ASSESSING NEIGHBORHOOD CONTEXTS OF LIHTC PROPERTIES

Housing affordability remains a pressing social issue in cities across the U.S. Low-Income Housing Tax Credits (LIHTC) are a popular tool for developing and rehabilitating housing units and incentivizing developers to respond by increasing the affordable housing supply. In a period where movement-to-opportunity policies have dominated housing decisions, it is crucial to understand the characteristics of the neighborhoods where LIHTC properties are located. I am interested in the difference between census tracts with and without LIHTC properties across the Atlanta-Sandy Springs- Roswell Metropolitan Statistical Area. Key questions include whether LIHTC sites are as proximate to schools, parks, and transit stations as non-LIHTC sites and if these neighborhoods have similar demographic and economic compositions. The research aims to identify the characteristics associated with neighborhoods with active LIHTC developments, hypothesizing that LIHTC properties are situated in more diverse communities and may have better access to amenities due to the Department of Housing and Urban Development’s existing reward system.

In the last few decades, the percentage of cost-burdened renters has increased precipitously  _{[1, 2]} . More households struggle to find affordable, adequate, and available (AAA) housing in the nation’s largest urban areas. Nearly one-third of U.S. households are cost-burdened  _[3] . In the five-county core of Atlanta, that number increases to 50% of renters  _[4] . As the cost of most items has increased, public and private entities are looking for ways to mediate these effects.

“Through this program, private investors receive a federal income tax credit as an incentive to make equity investments in affordable rental housing”  _[5] . Between 1987 and 2021, 3.55 million units have been added through low-income housing tax credits. An affordable LIHTC unit is one where housing costs, including utilities, are at or below 30% of the tenant’s income. LIHTC units are affordability-preserved for 30 years, upon which the property owner can elect to convert it to market rate.

“Created by the Tax Reform Act of 1986, the LIHTC program gives state and local LIHTC-allocating agencies the equivalent of approximately $10 billion in annual budget authority to issue tax credits for the acquisition, rehabilitation, or new construction of rental housing targeted to lower-income households”

This source of financing can be used for the construction, refinance, or rehabilitation of properties. The program originated in the Internal Revenue Service Office and is administered via federal regulations through the Department of Housing and Urban Development. The credits are then given to states for distribution, which rely on housing agencies to allocate the credits to specific developments. Initially, the program was set up to provide units for households between 20% and 60% of the area median income (AMI). “ The Consolidated Appropriations Act of 2018 (the federal omnibus appropriations bill passed in March 2018) amended this standard: now, households earning up to 80 percent of AMI are allowed in LIHTC-assisted units as long as the average income of all households in assisted units is 60 percent of AMI or below”  _[5] . The change can capitalize on the 80% AMI rents to cover the cost of producing units at 20% of AMI and incentivize a wider income spread within the building. In exchange for the tax credits, the property must adhere to investment regulations for a minimum of 15 years and affordability standards for a minimum of 30 years. Tax benefits can be rescinded if the properties fail to report on tenant demographics and building status. Two types of credits are offered – 4% and 9% - both can be applied to new construction, rehabilitation, or refinancing, but their uses and benefits differ.

I calculated the Simpson’s index in each Census Tract to create a composite measure of racial diversity. The formula was created to measure biodiversity but can be applied to studying racial diversity within a geographic area  _{[7, 8]} . A higher Simpson’s Index value indicates lower diversity, meaning a few race categories dominate the population. A lower value indicates greater diversity, suggesting the population is more evenly distributed across different race categories. For all the census tracts within the metropolitan statistical area, the median index value is 2.1, with a range of 4.2.

Census tracts with at least 1 LIHTC property across the MSA and within the city of Atlanta have a lower median household income, higher poverty rate, and lower Simpson’s index, indicating that it is more diverse than the MSA overall and than non-LIHTC census tracts. LIHTC census tracts within the city have the highest levels of diversity but the lowest average median household income. The average income spread between Atlanta census tracts with and without LIHTC properties is approximately $47,000, or 74% greater in non-LIHTC tracts. These data reflect the previous studies that show that LIHTC units are more often located in poorer neighborhoods but with increased levels of diversity.

In addition to measuring the differences in the tract's socioeconomic and racial composition, I also wanted to explore the variation in amenity accessibility. The resulting analyses use the centroid of LIHTC neighborhoods (census tracts) and non-LIHTC neighborhoods for a fairer comparison. I also performed the tests on LIHTC property locations for sensitivity testing. Unless stated otherwise, I used a closest facility network analysis to determine the distance from the centroid to the amenity. Because most metro Atlantans drive, the distances along the network represent miles traveled in a personal vehicle from the origin to the destination. I ran the analysis to determine the direction of travel to the points of interest and did not include any traffic barriers. I used Open Street Map data and Python code to generate a road network across the MSA for all subsequent routing inquiries.

