DRAFT Analysis of Impediments Study Policy Development Division, Office of Policy Development and Research January 27, 2009 DRAFT Contents Overview ........................................................................................................................ 3 Affirmatively Furthering Fair Housing ....................................................................... 3 Research Purpose ........................................................................................................ 4 Methodology .................................................................................................................. 4 Sample......................................................................................................................... 4 Four Dimensions of Assessment ................................................................................. 5 Analysis....................................................................................................................... 6 Findings: Background .................................................................................................. 6 Non-response .............................................................................................................. 6 Timeliness of AI Updates ........................................................................................... 7 Authorship of AIs ....................................................................................................... 7 Findings: Completeness and Quality of AIs ............................................................... 8 Role of Authorship ...................................................................................................... 8 Data Sources Used in AIs ........................................................................................... 9 Regional Variation .................................................................................................... 10 Improvement over Time? .......................................................................................... 10 Scores for Specific Dimensions ................................................................................ 11 Background and Purpose 11 Qualitative Data Collection and Analysis 11 Quantitative Data Collection and Analysis 12 Response and Accountability 12 Findings: Impediments Identified in AIs .................................................................. 13 Numbers of Identified Impediments ......................................................................... 13 Types of Impediments............................................................................................... 13 Conclusions .................................................................................................................. 15 Summary ................................................................................................................... 15 Policy Recommendations.......................................................................................... 16 Appendix A: HUD's AI Policy ................................................................................... 19 AFFH and Affordable Housing Guidance ................................................................ 19 HUD Evaluation and Enforcement of AIs ................................................................ 19 Appendix B: Assessment Instrument ........................................................................ 21 Appendix C: Field Visits ............................................................................................ 24 ATLANTA: Interview with Region 4 FHEO Director and staff .............................. 24 Interview with Director of Metro Fair Housing in Atlanta ....................................... 25 CHICAGO: Interview with Region 5 FHEO Director and staff............................... 26 Interview with Director of HOPE Fair Housing, Bernard Kleina ............................ 26 Interview with Chicago Housing Authority (CHA).................................................. 27 2 of 27 DRAFT Overview Affirmatively Furthering Fair Housing HUD requires jurisdictions to conduct an "Analysis of Impediments to Fair Housing" in order to be eligible for the Department's formula grant programs. The requirement arises out of a statutory requirement for HUD and its grantees to "affirmatively further fair housing."1 The Community Development Block Grant statute states that grants may be made "only if the grantee certifies to the satisfaction of the Secretary that...the grant will be conducted and administered in conformity with the Civil Rights Act of 1964 and the Fair Housing Act, and the grantee will affirmatively further fair housing...." 2 HUD's regulations on the Consolidated Plan expand on the meaning of the requirement to affirmatively further fair housing:3 (a) General. The following certifications, satisfactory to HUD, must be included in the annual submission to HUD.4 (1) Affirmatively furthering fair housing. Each jurisdiction is required to submit a certification that it will affirmatively further fair housing, which means that it will conduct an analysis to identify impediments to fair housing choice within the jurisdiction, take appropriate actions to overcome the effects of any impediments identified through that analysis, and maintain records reflecting the analysis and actions in this regard. An analysis of impediments to fair housing choice (AI) thus is the foundational element of an AFFH certification, setting the strategic direction and plan for a jurisdiction's further affirmative actions. Note that in the context of AFFH, "fair housing" means not only the Fair Housing Act itself, but other key laws including the Americans With Disabilities Act and Section 504 of the Federal Rehabilitation Act of 1973. HUD provides guidance on AIs in the Fair Housing Planning Guide.5 HUD suggests that jurisdictions conduct or update their AI at least once every 3-5 years (consistent with the Consolidated Plan cycle) (pages 2-6). In a joint CPD-FHEO memorandum to field office directors in 2005, jurisdictions were to be reminded, "especially at the beginning of a new Consolidated Plan five-year planning cycle," that "Although AIs are not submitted or approved by HUD, each jurisdiction should maintain its AI and update the AI annually where necessary."6 1 2 Section 808(e)(5) of the Fair Housing Act. Section 104 (b) of the Housing and Community Development Act of 1974. 3 24 CFR 91.225. Certifications. 4 (See definition of "certification"' in Sec. 91.5.) 5 http://www.hud.gov/offices/fheo/images/fhpg.pdf 6 "Analysis of Impediments to Fair Housing Choice Reissuance." See http://www.hud.gov/offices/fheo/library/finaljointletter.pdf 3 of 27 DRAFT Research Purpose This report is an effort to assess the content of these AIs in a nationally representative way and to identify opportunities to strengthen the AI process, the documents, so they can be improved to have a positive impact on equal housing opportunities. Assessments were conducted by staff of the Policy Development Division (Office of Policy Development and Research) and by Fair Housing and Equal Opportunity (FHEO). Additionally, field visits to two larger jurisdictions, Atlanta and Chicago, were used to validate and provide context to the results of document reviews. Notes from interviews with key persons in these jurisdictions are included in the Appendix. Methodology Sample The research team reviewed 45 AIs for this study. Beginning with a universe of all 1,216 grantees that received FY 2008 Community Development Block Grant (CDBG) allocations in FY 2008, several types of jurisdictions were excluded so as to focus on municipal grantees facing comparable situations: o o o 51 State programs (which include Puerto Rico but not the District of Columbia). 186 Urban County programs, of which 78 are Consortia under the HOME program. 