Worst Case Housing Needs: 2021 Report to Congress

WORST CASE

HOUSING NEEDS

2021 REPORT TO CONGRESS

U.S. Department of Housing and Urban Development | Ofce of Policy Development and Research

Visit PD&R’s website

huduser.gov

to nd this report and others sponsored by HUD’s Ofce of Policy Development and Research (PD&R).

Other services of HUD User, PD&R’s research information service, include listservs, special interest

reports, bimonthly publications (best practices, signicant studies from other sources), access to public

use databases, and a hotline (800-245-2691) for help accessing the information you need.

WORST CASE

HOUSING NEEDS

2021 REPORT TO CONGRESS

Prepared for

U.S. Department of Housing and Urban Development

Ofce of Policy Development and Research

Prepared by

Thyria Alvarez

Barry L. Steffen

July 2021

WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Acknowledgments

The authors of this report gratefully acknowledge David A. Vandenbroucke

for his invaluable contributions to the Worst Case Housing Needs Report to

Congress over the years. David Vandenbroucke was a senior economist for the

U.S. Department of Housing and Urban Development who served for 28 years

before retiring in 2020. Mr. Vandenbroucke managed the American Housing

Survey and led its two survey redesigns in 1997 and 2015.

The authors would also like to thank Jennifer Turnham for her leadership and

expert advice in this report.

WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Foreword

The U.S. Department of Housing and Urban Department (HUD) is pleased to

present to the U.S. Congress its 18th biennial report on Worst Case Housing

Needs. The 2021 report on Worst Case Housing Needs provides national data

and analyzes the critical problems facing low-income renting families. The report

primarily draws on data from the American Housing Survey (AHS) sponsored

by HUD and conducted by the U.S. Census Bureau. AHS is a comprehensive

national longitudinal housing survey conducted since 1973.

Households with worst case housing needs are very low-income renters—

households with incomes at or below 50 percent of area median income—who

do not receive government housing assistance and who pay more than one-

half of their income toward rent, live in severely inadequate conditions, or both.

The report finds that in 2019, before the COVID-19 pandemic, 7.77 million

households had worst case housing needs. This is an improvement from the

record high of 8.5 million in 2011 but is substantially higher than the 5 million

households with worst case housing needs in 2001. There was no significant

change in the number of households with worst case housing needs between

2017 and 2019 despite favorable economic conditions.

While the overall number of worst case needs was unchanged, there were some

changes in the underlying drivers of worst case needs that push the count in

different directions. Two changes that lessened worst case housing needs were

modest income growth among households at the top of the very low-income

range (those with incomes between 30 and 50 percent of area median income)

and a modest increase in the availability of quality, affordable housing stock

for very low-income renters. Countering these positive developments were an

increase in total very low-income renters due to household formation and a

modest decrease in the number of very low-income renters receiving housing

assistance.

The 7.77 million households with worst case housing needs in 2019 included

2.27 million households with children, 2.24 million households headed by an

older adult (62 years or older), and 2.54 million single adults. About 13 percent

of households with worst case needs included people younger than 62 who

have disabilities, and about one-half were non-White or of Hispanic ethnicity.

Among very low-income renters, more than one-half of Asian, Native Hawaiian,

and Other Pacific Islander households had worst case needs, as did more

than 45 percent of Hispanic households, 44 percent of non-Hispanic White

households, and 36 percent of non-Hispanic Black households.

The key to ending worst case housing needs is increasing affordable housing.

In 2019, only 62 affordable units were available for every 100 very low-income

renter households. Only 40 affordable units were available for every 100

extremely low-income renter households. HUD is committed to ending worst

case housing needs and homelessness in America by increasing affordable

housing access. Key policy levers include increasing incomes of very low-

FOREWORD

WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

income renters, substantially expanding rental assistance, preserving the

existing assisted and affordable housing stock, and reducing barriers to the

production of new affordable housing.

This report captures housing need in the period immediately before the onset

of the COVID-19 pandemic and associated economic recession early in 2020.

The financial shock to the labor market and household incomes may cause

substantial increases in worst case needs when next measured with the 2021

American Housing Survey. The potential increase in worst case needs could

be dampened by strong fiscal relief packages that provide rental assistance to

sustain affordable housing and prevent homelessness during the pandemic.

Increased federal resources provided by pandemic stimulus packages,

fiscal year 2021 appropriations, and 2022 budget proposals are estimated

to generate affordable and assisted housing opportunities for approximately

330,000 households who would otherwise be at risk of worst case housing

needs. Further, targeted tax credits and resources proposed in the Biden

administration’s initial infrastructure plan would build and modernize more than

two million affordable and sustainable places to live. A Special Addendum in this

report discusses the recession, key features of the relief legislation, expected

impacts of the recession on housing outcomes, and potential implications for

future worst case needs estimates. The focus on the pandemic and response,

however, should in no way distract from the persistent, underlying structural gap

in the affordable housing market that is consistently documented in Worst Case

Needs reports. Ideally, a policy response that begins to bridge this affordable

housing gap will also seek to address geographic disparities in resource

allocation that contribute to inequities and pockets of distress.

