Developing a Standard Measurement of Housing Insecurity in Surveys - - PowerPoint PPT Presentation

Developing a Standard Measurement of Housing Insecurity in Surveys

Jessica Graber, Matt Virgile, and Dave Tuttle U.S. Census Bureau Nicole Watson and George Carter U.S. Department of Housing and Urban Development DC-AAPOR July 12, 2019 Washington, DC

1

Disclaimer: This presentation is released to inform interested parties of research and to encourage discussion. The views expressed are those

• f the authors and not necessarily those of the U.S. Census Bureau or the U.S. Department of Housing and Urban Development. The Census

Bureau’s Disclosure Review Board and Disclosure Avoidance Officers have reviewed this product for unauthorized disclosure of confidential information and have approved the disclosure avoidance practices applied to this release. CBDRB-FY19-CED002-B0001

Background

• The concept of Housing Insecurity (HI) has not been measured in a

consistent way across surveys in the US

• This creates a challenge for tracking prevalence of HI over time and examining

its relationship with other variables

• There is no transferable instrument for measuring HI that is currently used by

national surveys

• US Department of Housing and Urban Development (HUD) reviewed

research on HI, consulted experts, developed a working definition of HI tapping key concepts cited in the research, and drafted questions for a HI research module

2

Housing Insecurity Definition

• HI as “a significant lapse for a given household in one or more elements of

secure housing”

• 1. Affordability
• Shelter costs are manageable over the long term without severely burdening or

compromising other consumption that is normally essential for health and well-being

• 2. Stable Occupancy
• Household does not face substantial risk of involuntary displacement for economic or

non-economic reasons

• 3. Decent and Safe
• Unit has physical attributes that satisfy functional needs for well-being related to

health, security, and support for the activities of daily living

3

Research

• HUD will conduct a new HI follow-up survey with an opt-in sample of

eligible households who completed the 2019 American Housing Survey (AHS)

• Usually the AHS is conducted in odd-numbered years between May–September with

a sample of about 116,000 US households, by personal visit or telephone

• Households falling below 300% of the official poverty threshold (based on income

and number of people) will be invited to participate in HI follow-up telephone survey

• Upon completion, respondents will receive a $40 incentive
• A target of 4,000 responses will be collected in follow-up survey
• Census Bureau conducted cognitive pretesting on HI questions in 2018 and

made final recommendations to HUD

• This research module is the first step in HUD long-term goal to:
• Evaluate feasibility of collecting HI data and better define HI
• Measure quality of HI questions and develop a composite HI scale

4

Cognitive Interviewing - Data Collection

• Eligible respondents were at least 18 years old and fell below 300% of the

household poverty threshold

• 15 cognitive interviews
• 10 respondents from existing frame for a prior Census Bureau study
• Email invitation and phone calls made to those who were not selected for prior study
• Satisfied one of three criteria: (i) Currently rent a home, or own with a mortgage or loan; (ii)

Live with any children under 18 years old; or (iii) Moved within the past two years

• 5 respondents from new recruitment
• Recruitment ads posted on Craigslist and Nextdoor
• Respondents lived with at least one adult nonrelative
• All respondents pre-screened by telephone for eligibility, interviewed face-

to-face, and given $40 cash incentive

• Concurrent probes asked how respondents arrived at an answer, whether question

was easy or difficult to answer, and what specific term or phrase in a question meant to them

• Final debriefing section asked about overall experience, including difficulty and

sensitivity

5

Cognitive Interviewing - Respondents (n=15)

Demographics Total Sex Male 1 Female 14 Age Under 35 4 35-49 5 50 and above 6 Education Less than Bachelor’s Degree 7 Bachelor’s Degree or Higher 8 Race White 3 Black or African-American 9 Asian 1 Multi-Racial 2 Hispanic Origin Hispanic 2 Non-Hispanic 13

6

Cognitive Interviewing - Protocol

Section

• No. Questions

Content

• I. Screener*

15

• Home type and tenure
• Mortgages (if applicable)
• Household roster and demographics
• Length of time living at current home
• Tenure of previous home (if applicable)
• II. Affordability

25

• Difficulty affording rent or mortgage payments
• Level and frequency of stress over payments
• History falling behind on payments
• Difficulty paying for other bills and expenses
• Assistance offered for housing costs

7

*Screener section included for context, not for testing.