31 Entitlement cities in U.S. territories (American Samoa, Guam, Marianas Islands, Puerto Rico, and Virgin Islands). The remaining sampling frame comprised 948 municipal Entitlement jurisdictions, including 60 Consortia (e.g., Boulder, Tucson, Sarasota, Tacoma). We stratified by the population of the jurisdiction in 2000 (less than 250,000, and 250,000 or more) and by the four Census regions. There are 188 jurisdictions in the Northeast region, 209 in the Midwest, 279 in the South, and 272 in the West. A random sample of 70 jurisdictions was drawn from this pool. FHEO asked fair housing directors in each HUD region to obtain Analyses of Impediments for the jurisdictions in their region. Of the 70 AIs sought by the research team, 45 jurisdictions provided AIs to review. Number of Jurisdictions Universe Selected Sample Responding Jurisdictions 948 70 45 Percent of Jurisdictions by Census Region Northeast 19.8% 31.4% 24.4% Midwest 22.0% 21.4% 31.1% South 29.4% 27.1% 20.0% West 28.7% 20.0% 24.4% The Northeast and Midwest, relative to the South and West, are over-represented among responding jurisdictions. This modest overweighting of certain regions, and--more importantly--the likelihood that jurisdictions' response patterns correlate with characteristics of their AIs, create "non-sampling error" in statistical estimates. Additionally, the relatively 4 of 27 DRAFT small sample contributes a measure of sampling error. Together, these factors imply that the results of this study should not be construed as more than generally representative of jurisdictions producing AIs. Four Dimensions of Assessment Each AI was reviewed by a two-person team using a standard assessment instrument. The instrument focused on whether key substantive elements were present in the documents. Although assessing the quality of the AIs was not the primary purpose of the analysis, qualitative judgments are inherent in determining whether a certain aspect of AI narrative satisfies the need represented by a given criterion. The assessment instrument is shown in Appendix B. The substantive elements conceptually fall into four clusters or "dimensions," as follows: o Background and Purpose. This dimension addresses whether the AI is transparent and effective in communicating to the public, with appropriate balance between clarity and level of detail. Aspects include the legal foundation, issues, and importance of fair housing--both generally and within the jurisdiction; the process by which the AI was prepared, including the extent of public input and feedback; and identification of the authors and major contributors. Qualitative Data Collection and Analysis. This dimension refers to the nonnumerical information that jurisdictions provide as evidence that impediments to fair housing may or may not exist. In many ways, this is core to an AI, as the 2005 CPD-FHEO memo states that "The [Fair Housing Planning] Guide defines the AI as a comprehensive review of a state's or entitlement jurisdiction's laws, regulations and administrative policies, procedures and practices." Data covered by this section include contextual information about local history, laws and administrative practices, population groups, and characteristics of housing markets and neighborhoods. Information gathered from the populace through surveys or other methods is covered here, and the geographic span of coverage relative to jurisdiction boundaries is noted as well. Both the presentation of the evidence and analysis of the evidence were assessed. Quantitative Data Collection and Analysis. This dimension complements the qualitative data dimension with an assessment of the presentation and analysis of numerical data. These data types include Census data about demographic and economic characteristics, housing market information including mortgage application and denial rates, and results from discrimination testing or victimization surveys. Another critical data source covered under this dimension is administrative data about local fair housing complaints and case outcomes, as well as related responses to Voluntary Compliance Agreements (VCAs).7 However, no AI provided information about VCAs, and the team was unable to locate a list of VCAs to provide independent information about their presence in sampled jurisdictions. Response and Accountability. The final dimension concerns the extent to which the AI demonstrates accountability and appropriate response to identified fair o o o 7 Section 504 Voluntary Compliance Agreements are agreements between negotiated by HUD and public housing agencies to address issues of inadequate accessibility of housing for persons with disabilities. 5 of 27 DRAFT housing issues. This includes timely updates of the AI, clear identification of impediments that are supported by the evidence, identification of recommended actions that are "reasonable, feasible, practical, and meaningful" to resolve impediments, and timelines, action plans, and public meetings to support accountability. Analysis After reviewing and scoring the AIs, the research team cleaned the resulting data for consistency. Some data were recoded into categorical variables to facilitate analysis. For example, analysts assigned the impediments identified in AIs to one of 17 categories, which are presented below. The team also prepared cross-tabulations of variables to explore the relationships between key factors affecting content and quality of the AIs. Findings: Background Non-response A substantial proportion of jurisdictions in the sample, about 35 percent, either were unable to produce an AI upon request, or were not sufficiently directed to submit their completed AI by the appropriate HUD field staff. To the extent that non-response is attributable to lowperforming jurisdictions, the results reported here may be more favorable than the entire sample would warrant. Only a minority of jurisdictions have an AI readily available to the public via the internet. Further, the research team found while attempting to obtain AIs from city offices that, in at least a few jurisdictions, the location or existence of AIs is not widely known and sometimes may be limited to the author. Citizens seeking to obtain AIs would not consistently find them readily available. Additionally, a few jurisdictions that responded did so by producing a newly-issued AI after a short delay, suggesting that they may have scrambled to pull together an AI in response to this research request. In one case, the newly-produced document was more like a responding letter than an AI report as prescribed. Jurisdictions that were unable to submit an AI for review provided a variety of explanations: o o The AI is currently being prepared; The jurisdiction is not required to prepare an AI because it is not a direct recipient of CDBG funds. (This explanation is suspect given that the jurisdictions were selected from a database of CDBG recipients.); An AI was not available because the jurisdiction has only been an entitlement community for approximately one year; The AI is available only in a bound volume. o o These non-response issues indicate that AIs are not consistently fulfilling their purpose in raising awareness and focusing efforts on improving fair housing. This shortfall suggests that HUD's reviews of AIs and monitoring of their implementation--for example by cross- 6 of 27 DRAFT checking with Consolidated Annual Performance and Evaluation Reports (CAPERs)-- perhaps should be strengthened.8 Timeliness of AI Updates In 2005, CPD and FHEO issued a joint memorandum reminding grantees that the required five-year update of Consolidated Plans also requires an update of their AIs. Of the 45 AIs received for review, the most frequent year of update was 2005 (27 percent). A large proportion also were updated after 2005 (20 percent). While a significant number of AIs (9 percent) had a 2008 date, it is unclear whether they actually had been updated in 2008 or simply dated 2008 due to FHEO's request of the AI for review. About one-fifth of AIs, 18 percent, were produced before 2000 and have not yet been updated. This indicates that AIs remain a low priority for numerous jurisdictions. Authorship of AIs The authorship of AIs becomes important for discerning patterns of content and quality. Of the 45 AIs, about three-fourths were prepared by a single author or organization. The majority of AIs were prepared by public officials of the jurisdiction. Consultants were employed in preparing a significant number of AIs, but in most cases the consultants were not explicitly identified as specialists in fair housing issues. Authorship of AIs Single author Jurisdiction alone Fair housing group alone Other consultant alone Other or Co-authors Author unspecified Number of AIs (n=45) 35 27 2 6 8 2 Proportion of AIs 78% 60% 4% 13% 18% 4% Local fair housing groups had a primary author role in only a few cases, serving as sole author of two AIs and coauthor of two AIs. Fair housing groups more frequently were consulted and provided data. One AI was conducted by the jurisdiction in conjunction with a "fair housing team" comprising representatives from the local housing industry rather than fair housing advocates. Two AIs did not specify who conducted the AI. 8 The 2005 CPD-FHEO memo reinforces that jurisdictions should report in their CAPER on actions taken to overcome the effects of impediments. 