Todd Richardson

General Deputy Assistant Secretary for Policy Development and Research

U.S. Department of Housing and Urban Development

FOREWORD

vWORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Contents

Executive Summary ............................................................................................ vii

Few Significant Changes in Worst Case Needs Since 2017

....................... vii

Worst Case Needs Improved Slightly for Some Demographic Groups

and Household Types

.....................................................................................viii

Persistent Shortage of Affordable and Available Rental Housing

Is Especially Severe for Extremely Low-Income Households

....................ix

Conclusion

........................................................................................................xi

Section 1. Extent and Nature of Worst Case Needs

........................................ 1

Extent of Worst Case Needs in 2019

.............................................................. 1

Which Households Can Have Worst Case Needs?

........................................ 2

Severe Problems Trigger Worst Case Needs

................................................. 2

Inadequate Housing and Inadequate Income

............................................... 2

Progress in Reducing Homelessness

............................................................. 3

Prevalence of Worst Case Needs by Income

................................................ 4

Worst Case Needs Prevalence Among U.S. Households

.............................. 4

Demographics of Worst Case Needs

............................................................. 5

Worst Case Needs by Race and Ethnicity

...................................................... 5

Worst Case Needs by Household Type

.......................................................... 7

Disability and Accessibility in the American Housing Survey

....................... 10

Summary

.......................................................................................................... 11

Section 2. Shortage of Affordable Housing

..................................................... 13

Geography of Worst Case Needs

.................................................................. 13

Worst Case Needs and Housing Assistance by Region and

Metropolitan Location

.................................................................................... 14

Variation in Worst Case Needs Between Metropolitan Markets..................... 16

Factors Limiting Access to Affordable Rental Housing

............................. 18

How the Market Allocates Affordable Housing on a National Basis

.............. 18

Affordability, Availability, and Adequacy of the National Rental Stock

.......... 19

Measuring Whether Affordable Housing Stock Is Sufficient for Need

........... 20

CONTENTS

vi WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Rental Stock by Income ................................................................................. 21

Geography of Supply ...................................................................................... 22

Rental Stock by Region ................................................................................. 22

Rental Stock by Metropolitan Location .......................................................... 23

Summary ......................................................................................................... 24

Section 3. Understanding the Trend in Worst Case Needs ........................... 27

Drivers of Affordable Housing Demand .......................................................28

Other Factors Affecting Affordable Housing Supply and Demand ........... 29

Changes in Income Limits and Worst Case Needs ....................................... 31

Concluding Summary ..................................................................................... 32

Special Addendum. The Potential Effect of the COVID Pandemic

on Worst Case Needs ......................................................................................... 33

Economic Implications of the Pandemic ...................................................... 33

Federal Pandemic Relief Legislation .............................................................. 34

Eviction Moratoria ...........................................................................................35

What Percentage of Renter Households Are Behind on Rent

During the Pandemic? ....................................................................................35

How Well Will the AHS Measure Pandemic Effects in 2021? ..................... 37

Summary ......................................................................................................... 38

Appendix A. Detailed Data on Housing Problems and Supply

of Affordable Housing ........................................................................................ 39

Appendix B. Supplemental Exhibits ................................................................. 79

Appendix C. Federal Housing Assistance and Affordable

Housing Programs .............................................................................................. 81

Appendix D. Previous Reports to Congress on Worst Case Needs ............. 83

Appendix E. Data and Methodology ................................................................. 85

Using the American Housing Survey Data .................................................... 85

Household and Family Types ......................................................................... 86

Housing Assistance Status ............................................................................ 87

Housing Problems .......................................................................................... 87

Income Measurement .................................................................................... 88

Income Categories ......................................................................................... 88

Location ......................................................................................................... 89

Mismatch of Supply and Demand for Affordable Rental Housing ................. 89

Race and Ethnicity ......................................................................................... 90

Literature Cited ................................................................................................... 91

CONTENTS

viiWORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Worst case needs are a long-standing measure of the extent of unmet needs for

affordable rental housing of adequate quality. Renter households are defined

as having worst case needs for such housing if they have very low incomes—

household incomes at or below 50 percent of the area median income (AMI), do

not receive government housing assistance, and pay more than one-half of their

income for rent, live in severely inadequate conditions, or both.

Worst Case Housing Needs: 2021 Report to Congress examines the causes

of and trends in worst case needs using the most recent data from the 2019

American Housing Survey. The report finds that despite favorable economic

conditions in the 2017–2019 period, worst case housing needs persisted across

demographic groups, household types, and regions throughout the United

States. The unmet need for decent, safe, and affordable rental housing has

continued to outpace income growth and the ability of federal, state, and local

governments to supply housing assistance and facilitate affordable housing

production. As a result, the number of families with worst case housing needs

in 2019 remains modestly below historical high levels recorded since the Great

Recession of 2007–2009.