Cognitive Interviewing - Protocol (cont’d)

Section

• No. Questions

Content

• III. Stable Occupancy

35

• Level and frequency of stress over being forced to move
• History of eviction or foreclosure
• Reasons for moving from previous home (if applicable)
• Problem finding affordable housing in good condition/neighborhood
• Anyone temporarily staying in home due to financial difficulty
• Anyone experienced homelessness in last 12 months
• IV. Decent and Safe

10

• Frequency of musty smells in home
• Home repairs by owner started and solved quickly (if applicable)
• More people in home than can live comfortably
• Any major problems in home
• Home safety and security

8

Results - Affordability

• A few respondents were visibly upset by one or more questions during this

section, or indicated during the debriefing that this section was particularly sensitive

• One respondent had unstable employment and had difficulty affording many

expenses; the interviewer limited the number of probes to avoid upsetting her further

• Another respondent wept when asked a question about difficulty paying for medical

bills or prescriptions, saying her husband was mentally ill and her insurance would not cover his needs

• A third respondent said during debriefing “For me, any admission of sources of

income, where support comes from, that was hard to speak on. Any other questions that just kind of reveal poverty is kind of difficult.”

• Based on this, Census recommended moving the Affordability section to

come later in the survey, to ease respondents into sensitive questions and to collect more data before potential break-offs

• This section will come last for the 2019 survey production, after sections on Stable

Occupancy and Decent & Safe

9

Results – Affordability (cont’d)

• Some respondents living with nonrelatives had difficulty answering questions that

were about their entire household and could only answer for themselves

• Lacked financial knowledge about their housemates or roommates
• Indicated they would answer “Don’t Know” if forced to answer on behalf of entire household
• “Oh, our leases are entirely separate. I have my own lease … Our rent is not taken in total. We all

have individual leases.”

• “I don’t know about my roommate … I can’t speak on his finance, we’re not in a relationship or

anything … I can only say that as far as me.”

• Census reported this difficulty for HUD to determine a strategy for nonrelative

households

• For 2019 survey production the first question of this type will include a prompt for these

respondents: “Please consider the total amount paid by the household.”

• A new question at the end of the section will ask these respondents: “How difficult was it for

you to answer these questions about housing costs for your household?”

• Very difficult
• Moderately difficult
• A little difficult
• Not at all difficult

10

Results – Stable Occupancy

• Some respondents who had recently moved from their previous home

indicated that questions about hardship-related reasons for moving were burdensome

• These respondents did not have any hardship-related reason and answered “No” to

all questions; they moved simply because they wanted to be somewhere new or sold their old home

• “Instead of asking all these questions, why don’t you just ask ‘Why did you move away?’

first?”

• For 2019 survey production the preceding text for these questions “Did you move

away because” will be optional to read after the first few questions, to reduce repetition and burden

• Some respondents were unsure who to include in questions about

“temporary housing residents”

• For 2019 survey production text will be added to “Include yourself” and defining

such a resident as one that “has nowhere else to go and intends to stay here for a short time”

11

Results – Decent & Safe

• Some respondents had difficulty answering a question about frequency of “musty

smells” in their home, defined as “smells of dampness, mold, or mildew”

• Based their answers partially or entirely on visual signs of these, rather than strictly based on

smells

• “Recently there was a leak from the AC unit which flooded a portion of the room … I noticed mold and a

soaked box. Found the wet spot before we noticed any smell.”

• “There were incidents when I could see water damage collecting, especially being in a basement. When

I moved there, there was a moldy chute. But smelling, that only happened once.”

• For 2019 survey production the question will be revised to ask about musty smells “or other

signs of dampness, mold, or mildew” in the home

• Some respondents had difficulty answering a question about whether their

landlord starts major repairs “quickly enough” since they rarely or never needed repairs

• For 2019 survey production the question will be preceded with a “screener” yes/no question

about whether they’ve ever needed repairs; those answering “yes” will be asked the question and others will skip the question

12

Conclusions

• Questions about Affordability were particularly sensitive for a few

respondents

• This section will be moved to the end of the survey to ease respondents into

sensitive questions and to collect more data before potential break-offs

• Respondents living with nonrelatives had difficulty answering Affordability

questions for their household due to lack of knowledge

• The first question in this series will include a prompt for these respondents to answer

based on total amount paid by the household, and a new question will be added to the end of the section about difficulty answering these

• Questions about Stable Occupancy and Decent & Safe had fewer issues

with difficulty, and neither section was indicated to be sensitive

• Minor revisions in question wording will be implemented to reduce burden or clarify

meaning

13

Scale Development Plan

• HI Research Module (HIRM) Scale Development Process
• 2019: AHS HIRM data collection
• 2019-2021: HIRM Scale Development Contract
• Scale design experts will analyze 2019 HIRM data using psychometric methods to suggest

a smaller set of questions for a HI scale, and begin development of that scale

• HIRM Scale Development final report
• 2023-2024: Reduced set of HIRM questions asked of 2023 AHS full sample
• Recalibrate HI scale proposal with data from the 2023 AHS
• Report results of recalibrated HI scale proposal
• 2025-2026: Pursue National Academy of Sciences study on HI scale
• 2027: Informed by further study of HIRM questions, add revised questions to

the 2027 AHS core

14

Developing a Standard Measurement of Housing Insecurity in Surveys

Matthew.Virgile@census.gov George.R.Carter@hud.gov

15