7 of 27 DRAFT Findings: Completeness and Quality of AIs The assessment instrument used by the review team contained 20 elements consisting of yes/no responses that could be tallied into a Completeness/Quality index. The index treats each element the same rather than making judgments about relative importance of various factors by weighting them. Nevertheless, the composite index corresponds well with the reviewers' subjective impressions of the overall quality of the AIs. Distribution of AIs by Points Scored 14 12 10 8 6 4 2 0 0 1-2 3-4 5-6 7-8 9-10 11-12 13-14 15-16 17-18 19-20 The team defined quality categories on the basis of the Completeness/Quality index. Based on these categories, 7 AIs (16 percent) were rated "Superior," credited with elements totaling 17-20 points; 16 AIs (36 percent) were rated "Acceptable," receiving 13-16 points; 14 AIs (31 percent) were rated "Needs Improvement," with 7-12 points; and 8 AIs (18 percent) were rated "Poor," receiving 0-6 points. In sum, just over half of the AIs are rated as acceptable or better. No AI received more than 18 points out of the 20 possible. The categorical data are presented in the following table. Role of Authorship As previously noted, most AIs were prepared by officials of the jurisdiction. The ratings of the resulting AIs on the Completeness/Quality index varied widely. Officials of three jurisdictions produced superior AIs, while six produced poor AIs. Of the rest, about half were "acceptable" and half "needed improvement." It seems probable that AIs with unspecified authorship also were prepared by jurisdictions, with the "poor" ratings for these AIs reflecting the lack of basic documentation, among other deficiencies. No AIs prepared by co-authors were rated as poor, although ratings across the higher categories were similar in proportion to single-author AIs. The absence of the worst products from the co-authors group suggests that there may be benefit in encouraging jurisdictions to seek partners when authoring AIs, or at least suggesting that they consider it. 8 of 27 DRAFT Number of AIs by Completeness/Quality Rating Authorship of AI Poor, 0-6 pts Single author Jurisdiction alone Fair housing group alone Other consultant alone Other or Co-authors Author unspecified Total number Total percent * 2 8 18% 14 31% 16 36% 7 16% 3 6 6 Needs Improvement, 7-12 pts 11 10 1 Acceptable, 13-16 pts 13 8 1 4 3 2 2 Superior, 17-20 pts 5 3 Total 35 27 2 6 8 2 45 100% * Percentages may not sum to 100% due to rounding error. Data Sources Used in AIs Most AIs included Census or other data sources to inform about community conditions, such as demographic composition, household incomes, housing tenure, and housing stock. Maps were frequently used to present these data in a spatially relevant way. Over 78 percent of AIs included Census data, often in combination with other data sources. Number of AIs by Quantitative Data Sources Used Authorship of AI Any AI with Local Paired Testing data Other AIs with Census Data Census alone With HMDA With Other With HMDA and Other Other or Unspecified Data Total Single author Jurisdiction alone Fair housing group alone Other consultant alone Other or Co-authors Author unspecified Total number Total percent * 4 2 2 1 5 11% 5 4 1 1 6 13% 4 3 1 1 9 9 9 7 2 5 4 4 35 27 2 6 8 1 5 11% 10 22% 14 31% 1 5 11% 2 45 100% * Percentages may not sum to 100% due to rounding error. 9 of 27 DRAFT Another common data source was Home Mortgage Disclosure Act (HMDA) data, which lending institutions report about the mortgage applications they receive and loans they award. HMDA data are useful for assessing fair housing conditions related to homeownership, and 42 percent of AIs included these data. About one out of ten jurisdictions (11 percent) included local paired testing data in their AIs. Paired testing is the most rigorous available method for identifying disparate treatment in the housing market.9 The two AIs authored by fair housing organizations both included testing data that they had generated. Although not readily apparent in this table, nearly all of the AIs with testing data also reported both Census and HMDA data. Regional Variation Some regional variation in quality and completeness of AIs was observed. In general, AIs of Midwestern jurisdictions received 1.2 points less than the national average, and Western jurisdictions received 1.3 points more than the national average. Jurisdictions by Census Region Northeast Mean Score 11.9 Midwest 10.4 South 11.8 West 12.9 Total 11.6 These results mask substantial variation within regions. California jurisdictions produced several superior AIs, as well as one inadequate AI. Improvement over Time? The scores of the AIs do little to suggest that jurisdictions are systematically and consistently improving the content and quality of AIs as they bring them up to date. The bottom axis of the scatter-plot below shows the year in which the sampled AIs were produced. The vertical axis shows the scores that the research team awarded to the AIs. Fewer than 45 dots are shown because some represent several AIs with the same values. Solely for purposes of this exhibit, the three AIs that were undated are assigned the year 1990 so that they appear on the left side. 9 Paired testing involves researchers, matched on relevant characteristics except membership in a protected class, who engage in real estate transactions to identify disparities in their treatment attributable to discrimination. See for example "All Other Things Being Equal: A Paired Testing Study of Mortgage Lending Institutions" and "Discrimination in Metropolitan Housing Markets: National Results from Phase 1, Phase 2, and Phase 3 of the Housing Discrimination Study (HDS)" at http://www.huduser.org/publications/hsgfin.html. 10 of 27 DRAFT The chart suggests that jurisdictions who produced AIs a decade ago were more likely to score 15 points or higher than those produced more recently. Of concern, those produced since FHEO issued the notice in 2005 display greater variance in completeness and quality, with scores spanning the full range from 0-18 out of 20. The evidence suggests that there may be need for greater resource commitments. 20 15 Aggregate score 10 5 The pattern also suggests that characteristics of the jurisdictions may affect both timing and quality of the AIs: 0 i.e., "self-starting" jurisdictions 1990 1995 2000 2005 2010 that updated their AIs during date of AI (missing shown as 1990) 2001-2004 were less likely to have submitted very weak products. Lagging jurisdictions that were motivated to update AIs only by the FHEO notice in 2005 or because of the five-year Consolidated Plan update cycle may have been more likely to make superficial efforts that resulted in low-rated documents. Scores for Specific Dimensions The four conceptual dimensions of completeness and quality that were previously described provide a framework for summarizing the results. Background and Purpose The majority of AIs scored high on the elements categorized as Background and Purpose. This dimension focuses on communicating who was involved in preparing the AI and clarifying the purpose of the AI. The average AI score was 2.8 points out of the 4 points available in this category. Four jurisdictions received zero points for this dimension. One of these AIs consisted of a four-page update to a previous AI that was embedded within the jurisdiction's Consolidated Plan. Another submission was a compilation of action plans from four years--an approach approved by the FHEO field office. In such cases, the score may or may not fairly represent the quality of the jurisdiction's effort. Qualitative Data Collection and Analysis AIs completed by jurisdictions varied substantially in terms of quality of qualitative data collection and analysis. The average AI scored 4.1 points out of 8 elements that could be tallied. The most prevalent shortcoming was an inadequate discussion of the rental housing market. 