The report captures housing need in the period immediately before the onset

of the COVID-19 pandemic and associated economic recession early in 2020.

The financial shock to the labor market and household incomes can cause

substantial increases in worst case needs when next measured with the 2021

American Housing Survey. The major federal legislative response, however,

complicates the expectation and measurement of worst case needs. A Special

Addendum in this report discusses the recession, key features of the relief

legislation, expected impacts of the recession on housing outcomes, and

potential implications for future worst case needs estimates.

Few Signicant Changes in Worst Case

Needs Since 2017

There were 7.77 million renter households with worst case needs in 2019, a

statistically insignificant increase of 50,000 cases compared with 7.72 million in

2017 (exhibit ES-1).

The latest figure continues to represent an improvement from the overall record

high of 8.5 million in 2011 but remains higher than during the years preceding

the 2007–2009 recession when there was greater availability of affordable

housing stock.

EXECUTIVE SUMMARY

Executive

Summary

viii WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

0.0

6.0

8.0

9.0

5.0

7.0

4.0

2.0

3.0

1.0

2009 2011 2013 2015 2017 2019

Renters with worst case needs (millions)

7.10

8.48

7.72

8.30

Exhibit ES-1. Change in Worst Case Housing Needs, 2009–2019

Source: HUD PD&R tabulations of American Housing Survey data

7.72

7.77

10.0

The rate at which very low-income (VLI) renters experience

worst case needs also has improved only modestly in recent

years. The percentage of VLI renters experiencing worst

case needs (the “prevalence”) was 42.2 percent in 2019, a

slight reduction of 0.5 points from 42.7 percent in 2017. The

number of worst case needs increased more slowly (0.6

percent) than the number of VLI renters (1.8 percent). The

prevalence has improved moderately from the highest rate

observed since the 2007–2009 recession, 44 percent in

2011. The most recent biennial change is attributable to three

factors: (1) modest income growth among households at the

top of the VLI range (those with incomes between 30 and

50 percent of AMI); (2) a modest decrease in the number of

renters with very low incomes receiving housing assistance;

and (3) a modest increase in the availability of quality,

affordable housing stock for very low-income renters.

Although the relative shares of renters with incomes at

and below 30 percent of AMI (known as extremely low-

income, or ELI

) and with incomes between 30 and 50

percent of AMI did not change, the prevalence of worst case

needs increased among the lowest-income group while it

decreased among the next income group. As a result, ELI

renters account for the majority of worst case needs cases:

74 percent in 2019, a proportion that has not been seen since

2005.

The Department of Housing and Urban Development (HUD) is required by law to set income limits or cutoffs that determine the eligibility of applicants for HUD’s

assisted housing programs. Extremely low-income (ELI) cutoffs for Section 8 programs historically meant household incomes at or below 30 percent of the

HUD-adjusted median family income for the metropolitan area. The FY 2014 Continuing Appropriations Act required ELI cutoffs to be set at the greater of 30

percent of HUD-adjusted median family income or the federal poverty guideline as published by the Department of Health and Human Services. The ELI cutoff

is capped by the VLI cutoff. See the Income Categories discussion in appendix E for further information.

Worst Case Needs

Improved Slightly for Some

Demographic Groups and

Household Types

The percentage of very low-income renters experiencing worst

case needs varied among demographic groups. In 2019, the

prevalence of worst case needs was 55 percent among Native

Hawaiian or Other Pacific Islander households, 53 percent

among Asian households, 45 percent among Hispanic

households, 44 percent among non-Hispanic White

households, 36 percent among non-Hispanic Black households

and the other race and ethnicity group, and 23 percent among

American Indian or Alaska Native households. The prevalence

of VLI renters with severe problems narrowly decreased by

2 percentage points for non-Hispanic Whites and 1 point for

Hispanics but increased by 1.6 points for non-Hispanic

Blacks and by 2 points for other races or other ethnicities.

The percentage of VLI renters receiving rental assistance

decreased for all racial and ethnic groups between 2017

and 2019.

Three regions in the country—Midwest, Northeast, and

South—had an average decline of about 2 percent in the

share of renter households reporting worst case needs in 2019,

while the West saw an increase of almost 7 percent, offsetting

the decreases in other regions. The prevalence of worst case

needs decreased in suburbs and non-metro areas between

2017 and 2019 but not in central cities. The greatest decline

was observed in rural suburbs.

The prevalence of worst case needs slightly declined among

all household types, with the exception of households headed

by older adults. As the older adult population has increased

during the past 10 years, so, too, has the number of older adult

households with severe housing problems. The prevalence of

worst case needs decreased by 1 point among households

headed by someone younger than 62 while it increased by 1

percentage point among households headed by an older adult.

