R-9: Fair Housing Compliance Assessment and Advocacy


Methods

While research on environmental factors is increasing rapidly, there is not much empirical research on housing accessibility or compliance of existing housing with federal housing laws. 

Even more, while advocacy interventions have shown promise for addressing a variety of systemic issues systematic comparative evaluation of advocacy to improve housing accessibility has not been reported. This project addresses these gaps in four phases, including (1) development and validation of two new rapid housing accessibility and usability assessment tools, (2) application of these new tools to assess accessibility and usability on a large population sample, (3) implementation of a traditional social action and a “New Governance” model of social change, and (4) evaluation of the effects of the two models of advocacy in promoting improvements in accessibility and usability. 

Sample

The figure below presents the study’s Consort diagram estimating sample size for the target population of people with disabilities living in housing units covered by federal housing laws.  The diagram includes estimates for recruitment of residents with disabilities, usability survey return rate and assignment of eligible participants into one of two intervention groups. 

The study sample will include 198 adults with disabilities aged 18 to 70 who live in housing units constructed since 1991 (i.e., units covered by federal housing laws) that are in one of three U.S. communities selected through a competitive application process (see below).  

To create the study sample of individuals with disabilities living in targeted housing, we will identify the existing housing complexes in a given geographic location using publicly available building permit records.  Next, we will randomly select housing complexes to reach a target of 2,000 housing units (e.g., apartments) in each community. The rate of rented housing units occupied by someone with a disability is 17.6%. 

Hence, we project the random housing unit sample of 6,000 units will include 1,056 units occupied by at least one person with a disability. Assuming a 25% response rate to a post-card solicitation and a 60% response to the survey, we arrive at our target survey sample of 198 individuals. Then, assuming 30% of these individuals are eligible (i.e., report usability problems with their housing and are willing to participate in the intervention), we will have 60 individuals in the experiment.  

To recruit study participants with disabilities, we will apply Dillman’s Tailored Design methods.  These evidence-based recruitment methods specify details for writing cover letters, soliciting timely survey returns, and conducting timed follow-up.  In keeping with these procedures, we will send a letter to each randomly selected housing unit, soliciting participation of anyone in the household aged 18-70 who could answer “yes” to at least one of five disability screening items from the American Community Survey (ACS). 

Inclusion criteria will focus on individuals with impairments that call for architectural accommodations. Residents with cognitive or emotional impairments but who do not have mobility impairments will be excluded as establishing consensus on environmental modifications to accommodate those issues is beyond the scope of this study.  The questions we will ask are:

  • Are you deaf or do you have serious difficulty hearing?
  • Are you blind or do you have serious difficulty seeing even when wearing glasses?
  • Do you have serious difficulty walking or climbing stairs?
  • Do you have difficulty dressing or bathing?
  • Because of a physical, mental, or emotional condition, do you have difficulty doing errands alone such as visiting a doctor’s office or shopping?

Along with each letter, we will include a pre-coded, anonymous, postage paid-post card that individuals can simply put in the outgoing mail to receive the housing usability survey. 

In another NIDRR-funded study that used this recruitment method, nearly 30% of the eligible disability population across three communities returned a survey, making this a very efficient method for drawing population-based samples. However, while our response rate estimates are based on our previous research, there is very little population-based housing research to further guide these estimates. If we fail to meet our target sample with these methods, we will use Respondent-Driven Sampling, a derivative of snowball sampling, to identify more intervention participants who are representative of the target population.

Design

We propose to use a true randomized control group experimental design with participants living in non-compliant housing randomly assigned to treatment by randomly assigning participants living in non-compliant housing to intervention (New Governance) or control (Complaint) groups.

First, we will assess the usability of housing covered by federal housing laws (i.e., multifamily complexes with four or more units constructed since 1991). Second, we will assess the accessibility of housing complexes and units identified as problematic in the usability survey.  Third, we will randomly assign individuals who live in housing units/complexes with verified accessibility compliance violations to either intervention or control experimental conditions and implement the interventions. Finally, we will repeat collection of the data six-months after intervention implementation. 

To implement the project, we will solicit participation of centers for independent living (CILs) through a competitive application process to receive contracts totaling $8,000 for each center distributed over four years. In another study, over 30% of contacted CILs submitted applications to participate in our research. We will select three centers in diverse geographic regions that demonstrate the capacity to identify housing units in their communities and to assist with the development and implementation of the measurement and intervention procedures of the project.  

Specifically, CILs will (1) recruit individuals to serve on our project advisory board, and identify the existing housing units for random selection by working with local building permit departments, (2) pilot test the usability survey and rapid accessibility assessment procedures and provide feedback in two-three iterations, (3) collect rapid accessibility assessment data, (4) participate in teleconference calls to shape the experimental and control intervention procedures, and (5) assign two staff to implement the interventions and reduce contamination (one for the experimental and one for the control conditions).   

Measures

We will develop two brief screening measurement instruments for this study.  These instruments will serve as pre- and post-test outcome measures for the proposed intervention.       

Rapid Usability Assessment. The usability assessment will screen residents for problems in housing complex common areas and individual unit kitchens and bathrooms.  These areas are specifically covered by federal housing laws.  Content of the survey will include usability ratings by resident respondents.  When respondents indicate low or no use, follow up questions will help identify whether or not a home visit and accessibility survey is needed.  



Rapid Housing Accessibility Assessment (RHAA).  We will base our brief accessibility instrument on methods used by the American Society of Home Inspectors that provide computerized reporting tools based on checklists completed during home inspections.  This method uses digital photography to highlight problems and assures that all member inspectors use the same criteria for assessing the adequacy of inspected structures. 