11 of 27 DRAFT A number of jurisdictions undertook survey-based data collection, with uneven results. The persons surveyed did not consistently reflect the population of protected groups or potential victims of discrimination, and AIs frequently provided poor documentation of survey processes and the range of responses. Overall, AIs completed by consultants seemed to do a better job with the inclusion of qualitative data collection and analysis. Quantitative Data Collection and Analysis AIs completed by jurisdictions also varied widely in terms of quality of quantitative data collection and analysis, while consultants typically did a superior job on this dimension. The average AI scored 2.1 points from the 4 elements tallied for quantitative data issues. The average score does not reflect the breadth of the range of effort that authors put into AIs. The documents range from those devoid of any quantitative data to those overburdened with numerical tables of Census statistics. The best AIs presented the statistics in a disciplined way while explaining how the data are relevant to fair housing. A variety of supplemental data sources also were presented, often quite usefully. A number of AIs included Home Mortgage Disclosure Act data, which are useful and typically report racial disparities in mortgage application rates and denial rates, but are limited by a lack of credit score data. One consultant studied real estate ads over a three-month period, including 5,000 classified ads and their depictions of diversity. Response and Accountability The average AI received 2.7 out of 5 points awarded for response and accountability. Again, AIs completed by public officials of jurisdictions displayed a wide variation in quality. Consultants seemed to do a better job of identifying impediments appropriately based on the evidence presented in the AI and provided meaningful recommendation to overcome impediments. The role of authorship by jurisdictions versus contracted consultants may have the clearest and most justifiable influence on this dimension of AIs. Consultants frequently appeared constrained by an inability to negotiate action plans among the range of responsible stakeholders, or to incorporate feedback from public response to a substantially-completed document if their contract did not provide for such revision. Public officials may in some cases be better positioned to make an AI an action-oriented document. However, jurisdiction-authored documents likewise were very uneven in their response to identified impediments. 12 of 27 DRAFT Findings: Impediments Identified in AIs Numbers of Identified Impediments The average number of impediments to fair housing identified in the reviewed AIs was 4.7. The greatest number of identified impediments was 16. At the other extreme, seven jurisdictions (16 percent) failed to explicitly list any impediments. In some AIs, impediments were mentioned in a discussion of recommendations rather than being listed separately or in conjunction with supporting evidence. Review teams made efforts to document any specifically identified impediments even if they were embedded in a broader discussion. The average number of impediments included in AIs varied by author type. The mean number of impediments for AIs completed by officials of the jurisdiction was 3.8. Consultants who authored AIs identified 4.7 impediments on average, and the two fair housing organizations that authored AIs found 9.0 impediments on average. Other types of authors or coauthors found 6.5 impediments on average. Authorship of AIs Single author Jurisdiction alone Fair housing group alone Other consultant alone Other or Co-authors Author unspecified Total Average Number of Impediments Identified 4.2 3.8 9.0 4.7 6.5 5.0 4.7 Types of Impediments The research team found that categorizing impediments required a substantial degree of judgment. First, the review teams were required to document the identified impediments, which often involved paraphrasing them. Then the analysis team had to sort them into reasonable categories. The documentation of impediments was made more difficult by the failure of AI authors to specifically identify impediments. Additionally, a number of AIs discussed impediments, or potential impediments, or sometimes even non-impediments, in connection with their recommendations rather than with their supporting evidence. These complexities introduce a degree of subjectivity to the following results. o Affordable Housing. Impediments related to a lack of affordable housing were the most prevalent type of impediment, found in 38 percent of AIs. Of the seventeen AIs identifying an affordable housing impediment, four listed multiple impediments related to affordability. A number of these impediments may have been identified inappropriately. 13 of 27 DRAFT Lack of affordable housing is indeed related to fair housing, but HUD has cautioned jurisdictions against finding affordable housing an impediment without analyzing underlying factors. Improper findings focused on affordability can become a distraction from issues related more directly to fair housing law. In extreme cases a jurisdiction can ignore fair housing issues completely and improperly cite a generic lack of affordable housing or lack of income as the sole impediment in their area. For a more detailed explanation of this distinction, see Appendix A on this topic. o o Housing Discrimination. Twelve AIs (27 percent) identified housing discrimination in general as an impediment to fair housing. Lending Discrimination/Predatory Lending. Eleven AIs (24 percent) identified impediments involving predatory lending or disparities in lending applications or lending outcomes between races. Redlining of Lending or Insurance. Two AIs (4 percent) listed impediments specifically addressing redlining. One involved mortgage lending and the other addressed home insurance. Enforcement. Twelve AIs (27 percent) had at least one impediment involving the inadequacy of fair housing laws, enforcement structures or resources in their jurisdiction. Of these AIs, four listed multiple impediments--including one AI listing four impediments of this nature. Education. A variety of education-related impediments were identified. Fourteen AIs (31 percent) listed an impediment involving a generalized or unspecified need for fair housing education. Six AIs (13 percent, with some overlap) identified a need for fair housing education specifically for the housing and real-estate industry. Four AIs (9 percent) had an impediment involving needed fair housing education for Limited English Proficiency (LEP) households. Eight AIs (18 percent) had impediments addressing the need for homeownership counseling and education. Finally, three AIs (7 percent) included impediments related to educating the public to distinguish between fair housing law and landlord/tenant law. Accessibility. Seventeen AIs (42 percent) identified impediments based on accessibility of housing for persons with disabilities, with two AIs finding multiple accessibility impediments. Nevertheless, 58 percent of jurisdictions did not address accessibility as an impediment. Inadequate/Substandard Housing. Twelve AIs (27 percent) included an impediment involving inadequate or substandard housing or unavailable housing. Of these AIs, five identified multiple impediments--up to four of this type. Public and Assisted Housing and Housing Mobility. Six AIs (16 percent) included impediments involving public and assisted housing. Several addressed limited housing mobility for low income (minority) households and the fact that Section 8 is not a protected class (source of income) under the Fair Housing Act/local fair housing ordinance. Housing Location and Neighborhoods. Eight AIs (22 percent) included an impediment that addressed spatial and neighborhood issues. These impediments included mismatch between (affordable) housing and job opportunities, lack of adequate public transportation, neighborhood segregation, and neighborhood conditions. o o o o o o o 14 of 27 DRAFT o Regulatory Barriers. Nine AIs (20 percent) had at least one impediment addressing regulatory barriers. Four jurisdictions identified multiple impediments of this type, including two AIs with three impediments and one with four impediments. NIMBYism. Three AIs (7 percent) included as an impediment "Not in My Backyard" (NIMBY) responses by neighborhoods toward proposed developments of affordable or special needs housing. Other. Seventeen AIs (38 percent) listed impediments other than those listed above, of which 11 AIs included multiple impediments of this type. o o Conclusions Summary This review of AIs showed that there was a wide variation in completeness and overall quality of the documents. As stated, 45 of 70 AIs selected in the national sample were received. Some of the jurisdictions from which an AI could not be obtained may have produced one at some point, but reviewers were unable to obtain them. Thus, the total number of the sample that did not have an