Despite minor changes, worst case needs remained a serious

and prevalent problem among all household types in 2019:

40 percent among families with children and among

households headed by older adults, 44 percent among

“other family” households (including multiple family members

without children), and 46 percent among “other nonfamily”

households (mostly single individuals). In absolute terms,

worst case needs involving other nonfamily households

increased during the last biennial period. In 2019, the worst

EXECUTIVE SUMMARY

ixWORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

case needs tally included 2.5 million “other nonfamily”

households, compared with 2.3 million families with children,

2.2 million older adult households, and 0.7 million “other

family” households.

About one in eight renter households with worst case

needs—13 percent—included people younger than 62 who

have disabilities. Those 1 million households reflect limited

improvement since 2011, when national levels of worst case

needs peaked at 8.5 million households.

Persistent Shortage of

Affordable and Available

Rental Housing Is Especially

Severe for Extremely Low-

Income Households

For most households, worst case needs are caused by

severe rent burdens—that is, paying more than one-half

of income for rent. Inadequate market supply, competition

for affordable units, and a shortage of rental assistance

continued to pose a substantial challenge for VLI renter

households in 2019. Inadequate housing quality caused only

3 percent of worst case needs.

The net increase in worst case needs by 50,000 cases

between 2017 and 2019 is attributable to a combination

of the demographic changes affecting the number

of unassisted VLI renter households and the housing

market’s response to such quantitative drivers of affordable

housing demand. An attribution analysis estimated the

independent contribution of each of four increasingly focused

demographic factors to assess its effect on the number of

unassisted VLI renters and thereby on the number of worst

case needs. The positive or negative effects attributed to

the four demographic factors are represented by the first

four bars of exhibit ES-2: household formation increased

worst case needs because there was a net increase in

new households from population changes; tenure shift

reduced worst case needs because the growth in renters

lagged growth in homeowners; renter household income

shifts increased worst case needs because there was a net

increase of those with income below 50 percent of AMI; and

the housing assistance gap increased worst case needs

because there was a net increase in VLI renters lacking

rent subsidies from the federal, state, or local government.

(The columns of ES-2 are cascading in the sense that each

column begins where the previous one ends; the gray box

shows the net change in worst case needs.)

Source: HUD PD&R tabulations of American Housing Survey data

159,000

increase from

population

growth

45,000

decrease from

homeowner

increases

relative to

renter share

20,000 increase

from renter

income changes

adding to VLI

numbers

218,000 decrease

from more

affordable and

available units for

VLI renters

Household

formation

Renter share Renter income

changes

Rental

assistance

gap

Affordable unit

competition

135,000 increase

from reduction in VLI

renters receiving

assistance

Net change:

50,000

increase

Exhibit ES-2. Contributions to Worst Case Needs

from Household Formation and the Rental

Assistance Gap were Offset by Less Competition

for Affordable Units from 2017 to 2019

Contributing most to the increase in worst case needs

were household formation, primarily among households

with extremely low incomes, and the widening of the

already unsettling gap in housing assistance relative to

households eligible to receive it. Although rising incomes in

a strengthening economy lifted some renter households with

incomes between 30 and 50 percent of AMI out of the VLI

population, there were larger increases in the number of ELI

renters. The only demographic factor that helped improve

the worst case needs picture between 2017 and 2019 was

the modest increase in homeownership rates. The primary

force helping to reduce housing problems in 2019 could be

considered economic rather than demographic: improvement

in the availability of affordable units in the housing market

associated with slightly less severe competition.

The four demographic factors together created the potential

for a substantial net increase in worst case needs between

2017 and 2019 by increasing the unassisted VLI renter

population. The market’s easing of competition among

renters for affordable units, however, successfully offset

much of the potential increase in worse case needs through

2019. The net increase attributed to demographic changes

was reduced an estimated 81 percent by modest expansion

in rental supply and associated changes in the availability

of affordable VLI units, as indicated by the fifth bar of

the exhibit. If the supply of affordable rental units fails to

increase at the same rate as the renter population, greater

demand would be expected to increase competition for

affordable units, drive up rents, and increase the prevalence

EXECUTIVE SUMMARY

x WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

of worst case needs. Competition may include higher-

income households choosing to occupy units that would be

affordable to households with significantly lower incomes,

making those units unavailable to those with greater needs.

By the end of 2019, an increase of 771,000 rental units

affordable and available to VLI renters exceeded the increase

of 321,000 VLI renter households. For ELI renters, however,

the increase of affordable and available units by 137,000 fell

short of the increase of 200,000 renter households in this

group. Additions to the total supply of rental units, including

converted owner-occupied units, were limited to a 1-percent

increase between 2017 and 2019.

With modestly improved supply, rents did not increase as

much as renter incomes between 2017 and 2019. Median

housing costs

increased by 8.1 percent, building on a similar

increase in the prior period incomes (see exhibit 3-2). The

mean change in renter income during 2017 to 2019 was 13

percent, greater than the median change of 10.8 percent

(thus showing that the distribution of gains was skewed).

This mean value was influenced by a 17 percent increase

in income for the subgroup of households with incomes

exceeding 120 percent of AMI and, on the other hand, by an

increase of about 4 percent for ELI renters—which was less

than one-half that of any other income group (see appendix

A-14).

