of project budgets revealed that social service cost reserves accounted for no more than 3% of total development cost in any of the sampled subsidized projects. The average number of units and total building square footage per unit (including non- residential space) are similar across Homekey acquisitions and LAHD-financed developments, but the average Homekey dwelling unit is somewhat larger. Buildings were also distributed across similar neighborhoods, with comparable average and median incomes. Notably, rents and shares of non-Hispanic white and homeowner households were slightly higher in Homekey acquisition neighborhoods. Lower costs for Homekey acquisitions are not explained by these modest differences, and may even indicate that Homekey acquisitions took place in slightly more desirable or higher-opportunity neighborhoods. Approximately 4.0% of total building area in LAHD-financed developments was dedicated to B and A-3 occupancy group uses, which may be used for offices providing supportive services and community rooms, respectively. Only 0.2% of building area was linked to these uses in the Homekey projects, but this disparity cannot explain the cost differential between acquisitions and subsidized developments. Furthermore, only 14 out of 24 LAHD-financed projects included any B or A-3 occupancy group space, indicating that these uses are not essential in all affordable or supportive housing contexts, though they are certainly vital in many cases. 4. Reduced financing complexity and fewer non-essential project requirements likely play a role in the lower cost of Homekey acquisitions. Flexibility in the use of public funds will be critical to the success of future acquisition efforts. Typically, subsidized developments require a larger number of funding sources than unsubsidized projects, and the disparity has widened as development costs rise (Kneebone and Reid, 2021). Managing the different timelines and requirements of multiple funding sources can delay projects and increase development costs. While there are many state and local efforts intended to streamline financing for affordable housing development, acquisition likely remains the simpler approach for increasing the stock of below-market homes. Future acquisition programs should utilize the Homekey model, in which a pool of funds is secured and then deployed over time to acquire a portfolio of buildings, balancing priorities LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 20 ///////////////////////////// such as cost-effectiveness, unit mix, and geographic distribution. This is in stark contrast to the subsidized development model in which each project assembles funding independently. Development projects financed with public funds must pay prevailing wages. Often, they must also provide neighborhood or resident amenities, such as community-serving commercial space. These requirements increase development costs and reduce the number of affordable homes available to low-income households, but they also benefit different communities. Ultimately, policymakers must decide which goals to prioritize. However, even if affordable housing is not the highest priority, there may be more cost-effective ways to meet other goals and address community needs. 5. With only nine Homekey acquisitions to date in the city of Los Angeles, more research is needed to assess the model’s applicability to other contexts, as well as its potential to scale up to serve more households. Homekey acquisitions took place in the midst of the COVID-19 pandemic, so it is unclear whether or not similar cost savings could be realized under more typical housing and financial market conditions. Many successful pilot programs struggle to maintain their effectiveness after scaling up. While caution is warranted, there is reason to believe that an expanded acquisition program could succeed. Since 2019, HACLA’s market-rate acquisitions represent less than 1% of housing units entitled by the city, and acquisitions shouldn’t necessarily be limited only to new buildings. A large stock of both new and older buildings represents countless opportunities for public officials to identify cost-effective acquisitions across a range of unit types and neighborhoods. 6. Homekey acquisitions draw attention to the fuzzy line dividing “luxury” market-rate housing from affordable below-market housing, and the critical role that subsidies play in producing affordable homes. This analysis shows that market-rate homes are not inherently more expensive to build than below-market homes. If anything, the opposite is true. From an affordability perspective, the most salient difference between the two is that below-market homes are usually publicly subsidized, while market-rate homes are not. Homekey illustrates how subsidies can be used to lower the price of housing at any point in its lifecycle, regardless of who builds it or to whom they intend to rent or sell. Compared to below-market housing, the rent paid by tenants of market-rate housing is higher, but the cost to build and acquire it is often lower — and costs ultimately determine how much housing can be purchased with limited subsidies. