The Devils in the Details: Key Issues in Implementing the New AFFH Rule

Posted by Dan Immergluck and Mindy Kao on July 5, 2016

For most of the Fair Housing Act’s history, its requirement to “Affirmatively Further Fair Housing” has been largely dormant. With the advent of the new AFFH rules in July 2015, however, there is some promise that this provision might be taken more seriously.

We’re not here to speculate on whether HUD will actually employ strong sanctions under the new rules. Many advocates recognize that the rules can have utility beyond punishing the worst actors: They may nudge some communities to do be more cognizant of fair housing barriers. They may help well-intentioned local officials utilize better data and metrics to look carefully at how they utilize scarce housing resources.

But there are some critical issues with AFFH implementation that are going to need addressing.

The New Rule

First, what is the new rule? The new AFFH rule was produced in response to criticism of weak implementation and enforcement of the AFFH provision, and in particular to criticisms of the process of producing the required local Analysis of Impediments to fair housing, or AIs. Since  AIs were not required to be submitted to or reviewed by HUD, compliance with the AI process was weak. Many local governments viewed AIs as a mere item on their compliance checklists, often outsourcing them, and frequently cranking them without much critical attention and effectively no community engagement. The GAO examined over 400 AIs and found that 29 percent of entitlement communities’ AIs were over five years old, many did not include timeframes for implementation, and others were not signed off on by top officials. More fundamentally, there is little sense that the AI process generated improvements in fair housing.

Spurred partly by the 2010 GAO report on AIs, HUD proposed a new set of AFFH rules in 2013; the final rule was issued in July 2015. It provided for the following changes:

1. More than 1,200 entitlement communities and 3,400 public housing authorities are now required to complete and submit an Assessment of Fair Housing (AFH) to HUD. The AFHs are required every five years.

2. A new set of HUD-provided data and a web-based GIS tool are to be used in the preparation of the AFH. Funding recipients can also add local data.

3. The AFH is to include the following components:

a. Summary of fair housing issues and capacity

b. Analysis of HUD-provided and possibly other data

c. Assessment of fair housing issues

d. Identification of fair housing priorities and goals

e. Strategies and actions

f. Summary of community participation

g. Review of progress achieved since submission of the prior AFH

HUD must review all AFHs and accept them as meeting AFH content requirements. Such acceptance, however, does not imply a certification that the recipient is satisfactorily affirmatively furthering fair housing.

Some Key Issues in Implementing the AFFH Rule

Beyond the issue of what sanctions HUD might actually apply if recipients are found to submit insufficient AFHs, or found not to be affirmatively furthering fair housing, another major issue that communities will need to address is their balance between “place-based” and “mobility” strategies. The final rule explicitly allows for employing both types of strategies, although this “both and” language leaves some observers uncomfortable.

Our focus here is less on these more ideologically charged issues, and more on some conceptual or pragmatic–but still critical, questions that are likely to arise during implementation:

1) HUD’s Capacity to Review and (Eventually) Accept (or not Accept) Thousands of AFHs

With over 4,600 entitlement communities and PHAs, HUD may eventually need to review 900 or more AFHs per year. (Joint or regional AFHs are possible, which may reduce the number of AFHs that HUD needs to review, but such AFHs may also be more complex.) HUD has 60 days to notify a locality that it is not accepting the AFH; otherwise the AFH is deemed “accepted.”

There are two classes of reasons for not accepting the AFH: (1) that it is inconsistent with fair housing or civil rights requirements; or (2) that it is substantially incomplete. Obviously, these two areas leave a great deal of uncertainty. Especially in the first five years of the new rule, the interpretation of “substantially incomplete” is likely to be a critical one. It may also be possible that some AFHs will go through multiple iterations before being deemed acceptable. These issues suggest that the administrative and staffing challenge for HUD will be a substantial one.