AI is unknown. Still, this represents a significant proportion of jurisdictions, which is cause for concern. In addition, many of the AIs obtained were completed over ten years ago and need to be updated. However, there is evidence that recently-produced AIs are not consistently improving upon the quality and completeness of AIs produced in the 1990s. A number of acceptable and even superior AIs have been produced since 2005, but the range of quality is quite broad. The data hints that "self-starting" jurisdictions who updated their AIs before the 2005 HUD Notice reminding grantees that the required five-year update of the Consolidated Plan also requires an update of their AI. The review also showed that a sizable proportion of the AIs reviewed did not contain key aspects recommended for inclusion by the Fair Housing Planning Guide. While the Guide provides only recommendations for jurisdictions and not binding requirements, it was clear that some AIs were completed in a cursory fashion only. However, in a few cases--for example with smaller jurisdictions--a complete consideration of all the Guide's recommendations may not have been necessary or appropriate. There are some inherent limitations in assessing where particular AIs would rank according to a subjective qualitative assessment. This is due to inherent subjectivity of assessing complex policy issues in connection with varying sizes and characteristics (demographics, economic, etc.) of jurisdictions, as well as variations among reviewers who may have differing judgments of the document being reviewed. Despite these qualifications, reviewers found that numerous AIs were produced with considerable effort and had very high levels of quality. Reviewers also found several outstanding examples of AIs, indicating that identifying such examples to be used as models could be explored. The AI is a local document, and to be effective it not only has to cover a variety of issues completely, but also needs to be internalized by local leaders in order to drive planning and funding decisions. Local involvement in developing the document and public participation is important. 15 of 27 DRAFT The validity of the specific fair housing impediments identified by different AIs is subject to judgment. Just as fair housing violations can involve either disparate treatment or disparate impact, impediments also can involve community conditions that reduce housing choices for protected groups even though there is no disparity as a primary feature. Thus, many communities identify affordability, and factors that reduce affordability, as impediments. Affordability was the most frequently identified impediment, and indeed fair housing and the need for affordable housing options are inextricably entwined. In some cases, however, because AIs often represent self-reporting, jurisdictions with persistent and ingrained discrimination problems may find it easier to focus on economic factors rather than more sensitive social factors that may indicate fair housing violations. In such cases, a blanket statement that lack of income or affordable housing was the only impediment present, while ignoring the issues with fair housing compliance, could constitute a disregard of the AI requirement. It is apparent that the process used to identify impediments within communities bears strongly on the results documented in the AI. Surveys involving public officials, the housing industry, neighborhood groups, or fair housing advocates can produce widely varying results in the same jurisdiction. This points to the importance of ensuring that the process is broadly based and balanced, and includes opportunity for public feedback. There is cause for optimism, as many of the AIs evidence a significant effort in time, effort, and expense. Serious AIs drew on data collection, public surveys, focus groups, expert analysis, and involvement from many parts of the community including businesses, nonprofits and members of the public at large. Many jurisdictions have obviously taken the AI planning process very seriously. HUD needs to assess and work with our state and local partners, governmental and private, to explore options for improving the AI process and taking steps for translating it into positive action on the fair housing front. Some considerations for policy steps are considered next. Policy Recommendations This AI Review suggests several possible options for improving the Analysis of Impediments. First, in some cases, AIs were apparently not performed at all. The AI is required by HUD regulations that implement the statutory obligation for grantees to use federal housing funds to "affirmatively further fair housing." HUD needs to determine appropriate means for achieving compliance with this requirement. In other cases, the review showed that AIs varied greatly in terms of completeness, for instance in addressing key issues recommended for consideration in the Fair Housing Planning Guide. As stated above, there may be various reasons for this, including valid reasons such as different types of communities of various sizes that may or may not need to address the full range of the Guide's recommendations. In some cases, however, the reviewers found that the AIs were not appropriately complete documents to implement the requirement itself. Perhaps for the majority of jurisdictions, enhanced HUD guidance and assistance would increase completeness and quality. This assistance could take the form of providing better access to federal data tools, broad-based training options or in some cases perhaps more indepth technical assistance. Reviewers found that the vast majority of jurisdictions that identified the funding source for conducting the AI was the CDBG program (15 percent can 16 of 27 DRAFT be used for administrative and planning costs and 15 percent can be used for public services). CDBG funding can be tight, however, due to the need to maintain existing public services, so HUD should be mindful of the funding and capacity limitations of the over 1,200 CDBG entitlement jurisdictions.10 HUD should consider other possible revenue streams, such as dedicated technical assistance funds, or appropriations for state Fair Housing Assistance Program (FHAP) grantees to assist localities under their jurisdiction. In the area of guidance, the Fair Housing Planning Guide remains useful. However, there may be value in updating it, drawing on studies of fair housing compliance since it was issued in 1995. HUD also should consider a basic fact with AIs--that jurisdictions are not currently required to submit them to HUD. The current HUD practice is described in the Fair Housing Planning Guide: 2.13 HUD EVALUATION AIs will not generally be submitted to HUD for review. Instead, as part of the Consolidated Plan performance report, the jurisdiction will provide HUD with a summary of the AI and the jurisdiction's accomplishments during the past program year. The Department could request the AI in the event of a complaint and could review the AI during routine on-site monitoring.11 HUD does have the enforcement authority to decertify a jurisdiction's Consolidated Plan if the AI is inadequate. The 2005 CPD-FHEO memo describes the process as follows: HUD can require the submission of an AI in the event of a complaint or as part of routine monitoring. If, after reviewing all documents and data, HUD concludes that (1) the jurisdiction does not have an AI; (2) an AI was substantially incomplete; (3) no actions were taken to address identified impediments; (4) the actions taken to address identified impediments were plainly inappropriate; or (5) the jurisdiction has no records, the Department would notify the jurisdiction that it believes the certification to be inaccurate, or, in the case of certifications applicable to the CDBG program, the certification is not satisfactory to the Secretary. In connection with this review, HUD will consider whether a recipient has made appropriate revisions to update the AI. HUD's approval and enforcement role, currently the subject of an internal review, needs to be examined carefully (see below). A possible first step could be simply to implement a submission requirement. HUD would need to devote staff and possibly information technology resources if a submission requirement is adopted. A basic step following submission of AIs to HUD, short of review and enforcement, could be to make all AIs available for public review--for instance through an internet-based clearinghouse. Providing public access to this new source of information could have benefits for a variety of people and organizations. The AIs would be of interest to State FHAP agencies, as well as regional planning bodies and housing organizations. Public dissemination also may help build public awareness of fair housing laws, in terms of how 10 In connection with this issue, an Atlanta official suggested that costs for preparation of the AI should be released from the CDBG cap on administrative activity. 