Our brief assessments of accessibility will use a similar method.  Checklists of access criteria will be generated based on FHAA and Section 504 of the Rehabilitation Act regulations.  Where potential problems exist, photos will be used to document the problem.  Checklists and photos of problem areas will be put together in a report that documents accessibility and compliance issues.  The RHAA will be collected by CIL staff through home visits, direct observation and completion of home access checklists and digital photography. 

In addition to these outcome measures, we will collect data on the advocacy process. These internet-based measures (e.g., Survey Monkey) will be completed by CIL staff and will include the number of actions taken, the time taken to complete the actions, and judgments of the social validity of the actions. 

We will examine the reliability, stability and validity of the RUA and RHAA measures as they are developed. Test re-test will be conducted over a two week period for self-report measures. The validity of the Rapid Usability Assessment will be assessed with data collected during the Rapid Housing Accessibility Assessment. Interrater reliability for the RHAA will be raised to a criterion of .80 or better. We have allocated 15 months for creating these measures.

Intervention

We will compare the relative effectiveness of two advocacy strategies.  The first, the self-advocacy toolkit involves a strategy based on a legal or “complaint” framework. This “control condition” will follow typical protocol for lodging complaints with building owners, property managers, and regulatory agencies (e.g., the housing authority).

The second advocacy strategy will be based on the NG framework. New Governance involves organizing networks of public agencies, private businesses, and citizen representatives to cooperate on solving a shared problem.  New Governance combines several advocacy strategies – monitoring with feedback, public education, watch-dogging, coalition building, etc. – into one comprehensive program that uses a collaborative, community development strategy rather than confrontational, social action tactics.  In this case, a CIL will organize a local consumer empowered team (CET). 



The CET will work with the researchers to collect access and usability data.  Together, the CIL and the CET team will develop a series of brief reports that will be distributed to potential stakeholders (e.g., Office of Public Housing, landlords, building and trades association). These reports will highlight the observational data on access, relate it to legal standards, explain the consequences of inaccessibility to individuals, and invite stakeholders to a series of meetings to address the problem.   

At the public meeting a facilitator will prompt discussions about gaps between observed access and usability, and federal housing regulations. The facilitator will insure that the perspectives of different stakeholders are noted and considered. Once the gaps have been identified and discussed, the facilitator will lead a brainstorming process to generate ideas on how to close the gaps, identify who has the authority and is responsible for taking the actions, and form participant groups who commit to follow up on proposed actions. 

The results of the discussion will be organized into a formal plan. The plan will be distributed to the participants, to city and county officials, and to local media.  The CET will use electronic media (website and email) to prompt planned actions by task groups, monitor their progress, and evaluate their impact. 

This second intervention will be developed by research staff with guidance from an advisory CET.  It will use techniques to educate and enlist collaboration among a variety of stakeholders including residents, building owners, property managers, housing regulatory officials and others (e.g., case managers). Stakeholders will maintain contact with each other using a combination of face-to-face meetings, email and phone contact, and a personal networking site.  All communication will be copied to investigators who will use it to assess fidelity of the intervention. In the event residents do not have email access, the CIL staff will review all outgoing and incoming communication over the phone to solicit input and participation in the advocacy process and provide hard copies of all communication to the research participants.

Based on prior work, we anticipate specific intervention components will include: 

  • Brief report backed by technical data with recommendations that include a grading system that assigns letter grades to housing complexes for ease of understanding; distributed to stakeholders including building owners, city and county officials, housing authority and local construction association. 
  • A stakeholder community meeting to present results of the usability and accessibility survey and to develop action plans that participant groups will use for implementing recommendations.
  • Regular email communication reporting on accessibility/usability problems and solutions as reported by residents and building owners.  When necessary, face-to-face meetings with stakeholders may be held to generate additional action steps and to problem solve issues.
  • Follow-up report about housing unit access to all stakeholders based on usability and accessibility post-test results that details action steps taken and results achieved.
  • Overall community report to all stakeholders at the conclusion of the intervention, detailing overall housing grades, actions taken and results achieved.

Both experimental and control conditions will be implemented by two CIL staff  (one assigned to the experimental and one to the control condition) who will meet research participants in person at least once and then rely on electronic communication except when residents do not have access to email.  

Data Analysis

All data will be entered into SPSS 19.0 (152) and checked for data entry errors.  Descriptive characteristics of all variables will be computed and the data distributions checked for normality deviations.  We will create two indices of compliance. First, we calculate a percentage score for each access and usability dimension for each housing complex and each community.  In addition, we will convert the percentage scores to a Grade scale: A-F. 

These results will be incorporated into the brief reports disseminated publicly as part of the NG process – with the grade scale providing an easy-to-understand ranking.   Second, we will compute an overall index that reflects the overall number of violations detected for each housing complex and report this by overall, as well as internal and external conditions.  Additionally, we will use the “weakest link” method to reflect compliance with all standards in Fair Housing and Section 504 of the Rehabilitation Act.  

We will assess the relative effectiveness of the intervention and control (i.e., NG and Complaint) by analyzing process and outcome data.  For outcome analysis, we will use a mixed model repeated measures general linear model to examine change over time both within and between groups.  A priori contrasts for the group X time interaction will indicate the relative effectiveness of each group.   For process outcomes, we will examine the amount of change produced relative to the level of effort and receptivity of the intervention by adding these variables as covariates to the model.  For example, a negative, significant effort covariate associated with a significant intervention X time interaction would indicate that significant outcomes are achieved with less effort.