2) Addressing Regional Barriers to Fair Housing that Cross Local Boundaries

The new rules explicitly recognize that barriers to fair housing choice often operate across and not just within jurisdictions. HUD provides an option of Regional AFHs, which, conceptually at least, allow for addressing such issues. However, if a group of majority people of color, lower-income suburbs forms a Regional AFH, will such a collaboration effectively address regional barriers to fair housing choice if a key interjurisdictional barrier is exclusionary zoning in wealthier, predominantly white suburbs? Will such suburbs be likely to join a Regional AFH together with more diverse or less affluent communities? Geoffrey Leonard has argued for funding incentives to encourage Regional AFHs and penalizing localities that are offered the opportunity to join a Regional AFH but refuse. The viability of such a proposal remains unclear however.

A related challenge is that, even if AFHs are conducted on a regional basis, is policy or programming likely to follow suit? Housing and community development programs are generally not funded regionally, so will a Regional AFH change how programs are deployed? Moreover, potential barriers like zoning are very unlikely to be coordinated regionally simply due to a Regional AFH.

Beyond the issue of Regional AFHs, another question is whether the sorts of analyses suggested by the AFH guidebook are likely to be of much utility for relatively homogeneous communities. Dissimilarity indices (which indicate how different groups are distributed within a geographic area) are not likely to be a terribly useful indicator for a locality that is 95 percent African-American or 95 percent white. What a homogenous area should instead would depend on the context; A predominantly white, affluent suburb is likely to have greater ability than one whose residents are predominately low-income and/or people of color to affect its demographic future. The latter is more likely to be a victim of intrajurisdictional disparities and segregation, rather than a contributor to them.

(3) The Focus on RECAPs and the Suburbanization of Poverty

The thresholds used for defining racially and ethnically concentrated areas of poverty (RECAPs) are relatively arbitrary and are not well suited the changing, and increasingly suburbanized, nature of spatially concentrated disadvantage. As Alan Berube and Natalie Holmes have argued, the thresholds of 50 percent non-white and 40 percent poverty do not pick up the large increase in suburban neighborhoods with poverty rates in the 20 to 40 percent range.  Over time, some of these neighborhoods may become RECAPs, but they may have already suffered substantial decline and, especially because many suburbs have weak commercial tax bases, may be at a severe disadvantage in terms of public education and other conditions. In a recent study we did of one large, diverse suburban locality in the Atlanta metropolitan area, we found that only two census tracts out of 113 were classified as RECAPs. Yet our analysis showed large levels of segregation by race, ethnicity, and income within the county, with substantial numbers of poor and racially segregated tracts.

4) Developing Reasonable Measures for Assessment and Goal Setting

Despite HUD’s efforts to provide detailed instructions on the AFH process, including a 222-page guidebook, the construction of assessment measures and “metrics and milestones” remains very open-ended. HUD left the responsibility for establishing metrics, benchmarks, and goals to the funding recipients. This lack of prescriptive guidance on metrics and goals was likely based on both pragmatism and political viability.

Localities will therefore face the challenge of developing their own indicators or ratios, and establishing benchmarks for these variables. For example, how does a PHA assess whether the distribution of its Housing Choice Vouchers is affirmatively furthering fair housing? What sort of measures should it use? What sort of measures might it use to establish reasonable goals? Community engagement will certainly be helpful in assessing AFFH activity and setting goals. However, some set of objective and comparable measures that can be tracked over time will also be critical.

In a separate paper, we develop and demonstrate what we call the “subsidy ratio” method to assess and help set goals for the spatial distribution of housing subsidies (e.g., vouchers, or Low-Income Housing Tax Credits). Others will no doubt develop additional measures or indicators.