11 See Appendix A for the full excerpt on HUD's review and enforcement authority, as described in the Fair Housing Planning Guide (1995): http://www.hud.gov/offices/fheo/images/fhpg.pdf 17 of 27 DRAFT they interact with local housing conditions. At a minimum, disparities in quality and completeness would become apparent to anyone who compared more than a few AIs. A more decentralized approach would be to strengthen requirements for jurisdictions to use technology for public review and dissemination of AIs on the internet. Considerations for more widespread HUD review, approval and/or enforcement need to take into account the issues, discussed in the summary above, concerning subjectivity and the nature of the AI as a locally driven document that is part of the block grant consolidated planning process. HUD also would have to ensure that necessary staff resources would be available for such an exercise. HUD would need to establish guidelines for staff review that balance objectivity and subjective judgments appropriately. Impediments often are too general to measure precisely, and mitigating steps typically are complex in design and implementation. A review would need to observe the fact that AIs are essentially local planning documents, and that options and resources available to localities vary widely. 18 of 27 DRAFT Appendix A: HUD's AI Policy AFFH and Affordable Housing Guidance Chapter 5 of the Fair Housing Planning Guide attempts to draw a distinction between the related concepts of affordable housing and affirmatively furthering fair housing (AFFH) actions: AFFH and Affordable Housing Clarification of the distinction between AFFH actions and affordable housing activities is often necessary. The two concepts are not equivalent but they are also not entirely separate. When a jurisdiction undertakes to build or rehabilitate housing for low- and moderate-income families, for example, this action is not in and of itself sufficient to affirmatively further fair housing. It may be providing an extremely useful service by increasing the supply of decent, safe, and sanitary affordable housing. Providing adequate housing and improving existing neighborhoods are vital functions and should always be encouraged. Additionally, the provision of affordable housing is often important to minority families and to persons with disabilities because they are disproportionately represented among those that would benefit from low-cost housing. When steps are taken to assure that the housing is fully available to all residents of the community, regardless of race, color, national origin, gender, handicap, or familial status, those are the actions that affirmatively further fair housing. (Page 5-4.) In addition, Chapter 2 of the Fair Housing Planning Guide states the following: Use a Fair Housing Perspective Where the community planning and development perspective looks directly at needs for housing and possible barriers to meeting those needs, the fair housing perspective focuses as much on the causes of needs of groups or persons protected by the Fair Housing Act as it does on the needs themselves. Thus, the explanation of barriers to affordable housing to be included in the Consolidated Plan may contain a good deal of relevant AI information but may not go far or deep enough into factors that have made poor housing conditions more severe for certain groups in the lower-income population than for others. Jurisdictions should be aware of the extent to which discrimination or other causes that may have a discriminatory effect play a role in producing the more severe conditions for certain groups. (Page 2-20.) HUD Evaluation and Enforcement of AIs Chapter 2 of the Fair Housing Planning Guide discusses HUD's current evaluation and enforcement of AIs. (Page 2-24). 2.13 HUD Evaluation AIs will not generally be submitted to HUD for review. Instead, as part of the Consolidated Plan performance report, the jurisdiction will provide HUD with a summary of the AI and the jurisdiction's accomplishments during the past program year. The Department could request the AI in the event of a complaint and could review the AI during routine on-site monitoring. In addition: 19 of 27 DRAFT If HUD's year-end review suggests that the AI or actions taken were inadequate, HUD could require submission of the full AI and other documentation. If, after reviewing all documents and data, HUD concludes that the AI was substantially incomplete or the actions taken were plainly inappropriate to address the identified impediments, the Department would provide notice to the jurisdiction that it believes the AFFH to be inaccurate and would provide the jurisdiction an opportunity to comment. If, after the notice and opportunity to comment is given to the jurisdiction, HUD determines that the AFFH certification is inaccurate, HUD will reject the certification. Rejection of the certification renders the Consolidated Plan substantially incomplete and constitutes grounds for HUD to disapprove the Consolidated Plan. HUD will work with the jurisdiction to determine actions necessary to make the certification accurate and the Consolidated Plan complete. The actions may take the form of a special assurance which describes the actions to make the AI complete or describes actions to overcome the effects of identified impediments and which includes a timetable for accomplishing these actions. NOTE: A jurisdiction cannot receive its CDBG, HOME, ESG, or HOPWA program grants until the Consolidated Plan is approved. 20 of 27 DRAFT Appendix B: Assessment Instrument Worksheet for Rating "Analyses of Impediments" to Fair Housing Assessment Dimensions and Elements Unless otherwise specified, enter Y for Yes and N for No. Jurisdiction: Reviewer: A. Background and Purpose A. 1 A. 2 Who conducted the AI? A) Jurisdiction, B) Fair Housing Group, C) Partnership of A + B, D) Other, E) Not specified. Is the AI transparent? Clarifies the purpose for lay readers. Explains the importance of fair housing. Explains the relationship of fair housing to local housing and community development activities, including federally sponsored activities. Did the process of preparing the AI include involvement and input from the public? Describes the extent of public involvement in study design and data collection. Describes opportunities for public review of draft products. Describes adjustments made in response to reflect comments. Is the AI's scope appropriate in level of detail and focus? Analysis demonstrates careful attention to locally relevant issues based on the jurisdiction's Housing Needs Analysis rather than a cursory or "boilerplate" presentation. Is the AI's overall organization and presentation useful? Non-technical readers are likely to find the evidence and assessment clear. Reviewer's comments/notes about this dimension of the AI. -- . . A. 3 . A. 4 . A. 5 A. 6 . . B. Qualitative Data Collection and Analysis B. 1 B. 2 Does the AI's community profile provide a local context? Discusses relevant history, demographics, and unique local conditions that have a bearing on fair housing. Does the AI provide a review of local laws, regulations, and administrative policies, procedures and practices? Describes protections and potential impediments to fair housing arising from local ordinances and regulations. Identifies and explains administrative practices that protect or create impediments. Does the AI provide an assessment of how those laws, policies and practices affect the location, availability and accessibility of housing? Examines the effects of local governmental actions on the availability of housing for all protected groups. Assesses impact of regulatory barriers, including zoning and land use. Discusses prevalence and effect of NIMBYism for affordable housing and assisted living/group homes. -- . . B. 3 . 