Similarly, compared with an increase in median housing

costs of 8.1 percent, mean housing costs increased by 9.1

percent among all renter households. For the ELI renter

subgroup, however, housing costs increased by 12 percent

during the 2-year period. As a result, the housing costs of

ELI renters increased almost three times faster than their

incomes from 2017 to 2019. This growing financial challenge

explains why the prevalence of severe problems among ELI

renters increased from 48.1 percent in 2017 to 49.2 percent

in 2019.

Access of VLI renter households to a sufficient supply of

naturally affordable rental units or assisted units is critical

to the extent of the worst case needs problem. Exhibit ES-3

presents how the availability of rental units affordable to VLI

households has responded to demand trends over the past

10 years.

Those housing costs include rent, utilities, property insurance, land rent, and association fees but exclude any separate security deposit or parking fees.

Source: HUD-PD&R tabulations of American Housing Survey data

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

20.0

40.0

60.0

80.0

100.0

2009 2011 2013 2015 2017 2019

Worst Case Needs (thousands)

Units per 100 Renter Households

Exhibit ES-3. Trends in Housing Supply

Mismatch and Worst Case Needs, 2009–2019

Affordable and available

units for ELI renters

(per 100 households)

Affordable and available

units for VLI renters

(per 100 households)

Worst Case

Needs

(thousands)

ELI = extremely low income. VLI = very low income.

Source: HUD PD&R tabulations of American Housing Survey data

Although the supply of rental units slightly expanded in

2019, rental housing production has significantly lagged

behind household formation since 2010. At the same time,

the number of households with rental assistance has risen

only modestly and has not kept pace with the increase in

the number of VLI households. Rental units that have been

added tended to be in higher-rent properties. As a result,

the ratio of affordable and available units to VLI renters

followed a downward path from 2009 to 2017. After 2017,

there was some improvement, with the ratio increasing from

59 units per 100 renter households in 2017 to 62 units per

100 renter households in 2019. For ELI households, the ratio

of affordable and available units did not change—there were

only 40 affordable and available units for every 100 ELI renter

households in both 2017 and 2019. A more generous supply

of affordable and available housing for VLI households did

not generate proportional benefits for their ELI counterparts.

Increasing affordable housing supply by providing rental

and sustainable homeownership options for households

across the income spectrum—including by expanding rental

assistance, particularly for poorer households—therefore, will

be important for reducing worst case needs during the next

decade.

Availability ratios are as important as worst case needs

measurement for understanding affordable housing

problems. Availability ratios demonstrate the critical role of

rental assistance in expanding affordable housing options

for VLI renters: among VLI renters with access to affordable

EXECUTIVE SUMMARY

xiWORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

housing, a large share have such access by virtue of the

rental assistance they receive. Availability ratios, when

compared with affordability ratios, also make clear the

striking competition for the most affordable housing. For each

affordability bracket, renters with incomes above the bracket

levels occupy large shares of units affordable to households

within the bracket. Such crowding-out affects 43 percent

of the units affordable to ELI renters, 40 percent of units

affordable at incomes of 30 to 50 percent of AMI, and 37

percent of units affordable at incomes of 50 to 80 percent of

AMI. As higher-income renters defer home purchases, they

continue to compete for affordable units and sustain rental

demand, limiting the availability of affordable rental units

for lower-income renters by two fifths. In short, the effect

of weak growth in the rental housing supply, a shortage of

rental assistance, and strong competition for available rental

units from higher-income renters seems to be having the

most detrimental effect on the availability of units affordable

to renters with incomes at and below 30 percent of AMI.

Improving the availability of affordable rental units for ELI

renters will be crucial to achieving substantial decreases in

worst case needs.

Conclusion

Worst case housing needs worsened slightly, but statistically

insignificantly, between 2017 and 2019 due to household

formation (new households formed as a result of population

increase) and widening of the rental assistance gap for

eligible very low-income households. Reductions in worst

case needs generally result when economic growth improves

household incomes and when the production of affordable

housing is sufficient to reduce market rents or, alternatively,

when rental assistance rates increase.

The leveling between 2017 and 2019 of housing problems

among the nation’s VLI renter households is primarily

attributable to a more adequate response of the housing

market to quantitative changes in demand for VLI-affordable

rental units. The progressive response of the housing

market blunted the potential increase in worst case needs

cases resulting from demographic and economic factors—

especially household formation, income loss, and the

widening gap between renter households eligible to receive

housing assistance and those receiving it. Households

reporting assistance decreased slightly even as the number

of VLI renter households expanded. An improved housing

market response that included modest housing production

helped increase the availability of affordable units for VLI

renters, although ongoing demand for more-affordable units

from higher-income renters continues to constrain availability

and prevent major reductions in worst case needs cases.

