Worst Case Housing Needs of Households with Disabilities

I discussed the Worst Case Housing Needs 2009: Report to Congress previously on a blog, and you may want to review my summary first. I mentioned that the Department of Housing and Urban Development was not very confident of their disability data, and that they would be following up with another report. That day has arrived, and the 2009 Worst Case Housing Needs of People with Disabilities has arrived.

This report isn’t as much a further examination of the housing needs of people with disabilities as it is a discussion of the surveying methods and an explanation of present and past differences in data.

 
Apparently, in 2009, the American Housing Survey (AHS), for the first time, included direct questions for data on households with persons with disabilities. Prior to that, for the Worst Case Housing Needs reports, HUD has used proxy measures using income sources typically associated with persons with disabilities such as Social Security and other benefits. The report compares the number of non-elderly households with persons with disabilities compared to this proxy report, and also to American Community Survey (ACS) and National Health Interview Survey (NHIS). 

QUESTIONS ABOUT DISABILITIES IN THE 2009 AMERICAN HOUSING SURVEY

•  Are you deaf or have serious difficulty hearing? 
• Are you blind or have serious difficulty seeing, even when wearing glasses?
•  Because of a physical, mental, or emotional condition, does anyone in this household have serious difficulty concentrating, remembering, or making decisions?
• Does anyone in this household have serious difficulty walking or climbing stairs?
• Does anyone in this household have serious difficulty dressing or bathing?
• [For all household members 15 years old or older] Because of a physical, mental, or emotional condition, does anyone in this household have difficulty doing errands alone such as visiting a doctor’s office or shopping?

 Source: HUD PD&R (2010)

.... These functional limitations are complemented by two questions that measure difficulties with Activities of Daily Living (ADL), such bathing and dressing (self-care difficulty), and Instrumental Activities of Daily Living (IADL), such as performing errands (independent living difficulty).

The 2009 American Housing Survey found that 11% of non-elderly rental households and 7% of non-elderly owner-occupied households reported affirmative to at least one of the six questions. I’m not sure why they did not include the statistics for elderly persons with disabilities for much of the report since the likelihood of having a disability rises with age. However, at one point, the report states that there are 987,000 households with worst case needs and disabilities, and there are also 619,000 households with worst case needs that include an elderly person with disabilities, with some overlap. So there actually are approximately 1,606,000 worst-case needs households with a person with a disability, elderly or not. 

This is one of the differences between using the AHS data in the latest survey with the past estimates using proxy data. As discussed above, the data for non-elderly worst case needs households with disabilities is 987,000, but proxy estimates places it at slightly above 1.1 million households.

When you look at renter households with non-elderly persons with disabilities, they are: 

more likely to have very low incomes, to experience worst case needs, and to pay more than one-half of the household income for rent, than those renter households without disabilities. In 2009, two out of three (or 66 percent) of renter households with disabilities had very low incomes, and only 46 percent of renter households without disabilities had very low incomes. Approximately 25 percent of renter households with disabilities experienced worst case needs, and only 19 percent of renter households without disabilities experienced worst case needs. Finally, renter households with disabilities were almost one and one-half times more likely to pay more than one-half of their income for rent than renter households without disabilities.

But renter households with disabilities were more likely to receive housing assistance, although there is still a huge number of low-income households (with or without disabilities) without assistance.

Of the households with disabilities considered having worst-case needs, people with ambulatory (mobility) disabilities (54%) and people with cognitive disabilities (48%) were the highest percentage.

The number of renter low-income households with disabilities considered having worst-case needs has increased from 1,059,000 in 2005 to 1,140,000 in 2009 (using proxy data). However, the percentage compared to the overall total of worst-case housing needs has actually decreased from 17% to 16% due to a large jump in the number of households with worst-case needs (most likely due to the economy and a higher number of households without a person with a disability paying more of their income toward rent).

Across the nation, worst case housing needs for households with disabilities tend to be greater in central cities with 41% of total, although the suburbs aren’t that far behind at 38% of total, while the non-metro areas had 21%. 

I believe the report also highlights the importance of understanding the survey design and the resulting data.

A number of possible reasons may explain differences in disability rates between ACS and AHS. These surveys have different purposes and context, with the AHS focusing on housing units and the ACS focusing on people. Results might also be affected by the different modes of collection. AHS uses only personal and telephone interviews and ACS relies more on a mailed form. Finally, AHS and ACS questions are grouped differently, have slight wording differences, and some of the questions are applied to different age groups. Although three questions in ACS are applied only to people 5 years old or older, they are applied to people of all ages in AHS. Even when surveys use similar questions, slight variations in wording or question sequence can yield different results.