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 21 7. Acquisition is not a substitute for increased housing production. Acquiring market-rate apartments is a useful strategy for increasing the stock of affordable below-market housing, but building housing is critical for stabilizing rents and home prices across the housing market, for people at all income levels, including lower-income households. One important benefit of subsidized housing production is that it adds to the overall housing stock — though on a less than one-for-one basis due to crowd-out of unsubsidized housing development (Eriksen and Rosenthal, 2010). Acquisition does not add to the housing stock, and must therefore be accompanied by targeted efforts to increase market-rate housing production. 8. Acquisition is also not a substitute for housing built to meet the needs of specific populations. Although only slightly more than half of LAHD-financed developments include B or A-3 occupancy group uses, these spaces remain essential for certain populations served by supportive or affordable housing. Further, while this analysis cannot account for unit design, some apartments must be designed to be accessible to people with disabilities, and this is a use case in which subsidized housing developers excel. Even if acquired properties can be renovated to accommodate these needs, doing so may not be the most cost-effective solution. 9. The state of California should expand funding for acquisition of new market-rate units and conversion to deed-restricted affordable housing, providing technical assistance to small and mid-sized housing authorities that are interested in pursuing the model. Unlike other affordable housing programs like the Low Income Housing Tax Credit and Housing Choice vouchers, there is no permanent source of funding for acquiring new multifamily buildings. Building on the success of the Homekey program, California should explore establishing a more permanent program with dedicated funding. Considering that other local housing authorities do not have the same staffing levels or resources as Los Angeles, the state should provide technical assistance to smaller housing authorities to ensure the public’s interests are fairly represented in acquisition deals. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 22 ///////////////////////////// 10. Federal agencies, in partnership with state officials, should consider allowing municipalities to leverage local investments with funding from new or existing housing programs to finance future acquisitions. Currently, acquisition of new market-rate housing is not integrated into affordable housing finance policies and programs in the same way as ground-up development or rehabilitation of older buildings, limiting local governments’ ability to leverage their investments with state and federal funding. As a result, acquisition may reduce the total cost of providing affordable housing, but increase costs to the local government (or the state, in the case of Homekey). Federal officials should explore how new or existing affordable housing funding programs can further encourage acquisition of new market-rate buildings. To ensure that acquisition complements rather than substitutes for housing production, funds should be restricted to local governments that meet prescribed production goals. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 23 Asquith, B. J., Mast, E., & Reed, D. (2023). Local effects of large new apartment buildings in low-income areas. The Review of Economics and Statistics, 105(2), 359-375. Been, V., Ellen, I. G., & O’Regan, K. (2023). Supply Skepticism Revisited. NYU Law and Economics Research Paper, forthcoming. Been, V., Madar, J., & McDonnell, S. (2014). Urban Land‐Use Regulation: Are Homevoters Overtaking the Growth Machine? Journal of Empirical Legal Studies, 11(2), 227-265. Dillon, L., & Poston, B. (2022, June 20). Affordable housing in California now routinely tops $1 million per apartment to build. Los Angeles Times. Retrieved from https://www.latimes.com/homeless-housing/story/2022-06-20/california- affordable-housing-cost-1-million-apartment Eriksen, M. D., & Rosenthal, S. S. (2010). Crowd out effects of place-based subsidized rental housing: New evidence from the LIHTC program. Journal of Public Economics, 94(11-12), 953-966. Gabbe, C. J. (2018). Why