21 of 27 DRAFT B. 4a Does the AI discuss the rental housing market in its jurisdiction? Assesses availability of suitable housing for elderly persons and families, accessibility to persons with disabilities, location and condition of the community's affordable housing stock, etc. If Yes, does the AI refer to Affirmative Fair Housing Marketing Plans for multifamily housing and Public Housing Agency Plans for public housing, including tenant selection plans? Does the AI discuss the homeownership market in its jurisdiction? Considers potential impediments related to real estate brokers and shopping for homes. Considers housing finance issues including subprime and predatory lending, mortgage application processes, and approval or denial of applications. Which key persons that were interviewed in preparing the AI? (List all that apply) A) Fair Housing Group B) Advocacy Group C) Local Government Officials D) Housing Developers E) Public F) Other. Discusses identification and selection of a population of key persons with special knowledge of fair housing issues and impediments. What geographic area does the AI cover? A) specific jurisdiction/community, B) metropolitan area, C) Other regional area spanning jurisdiction boundaries. Considers regional implications for fair housing of inter-jurisdictional issues such as exclusionary zoning, annexation, and coordination of transportation and land use. Does the AI describe methods used to obtain useful information about local conditions? Potential methods include written surveys, interviews, focus groups. Reviewer's comments/notes about this dimension of the AI - including other types of qualitative analysis included in AI. . B. 4b B. 5 . . B. 6 . B. 7 . B. 8 B. 9 . . C. Quantitative Data Collection and Analysis C. 1 Does the AI identify sources of quantitative data? (List all that apply): A) Census data, B) Housing Discrimination Survey, C) Home Mortgage Disclosure Act, D) Local paired testing data, E) Other (specify in C.7). Identifies significant targeted surveys and studies, potentially including local paired tests for disparate treatment, survey of fair housing victimization (i.e., discrimination) rates, or survey of fair housing awareness rates. Surveys and studies demonstrate sound, statistically representative methodology. Does the AI explain how the quantitative data are relevant and explain implications? Does the AI include a demographic profile? Summarizes census population data about race, ethnicity, age, family status. Considers local concentrations of various groups. Does the AI include data on fair housing complaints by cause and by outcome in jurisdiction? Summarizes available data on complaints about discrimination submitted to local fair housing agencies and HUD. Analyzes the distribution of reasons that complaints were filed. Evaluates the rate at which complaints are conciliated or become formal claims. Assesses trends in local complaints. Does the AI address previous Section 504 Voluntary Compliance Agreements? Identifies a public housing agency in the community that has a Section 504 VCA for housing persons with disabilities. If yes, does the AI address the housing needs of persons with disabilities in accord with the VCA? -- . C. 2 C. 3 C. 4 . . . C. 5a . C. 5b . 22 of 27 DRAFT C. 6a Does the AI address previous or existing fair housing case outcomes? (List all that apply): A) local Fair Housing Settlement, B) a Consent Decree, C) an Administrative Law Judge (ALJ) decision, D) Other identified outcome, E) None identified. If A, B, or C, does the AI address the provisions of the settlement, consent decree or ALJ decision? Reviewer's comments/notes about this dimension of the AI. . C. 6b C. 7 . . D. Response and Accountability D. 1 D. 2 When was the AI updated? Enter year (YYYY), or 9999 if not identified. Does the AI identify impediments appropriately based on the evidence presented? Impediments must be related to issues of fair housing (lack of affordable housing is not sufficient) and avoid omissions and distortions. Does the AI provide actions to overcome the impediments identified? Does the AI provide recommendations that are reasonable, feasible, practical, and meaningful? Does the AI provide an action plan with achievable tasks, sufficient detail, appropriate timelines, and assigned responsibilities? Were the AI's results and planned actions presented in a public forum? Reviewer's comments/notes about this dimension of the AI. -- . . D. 3 D. 4 D. 5 D. 6 D. 7 . . . . . E. Impediments Identified (use [brackets] to paraphrase) E. 1 E. 2 E. 3 E. 4 E. 5 E. 6 E. 7 E. 8 E. 9 E.10 E.11 E.12 E.13 E.14 E.15 E.16 Impediment 1 Impediment 2 Impediment 3 Impediment 4 Impediment 5 Impediment 6 Impediment 7 Impediment 8 Impediment 9 Impediment 10 Impediment 11 Impediment 12 Impediment 13 Impediment 14 Impediment 15 Impediment 16 . . . . . . . . . . -- 23 of 27 DRAFT Appendix C: Field Visits Background on Field Visits As part of the AI study, researchers made site visits to Atlanta and Chicago in September, 2008, and interviewed field office directors and private fair housing organizations. Research questions included the following: o o o o o o o How is the relationship between local, state and federal fair housing laws perceived at the local level? What is the level of cooperation in enforcement issues across the metro area? What is the process of developing the AI? How useful is the AI in practice? Is it used to motivate real changes in policies, procedures, practices, etc.? How seriously do local CDBG recipients take the AI? How has public comment been integrated in the AI? What are the main areas of public concern? Should local fair housing organizations play more of a role in the development of the AI? If so, how could they play more of a role? The following notes summarize the field interviews. ATLANTA: Interview with Region 4 FHEO Director and staff The interview was loosely structured so as to respond to local issues. The field staff generally find that AIs fail to fulfill their potential, and questioned whether a formal study of this type would prove worthwhile. AIs are simply one piece of a larger fair housing puzzle. Oversight of jurisdictions in completing and complying with AIs in Atlanta is quite limited, although the office is trying to improve efforts. Resources for monitoring. Severe resource constraints are preventing the office from investigating and resolving known fair housing issues. A key factor is that FHEO staff has undergone attrition of 40 positions while responsibilities for oversight of FHIP grantees has devolved from Headquarters to the field. The geographic allocation of staff resources also plays a role, because FHEO's staffing in the field is not proportional to the number of FHIP and FHAP agencies. Chronic deficiencies in PHA plans. The chief concern of Region 4 FHEO is with PHA plans, especially as they are being automated. The staff said that although the PHA planning process and automation is good, there is a lack of FHEO personnel and implementation to follow up on the AI aspects. Specifically, the staff viewed PHAs' "affirmatively furthering fair housing" certifications (AFFIs) to be meaningless for the vast majority of PHA plans. The big issue was the inability to apply pressure on PHAs to address long-standing violations of fair housing standards. Instead, FHEO can challenge AFFIs only after a fair housing complaint is received. 24 of 27 DRAFT As an example of ineffective AFFIs, the staff cited lack of accommodation of persons with disabilities in the Gainesville HA. After the PHA faced three fair housing charges in two years, FHEO was able to conclude a voluntary compliance agreement with the PHA. Yet, one disabled resident had spent 10 years without being accommodated with an accessible shower. Atlanta HA residents also have filed a race-based compliant, as well as a disability complaint regarding public housing demolition, which is an extensive and ongoing process in Atlanta. Lack of FHAP coverage. The State of Mississippi was forced to update its AI as part of the reconstruction grant following Hurricane Katrina. Yet there are no substantially equivalent FHAP grantees in either Mississippi or Alabama. In Atlanta, the AFFI certification has encouraged local jurisdictions to adopt substantially equivalent fair housing ordinances, which has led to a total of 28 FHAP grantees in 22 jurisdictions. In this case too, staffing and funds became a barrier to further progress. Insufficient data for monitoring. FHEO relies on PIH, Housing, and CPD program staff to identify potential fair housing issues. Program staff use risk-based checklists during frontend reviews of AFFI certifications, Community Planning and Development performance reports (CAPERs), etc. The Atlanta HA is a Moving to Work agency. As a result, it has been exempted from full reporting of tenant data, which prevents