Robust city planning emphasizes proximity to greenspaces, so the first analysis centers on the variation in distances to parks. I filtered the park data to exclude any golf courses or private greenspaces. I also excluded polygons below half an acre to ensure the accessibility map represented access to more than a pocket park or seating area. Building off of a study by Wang et al., I created points where the road network intersected the edge of the park polygon within 20 meters. These points represent entry points into the park and, more accurately, capture the distance between the origin and destination points. Where roads did not intersect with greenspaces, I used the centroid of the park feature as a proxy. I merged the two datasets to represent park point locations across metro Atlanta. I performed manual visual checks to ensure that entry points were present and aligned with existing road infrastructure for larger green spaces.

The metro-level analysis shows that LIHTC neighborhoods have greater proximity to parks than non-LIHTC neighborhoods. About 53% of LIHTC neighborhoods are within 1 mile of a greenspace, compared to 34% of non-LIHTC neighborhoods. The variation was fairly distributed across the first three distance thresholds. The sensitivity tests revealed that the centroid estimate was conservative, and nearly 70% of LIHTC properties are within a mile of a park. I altered the distance thresholds for the city because of the proximity to green spaces. Within Atlanta, there is less variation in access to parks. The data reflect the city's smaller relative size and inclusion of greenspaces in planning efforts.

Overall, it appears that the siting of LIHTC locations aligns with the federal incentives and statewide programs that promote developments in more diverse and accessible neighborhoods. While these neighborhoods tend to have lower median household incomes and higher poverty rates, the analyses revealed that they are often more proximate to the study's points of interest. Despite Atlanta's perception of racial polarity, there was still a difference between LIHTC and non-LIHTC neighborhoods. This index looks beyond just the percent of non-white populations, considering any racial concentration as "less diverse." The LIHTC neighborhoods truly represent a higher mix of backgrounds than non-LIHTC neighborhoods. Non-LIHTC neighborhoods are less diverse than the average at the city and metro levels. Park access varied more at the metro level as greenspace amenities became more scarce in the southern census tracts along the metro's periphery. On average, LIHTC neighborhoods are considerably closer to primary schools than non-LIHTC neighborhoods across the metro. The results indicate that the federal and state programs effectively encourage locating LIHTC properties in more accessible neighborhoods.

Because low-income housing tax credits are the most popular tool for producing affordable housing units, understanding their neighborhood context is essential to planning in Atlanta. It reflects not only federal regulations and statewide mandates but also the city and metro's approach to placemaking and economic development in affordable housing projects. Access to public parks, high-quality education facilities, and transit stations is essential. Retroactively, local governments and planning agencies can also work to improve the surrounding conditions of existing LIHTC properties to "move the opportunity" to existing residents by improving transportation networks, improving public schools, and enhancing green spaces. These interventions can improve the public realm without necessarily gentrifying existing neighborhoods. The heightened diversity of LIHTC tracts across the MSA presents an opportunity for placekeeping. 

Based on the results across the metro and previous work studying the impact of LIHTC developments on surrounding neighborhoods, this research supports the continued use of incentive programs that target specific areas for growth or maintenance. Georgia's Department of Community Affairs (DCA) 2024-2025 QAP states, "All Georgians deserve choice to live near quality services, schools, jobs, and transportation options"  _[9] . This can be seen across the metro when LIHTC properties are located in more amenity-rich areas. Their core plan also emphasizes local revitalization through quality construction and access to opportunities. They have also created target allocation requirements for 9% credits in new affordability outside the Atlanta metro and rural communities. Efforts to expand affordable housing across the state are a priority for DCA but can make accessibility more difficult as rural communities, by design, lack dense road connectivity.

 It is essential, however, to weigh the effects of specific incentives on the surrounding neighborhood. Going forward, more emphasis should be placed on siting developments in existing amenity-rich locations with diverse populations. However, this might not necessarily align with the currently difficult-to-develop areas. Spatially, studies can be repeated like this to identify amenity-rich locations to incentivize. Additional work should also be conducted on any gentrifying effects of developing LIHTC properties in lower-income areas. Specifically, in Atlanta, these properties tend to have higher density than the surrounding properties and can have a significant impact on the neighborhood's character.

Overall, this work helps identify differences in affordable housing neighborhoods across the Atlanta-Roswell-Sandy Springs metro. Many LIHTC properties are approaching the end of their 30-year affordability period. Incentives and support for more affordable housing, whether through new construction or rehabilitation, should holistically reflect what residents value in their community - access to schools, parks, transit, diverse population groups, and opportunities. As planners and GIS professionals, we must use our skills to study, support, and implement policies that improve our communities. This work is just one component of better understanding the spatial implications of our policy decisions.