Three of five ELI renter households and three of eight VLI

renter households continued to lack access to affordable

and available housing units as of 2019. Rental housing

assistance—such as that offered by HUD programs,

other federal programs, states, or localities—helps many

vulnerable renter households who have such limited incomes.

Among VLI renters in 2019, 27 percent of households

were able to avoid worst case needs because they had

rental assistance. But rental assistance is in short supply:

because of inadequate funding, only about one in four

eligible households receives rental assistance. Another 30

percent were able to avoid severe housing problems in the

unassisted private rental market. The remaining 42 percent,

however, were left with worst case needs for assisted or other

affordable housing, and almost three-fourths of those were

ELI households.

As the economy grew during 2017 to 2019, the production

and supply of affordable homes remained insufficient to

satisfy the demand for affordable and available units by very

low-income renters. A broad strategy at the federal, state,

and local levels has long been needed to continue to grow

the economy, support market production and access to

affordable homes, and provide rental assistance to the most

vulnerable households. Additionally, beginning in early 2020,

economic stresses associated with the COVID-19 pandemic

created new critical needs to prevent eviction.

Several Congressional pandemic responses, as discussed in

the Special Addendum, provided increased federal housing

resources during 2020 and 2021. Stimulus funding provided

$46 billion to states for Emergency Rental Assistance to

assist tenants and landlords with pandemic-related rent

arrears. HUD’s FY 2021 appropriation increased subsidies

for public and assisted housing operations by $3.2 billion

from FY 2020 levels to address lost tenant rent contributions.

The American Rescue Plan Act of 2021 provided $5 billion

that funded 70,000 new Emergency Housing Vouchers. The

Federal Housing Finance Agency announced a $711 million

allocation for the Housing Trust Fund in 2021, representing

twice the state funding for housing production as available

in 2020. The Treasury Department’s $10 billion Homeowner

Assistance Fund will help prevent foreclosures that ultimately

could increase the number of renter households with

worst case needs. The President’s FY 2022 Budget further

proposes to fund an additional 200,000 Housing Choice

Vouchers and increase housing production with $500 million

of increased funding for the HOME Investment Partnerships

program and $180 million to support 2,000 units of new

permanently affordable housing for older adults and people

with disabilities. Such housing supply-side and demand-

side resources are complemented by HUD’s work to reduce

regulatory barriers to affordable housing production and

EXECUTIVE SUMMARY

xii WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

provide technical assistance to local governments to assist in

removing barriers that drive up housing costs.

With the expected impact of the COVID-19 pandemic and

associated economic difficulties in 2020 and 2021, worst

case housing needs have potential to increase substantially

before HUD’s next report. A comprehensive approach to

housing policy is sorely needed to address the long-standing

and evolving challenge of worst case housing needs.

EXECUTIVE SUMMARY

1WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Extent and Nature of

Worst Case Needs

The U.S. Department of Housing and Urban Development (HUD) is the largest

federal provider of affordable rental housing. In response to a request by

Congress in 1991, HUD’s Office of Policy Development and Research (PD&R)

periodically reports on the severity of worst case needs for affordable rental

housing, as collected in the biennial American Housing Survey (AHS). This

report is the 18th in the series of core reports.

Extent of Worst Case Needs in 2019

HUD analysts examined the 2019 AHS data to understand the evolving

dimensions of a persistently expanding shortage of decent and affordable rental

housing for lower-income households. The basic facts presented and examined

in the following pages are—

- In 2019, 7.77 million renter households had worst case needs (see

exhibit 1-1). These are renters that have very low incomes,

lack housing

assistance, and have either severe rent burdens or severely inadequate

housing (or both).

PD&R supplements the core reports on worst case needs with periodic topical reports. For a list

of previous titles, see appendix D.

Very low income and extremely low income refer throughout this report to the income levels of

renters. Very low incomes (VLI) are those incomes of no more than 50 percent of the area median

income (AMI), and extremely low incomes (ELI) are those incomes of no more than 30 percent

of AMI—typically below the poverty line. HUD programs use AMI based on local family incomes

with adjustments for household size, more precisely known as HUD-adjusted area median family

income, or HAMFI (see appendix E). Nationwide, median very low-income and extremely low-

income levels were $32,250 and $21,330 per year, respectively, in 2019 (see exhibit 3-2). These

income levels are for a family of four. ELI and VLI families may have incomes much less than

these national thresholds if they have fewer than four members or live in areas with lower median

family incomes.

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

Section 1

2 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

0.0

6.0

8.0

9.0

5.0

7.0

4.0

2.0

3.0

1.0

2009 2011 2013 2015 2017 2019

Renters with worst case needs (millions)

7.10

8.48

7.72

8.30

Exhibit 1-1. Growth in Worst Case Housing Needs, 2009–2019

Source: HUD PD&R tabulations of American Housing Survey data

7.72

7.77

10.0

- Between 2017 and 2019, the number of very low-income

renters with worst case needs increased by a statistically

insignificant 0.6 percent, following a 7.1-percent

decrease observed during the 2015-to-2017 period. The

50,000 additional cases of worst case needs reflect a

significant increase of 225,000 cases among renters

with extremely low incomes, offset by a decrease of

175,000 cases among renters with incomes between 30

and 50 percent of AMI.