are regulations changed? A parcel analysis of upzoning in Los Angeles. Journal of planning education and research, 38(3), 289-300. Galperin, R. (2022, February 23). The Problems and Progress of Prop. HHH. Controller, City of Los Angeles. Khouri, A. (2021, August 23). Buy a luxury building, then lower the rent: A housing fix for California’s middle class? Los Angeles Times. Retrieved from https://www.latimes.com/business/story/2021-08-23/luxury-apartment-lower-rent- affordable-housing-california-middle-class Kneebone, E., & Reid, C. K. (2021). The Complexity of Financing Low-Income Housing Tax Credit Housing in the United States. Terner Center for Housing Innovation, UC Berkeley. Monkkonen, P., Lens, M., O’Neill, M., Elmendorf, C., Preston, G., & Robichaud, R. (2023). Do Land Use Plans Affirmatively Further Fair Housing? Measuring Progress. Journal of the American Planning Association, 1-14. Phillips, S. (2020). Increasing the Duration of Affordability Requirements for New Affordable Housing. UCLA Lewis Center for Regional Policy Studies. Phillips, S., & Ofek, M. (2022). How Will the Measure ULA Transfer Tax Initiative Impact Housing Production in Los Angeles? UCLA Lewis Center for Regional Policy Studies. Raetz, H., Forscher, T., Kneebone, E., & Reid, C. (2020). The Hard Costs of Construction: Recent Trends in Labor and Materials Costs for Apartment Buildings in California. Terner Center for Housing Innovation, UC Berkeley. Reid, C. (2020). The costs of affordable housing production: Insights from California’s 9% low-income housing tax credit program. Terner Center for Housing Innovation, UC Berkeley. Reid, C., Finnigan, R., & Manji, S. (2022). California’s Homekey Program: Unlocking Housing Opportunities for People Experiencing Homelessness. Terner Center for Housing Innovation, UC Berkeley. Saks, R. E. (2008). Job creation and housing construction: Constraints on metropolitan area employment growth. Journal of Urban Economics, 64(1), 178-195. References LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 24 ///////////////////////////// Address Units Average unit size (sq. ft.) Total cost Cost per unit Cost per sq. ft. (ZC) 3401 E 1st St 49 1,087 $34,153,498 $697,010 $418 9502 N Van Nuys Blvd 49 936 $32,186,545 $656,868 $635 2444 S Barry Ave 61 719 $38,278,805 $627,521 $898 923 S Kenmore Ave 75 496 $46,841,014 $624,547 $904 Table A2. LAHD-financed developments. Appendix List of All Homekey Acquisitions and LAHD-Financed Developments Included Within This Analysis Address Units Average unit size (sq. ft.) Total cost Cost per unit Cost per sq. ft. (ZC) 21121 Vanowen St 101 966 $55,240,000 $546,931 $566 10150 Hillhaven Ave 34 1,008 $17,750,000 $522,059 $518 5050 W Pico Blvd 79 705 $36,500,000 $462,025 $655 7639 Van Nuys Blvd 36 871 $16,600,000 $461,111 $529 2812 W Temple St 42 686 $18,580,800 $442,400 $644 6521 Brynhurst Ave 41 1,083 $17,340,000 $422,927 $391 916 N Alvarado St 27 428 $11,272,500 $417,500 $975 740 S Alvarado St 80 514 $31,000,000 $387,500 $754 1654 W Florence Ave 128 587 $49,500,000 $386,719 $659 Table A1. Homekey acquisitions. ///////////////////////////// UCLA LEWIS CENTER FOR REGIONAL POLICY STUDIES ///////////////////////////// 25 Address Units Average unit size (sq. ft.) Total cost Cost per unit Cost per sq. ft. (ZC) 11408 S Central Ave 64 853 $39,900,980 $623,453 $696 1035 S Berendo St 42 1,016 $26,003,817 $619,139 $671 6940 N Owensmouth Ave 80 1,513 $49,269,595 $615,870 $622 1141 S Crenshaw Blvd 43 689 $26,440,278 $614,890 $597 4719 S Normandie Ave 43 578 $26,130,074 $607,676 $1,111 418 N Firmin St 64 731 $38,372,768 $599,575 $804 4200 W Pico Blvd 54 509 $32,155,802 $595,478 $1,072 5050 N Bakman Ave 40 696 $23,798,875 $594,972 $798 316 N Juanita Ave*53 395 $28,554,279 $538,760 $842 1611 W Montana St 64 672 $34,368,962 $537,015 $775 10243 N Topanga Canyon Blvd 55 414 $29,501,339 $536,388 $1,102 2721 Wilshire Blvd 64 597 $34,079,355 $532,490 $891 12003 S Main St 56 372 $29,689,856 $530,176 $823 6576 S West Blvd 64 812 $33,648,592 $525,759 $647 316 N Juanita Ave*52 370 $26,871,012 $516,750 $922 1049 1/2 S Harvard Blvd 47 503 $22,017,881 $468,466 $1,058 5138 S Broadway 49 427 $22,148,687 $452,014 $948 16015 W Sherman Way 46 617 $19,627,749 $426,690 $822 6501 S Crenshaw Blvd 98 473 $40,057,844 $408,754 $772 9500 S Compton Ave 25 452 $9,403,369 $376,135 $933 Table A2. (continued) LAHD-financed developments. * The two projects at 316 N Juanita Ave are separate buildings developed as part of the Enlightenment Plaza complex in Koreatown. LESSONS FROM CALIFORNIA’S HOMEKEY PROGRAM | Adding Affordable Housing by Buying Market-Rate Apartment Buildings 26 ///////////////////////////// 2381 Public Affairs Building, Los Angeles, CA 90095 lewiscenter@luskin.ucla.edu lewis.ucla.edu © 2024