Some Optimism

The new AFFH rule is unlikely to mark a dramatic inflection point in the history of fair housing policy. However, beyond their utility to HUD and its enforcement of AFFH, the new rules – and the improved availability of important data – have the potential to provide a platform for government agencies, researchers, journalists, and, especially fair housing advocates to create new measures and analyses that will shine a brighter light on barriers to fair housing. By providing comparable data and a stronger call to employ such data, the new rule could help precipitate the sort of data-driven activism that helped increase attention to fair lending and community reinvestment after major improvements to the Home Mortgage Disclosure Act data in 1989. Data, by itself, is not a sufficient tool for enabling social change. However, it is often a critical, necessary ingredient. The AFFH rules have the potential to encourage civil society to increase public attention to issues of fair housing and equal opportunity in ways that could encourage greater housing opportunity.

(Photo credit: Pablo Viojo via flickr, CC BY 2.0)

About the author more »

Dan Immergluck is a professor in the School of City and Regional Planning at Georgia Tech in Atlanta, where he teaches and conducts research on issues of housing, community development, real estate and mortgage finance, and related topics. He is the author of three books, more than 40 scholarly articles, numerous book chapters and encyclopedia entries, and scores of applied research and policy reports. Mindy Kao Mindy Kao is a program associate for the Community Foundation for Greater Atlanta. Mindy works on the foundation’s community development initiatives and projects. She also supports the foundation’s community intelligence internal team in analyzing and disseminating community development issues, policies, and trends affecting the metro Atlanta region. Prior to joining the Community Foundation, Mindy worked as a graduate research assistant at the Federal Reserve Bank of Atlanta in the Community and Economic Development Department, conducting research and providing support on projects related to housing, neighborhood stabilization, workforce development, and community development finance.

If you like this article, please subscribe to Shelterforce in print or make a small donation to keep Rooflines strong.

COMMENTS

Fair Houser
12 Jul 16, 11:12 am

It is extremely wrongheaded to imply there is something wrong about jurisdictions “outsourcing” production of their Analyses of Impediments. HUD’s suggestion that the new Assessments of Fair Housing—which are nothing more than AIs on steroids—should be conducted in-house is extremely inappropriate and an invitation to fraud. An AI/AFH is like a financial audit—and everybody knows those should be conducted by an independent third party.

First, it is an inherent conflict of interest for a jurisdiction to conduct its own AFH or AI. Staff are severely compromised because an AI/AFH, conducted properly, is an critical examination the jurisdiction’s laws and practices. That puts staff in an untenable position of having to criticize the actions of their employer—both staff and elected officials. We have yet to see an purely internally-conducted AI that is remotely credible, much less competent.

Second, planning and community development staff are not trained in the areas needed to conduct an AI/AFH. These studies require a solid background in planning, sociology, lending practices as well as fair housing and zoning law—and I have yet to come upon a community development or planning department staffed with professionals with these skills. A competent AI/AFH requires understanding the factors that cause housing segregation and the tools that produce stable, integrated communities. Heck, even very few of the consultants that produce AIs exhibit that understanding—and it’s even worse among local government planning and community development staffs.

Finally, the new AFFH rule and AFH will essentially dilute the review of a community’s laws and practices because it requires examination of many topics that have only a peripheral relationship to affirmatively furthering fair housing and achieving the core goal of CDBG: fostering racial, ethnic, and economic integration. Recipients will continue to budget inadequate amounts of money to conduct a competent AFH and will continue to hire the 95% of consultants that produce whitewashed AI/AFHs that contain no analysis and who think that public education is the route to achieving AFFH. Having dealt with HUD on AIs for too long to remember, I have no doubt that HUD’s reviews of AFHs will be cursory, checklist types of reviews that fail to examine, much less understand, the substance in the AFH. Already Secretary Castro has hinted that HUD will be as lax in reviewing AFHs and it has been reviewing AIs.

So beware of the mess that HUD will inevitably make of the new AFFH rule and AFHs. It will not be pretty. Already just about everybody who has tried to use the new AFFH data tool is in despair over its poor quality, user-unfriendliness, outdated data, and confusing presentation of the data.

POST YOUR COMMENTS register or login