- Consistent with long-term trends, the primary problem

for worst case needs renters in 2019 was severe rent

burden resulting from insufficient income relative to rent.

Among all renter households, much of the 10.8-percent

increase in median incomes between 2017 and 2019

was consumed by an 8.1-percent increase in median

housing costs for renters. Severely inadequate housing

accounted for only 2.5 percent of worst case needs.

- Positive economic forces prevented worst case needs

from growing between 2017 and 2019. Competition for

affordable units eased slightly, and households moved

toward homeownership, a welcome contrast with the

recession-related challenges of mortgage foreclosures,

unemployment, and shrinking renter incomes that

increased worst case needs by 2.57 million households,

or 43.5 percent, between 2007 and 2011.

- In 2019, there were 18.39 million VLI renter households,

a 1.8-percent increase from 2017 levels that partially

reversed the 6.1-percent decrease seen in the 2015-

to-2017 period. In 2019, 42.2 percent of VLI renter

households and 49.2 percent of ELI renter households

had worst case needs.

- Worst case needs remained unchanged as a proportion

of U.S. households from 2017 to 2019 at 6.3 percent.

WHICH HOUSEHOLDS CAN HAVE WORST CASE

NEEDS?

By denition, households that can have worst case

needs are households that—

• Are renters.

• Have very low incomes—that is, incomes of no

more than 50 percent of the area median income

(as adjusted for family size).

• Do not receive housing assistance.

SEVERE PROBLEMS TRIGGER WORST CASE

NEEDS

Two types of severe problems determine whether

households have worst case needs:

1. Severe rent burden, which means that a renter

household is paying more than one-half of its

income for gross rent (rent and utilities).

2. Severely inadequate housing, which refers

to units having one or more serious physical

problems related to heating, plumbing, and

electrical systems or maintenance (problems

are listed in appendix E).

- Housing assistance prevents millions of renters from

experiencing worst case needs. The shortfall of housing

assistance relative to need increased between 2017 and

2019 as the number of assisted renters decreased by 2.7

percent. The share of VLI renter households receiving

housing assistance decreased by 1.3 points to 27.5

percent during the period.

- An important dimension of the affordable housing supply

gap is that affordable units are not necessarily available

to the renters who need them most; higher-income

renters occupy substantial shares of units that would be

affordable to the lowest-income renters.

With these key facts in mind, section 1 explores the current

extent and the demographic characteristics of worst case

needs—which households have such needs and what their

situations are.

Inadequate Housing and

Inadequate Income

Of the two types of severe problems that make up worst

case needs, severe rent burden is, by far, the more frequent

problem. As exhibit 1-2 illustrates, 97.5 percent of all

worst case needs renters, or 7.57 million households, had

severe rent burdens in 2019. Paying one-half (or more) of

a limited total income for rent leaves very little income for

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

3WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

other essentials, such as food, medical care, transportation

expenses, education, and child care.

Source: HUD PD&R tabulations of American Housing Survey data

Exhibit 1-2. Severe Rent Burdens Drove

Worst Case Needs in 2019

Severe rent

burden only

Both severe

problems

Severely

inadequate

housing only

95.2%

2.3%

2.5%

N = 7.766 million renters with worst case needs.

Severely inadequate housing alone made up only 2.5

percent of worst case needs in 2019; 4.8 percent of renter

households with worst case needs, 374,000, had severely

inadequate housing, either alone or in combination with

severe rent burdens. Although severe housing inadequacies

represent only a small fraction of severe housing problems,

the number and share of worst case needs households

experiencing such quality problems remained almost

unchanged, with a 0.3-point decrease in the 2017-to-2019

period.

That severely inadequate housing causes such a small

fraction of worst case needs is the result of a decades-long

trend of improvements to the nation’s housing stock. More

stringent building codes prevent the construction of units

without complete plumbing or heating systems, and obsolete

units are demolished each year.

In addition, a portion of

severe physical inadequacies reported in the AHS likely

results from or reflects maintenance or upgrade activity

occurring in occupied units. Among all renter households,

3.0 percent of those with very low incomes and 1.0 percent of

those with higher incomes

had severely inadequate housing

in 2019. Nevertheless, the housing stock is continually

aging, and thousands of renters continue to live in severely

inadequate units. The costs associated with repairing severe

Changes in the overall housing stock are primarily driven by new construction and losses due to demolition and natural disasters (Econometrica, 2016).

Homeowners reported severely inadequate housing at even lower rates than renters: 1.9 percent of VLI homeowners and 0.5 percent of homeowners with

higher incomes had severely inadequate housing. See exhibit A-1B.

Divringi et al. (2019) estimated repair costs associated with quality deficiencies identified in the 2017 AHS and found that units occupied by renters with incomes

at or below the poverty line accounted for $25.5 billion, or 56.7 percent, of the aggregate estimated repair costs associated with rental units in the United States.