FHEO from effectively monitoring fair housing compliance. Staff judged that the data fields related to fair housing should be subject to mandatory reporting. Region 4 has had substantial difficulty getting TRACS tabulations from the Office of Housing in order to assess segregation in Section 8 developments. The data revealed possibly significant fair housing issues, as 90 percent of Florida's multifamily developments were found to be segregated in 2005. Interview with Director of Metro Fair Housing in Atlanta Metro Fair Housing Services is a FHIP agency. A member of the agency's staff authored Atlanta's AI. Fair housing complaints have decreased due to a lack of education and outreach funding. Major types of issues include foreclosures, tenant issues. Metro Fair Housing is using grant funds to build wheelchair ramps. Disability complaints have increased in recent years. Metro won a $300,000 settlement for a race-based case. Counties in the Atlanta region fail to take fair housing very seriously, despite the recent FHEO memo. While Atlanta and the counties of Dekalb, Fulton, and Clayton fund Metro Fair Housing, Cobb County does not. Cobb County instead produced its AI in-house, reflecting a generally conservative approach. After HUD's last Housing Discrimination Study, additional funding for retesting was provided in order to develop complaints. The testing revealed that racial steering was occurring (from Alfaretta to southern Fulton County). Atlanta still has no fair housing ordinance. However, at least one community development official from a surrounding county that a local ordinance might be more trouble than it is worth. Georgia law prohibits jurisdictions from adopting more restrictive ordinances. 25 of 27 DRAFT CHICAGO: Interview with Region 5 FHEO Director and staff The Chicago field staff agreed that AFFH should be more clearly defined and incorporate terms such as "integration." They also asserted that currently there is no accountability to AFFH. Staff also stated that a key component of fair housing is that everyone should have the option of where they want to live. Tenants who receive vouchers should have the option to choose where they want to live. Staff reported incidents in which tenants have received housing vouchers with little assistance in helping them to secure housing. Staff has advocated for more assistance to households who want to move to low poverty areas. These efforts may be enhanced by collaborating with key affordable housing organizations such as Housing Choice Partners. Resources for monitoring. While HUD Headquarters periodically requests the field staff to review AIs, field staff stressed that AI guidance is weak. In addition, the field office lacks the capacity and resources to devote time to reviewing AIs and ensuring goals are met successfully. Instead, FHEO field staff has allocated their efforts to other pressing issues where they have more authority. This includes managing Title VIII discrimination cases and Section 504 disability compliance and code enforcement. In addition, FHEO has worked with the Chicago Housing Authority (CHA) to develop a Voluntary Compliance Agreement for approximately1,500 CHA units to be accessible by the time CHA's "Plan for Transformation" is complete. No accountability. Field staff discussed that currently there is a lack of accountability for grantees to affirmatively further fair housing because only in rare cases that involve severe litigation are CDBG funds taken away. They discussed their frustrations to enforce AFFH efforts when they lack a "stick." Section 8 Source of Income/Housing Mobility. Staff also discussed housing mobility issues for voucher households. While the City of Chicago currently has an ordinance that covers Section 8 as a protected source of income for purposes of affording rents, the County and State laws fail to include Section 8 as a protected income source. Section 504 requirement for CHA's "Plan for Transformation." In constructing new affordable housing, CHA is moving away from townhouse designs toward mid-rise elevator buildings. Staff provided an example of how the Horner development had problems with accessibility after they rehabbed their units. Interview with Director of HOPE Fair Housing, Bernard Kleina HOPE Fair Housing is a nonprofit organization that works with 30 counties in Illinois. The nonprofit primarily focuses its efforts in DuPage County. In 1971 HOPE sued DuPage County for exclusionary zoning. As a result of the suit, HUD had cut CDBG funds from DuPage County for three years, but it was observed that this type of enforcement action is no longer being taken.. HOPE's relationship with DuPage County was strained for several years after this case, but the relationship has improved and HOPE now gets funding from DuPage County for testing. HOPE also assists in writing the AI for DuPage County. Although Naperville in DuPage County has its own CDBG program, they have not funded HOPE. Fair Housing Organization's Role in AI. Kleina asserted that local fair housing organizations do not necessarily need to write the AI. Rather, jurisdictions should receive input from an organization that HUD determines to be a "qualified fair housing center." 26 of 27 DRAFT No Accountability. While jurisdictions identify impediments in their AI, many of them do not state actions to address those impediments. Kleina suggests that perhaps specific resources (i.e. CDBG funds) should be tied to specific impediments to ensure that grantees work to resolve them. Kleina argues that at a minimum there should be more stringent guidelines. This would help legitimize fair housing organizations requests for funding from jurisdictions. Precedent-setting AFFH case. Kleina also discussed a pending lawsuit in Westchester County, NY, that he believes will establish a precedent in lawsuits related to AFFH. The community had signed off on AFFH documents but they weren't actually doing anything to AFFH. Section 8 Source of Income/Housing Mobility. While State law does not protect Section 8 as source of tenant income, local Chicago ordinance protects Section 8 as source of income. Several fair housing complaints in Chicago have been lodged by displaced public housing residents based on source of income. Kleina noted that leases denied on the basis of source of income may be a pretext for racial discrimination. Interview with Chicago Housing Authority (CHA) Background. Generally for HOPE VI sites, the mix of housing is one-third market rate units, one-third affordable units and one-third public housing, except for sites developed under a consent decree. For example, at the Cabrini site the mix is 50 percent market rate, 20 percent affordable and 30 percent public housing. Because of the consent decree, 50 percent of the public housing units have no work requirement. As a result of the Gautreaux case, CHA is now placing residents in units that look like market rate units, having the same finishes and appliances. Relocation. CHA resident services staff said that they do not have fair housing issues involving relocation. They have successfully placed residents throughout the city, in all but two districts out of approximately seventy. This statement conflicts with what Chicago HUD staff said. CHA staff, however, did mention that they were working with the Housing Choice Partners to help residents find housing. Site and Neighborhood Standards. CHA development staff said they do not have any major issues with site and neighborhood standards since CHA is primarily building on their own sites. It was noted that zoning laws in the Ward where Cabrini is located actually favors public housing, requiring 10 percent of units to be Public Housing. Accessibility--Section 504 Compliance. Based on a needs assessment, CHA developed guidelines that 5 percent of units at each site covered by the "Plan for Transformation" (of Public Housing), including both family and senior developments, need to comply with the Americans with Disabilities Act and Uniform Federal Accessibility Standards. Further, 20 percent of units at each site need to be "adaptable" if reasonable accommodation is needed. CHA has an additional two years to comply with accessibility requirements for senior developments because there are greater accessibility needs and they are mostly rehabilitated units. CHA works with the Mayor's Office for People with Disabilities to review initial building plans for compliance. The City sends inspectors on site when construction is 50 percent complete, 90 percent complete, and fully complete. 27 of 27