Older single-family and multifamily units occupied by poor renters had higher median repair cost estimates—$2,096 and $1,355, respectively—than newer units.

Similarly, Wallace et al. (2019) found that repair costs increase with the degree of housing inadequacy as measured by the AHS, with median costs for repairing

moderately and severely inadequate units estimated at $2,440 and $3,346, respectively.

quality deficiencies present another formidable barrier to the

ability of lower-income households to improve their housing

conditions. Landlords offering lower-priced units for rent

may similarly delay or avoid high maintenance and repair

expenses as units age.

PROGRESS IN REDUCING HOMELESSNESS

Individuals and families experiencing homelessness

clearly have the greatest need for affordable

or assisted housing. People experiencing

homelessness, however, are not included in ofcial

estimates of worst case needs because the AHS

covers only housing units and the households

that live in them, and people experiencing

homelessness, by denition, do not live in a housing

unit and are not surveyed by the American Housing

Survey (AHS).

In the 2019 Annual Homeless Assessment Report

to Congress, HUD estimated that 568,000 sheltered

and unsheltered homeless people were in the

United States during a given night in January 2019.

Most of these, 63 percent, were staying in residential

programs for people experiencing homelessness,

and the remaining 37 percent were staying in

unsheltered locations (HUD-CPD, 2019).

Since 2007, total homelessness on a given night

has declined by 15 percent, and homelessness

among families with children continues to decline.

This long-term progress, however, is threatened

by recent local trends among unsheltered and

chronically homeless populations in certain areas

of the country. Total homelessness has increased

modestly since 2016. The increase has been driven

by a growing unsheltered population in high-cost

markets, particularly in California, even as total

homelessness has continued to decline outside of

those areas. Although the number of people staying

in emergency shelters and transitional housing

programs continues to decline as the inventory of

beds in rapid rehousing programs increases, the

number of people staying in unsheltered locations

grew by 20 percent between 2016 and 2019.

Between 2016 and 2019, families with children

experiencing homelessness decreased by 12

percent, but the number of individuals experiencing

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

4 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

homelessness increased by 11 percent. Chronic

homelessness among individuals grew by 24

percent even as the count of veterans experiencing

homelessness decreased by 6 percent during the

same period.

The AHS samples both occupied and vacant residential housing

units but excludes places such as group quarters or motels where

homeless persons may be sheltered (Census-HUD, 2017: 3–5).

Prevalence of Worst Case

Needs by Income

Because most cases of worst case needs are triggered by

severe rent burdens, the adequacy of household incomes

relative to rents of available units is crucial. Among the 18.39

million VLI renter households in 2019, 42.2 percent had worst

case needs (exhibit 1-3). The VLI category includes ELI renters,

who had an even greater prevalence of worst case needs

at 49.2 percent. ELI renter households constituted a large

share (63.9 percent) of VLI renter households in 2019, yet

their 1.7 percent increase between 2017 and 2019 was about

the same as the 1.9 percent increase of the 30–50 percent

of AMI population. ELI renter households experienced worst

case needs at a greater frequency in 2019, 49.2 percent, than

they did in 2017, 48.1 percent. As a result of their increased

prevalence of severe problems, ELI renter households account

for 74.4 percent of worst case needs in 2019, up from 72.0

percent in 2017, reflecting the difficulty of finding decent,

affordable housing at ELI levels.

Exhibit 1-3. Extremely Low-Income Renters Were Most

Vulnerable to Worst Case Needs in 2019

0–30% AMI

>30–50%

AMI Total VLI

Number of renters

(thousands)

11,748 6,640 18,388

Number that are worst

case needs renters

(thousands)

5,780 1,986 7,76 6

Percentage that are worst

case needs renters

49.2 29.9 42.2

AMI = area median income (HUD adjusted). VLI = very low income.

Source: HUD-PD&R tabulations of American Housing Survey data

Of the 6.0 million ELI renter households without worst case needs, 4.0 million (or 66.2 percent) received rental assistance subsidized by HUD or other federal,

state, or local programs. In other words, only 2.0 million of the 11.7 million ELI renter households nationally (or 17.2 percent) avoided severe housing problems in the

unassisted private market in 2019. See exhibit A-1A.

Previous Worst Case Needs reports have documented much more rapid growth during the preceding 10 years. During 1999 to 2009, the number of worst case needs

increased from 4.86 million to 7.10 million, an increase of 46 percent.

Worst Case Needs Prevalence Among

U.S. Households

The estimated number of worst case needs increased by a

statistically insignificant 50,000 cases (or 0.6 percent) from

2017 to 2019, halting the decline in worst case needs observed

during the previous biennial period from 2015 to 2017. Over

the 10-year span from 2009 to 2019, however, the number of

households with worst case needs had grown by 9.5 percent,

or 671,000 households (exhibit 1-4).

Worst case needs

minimally decreased as a proportion of U.S. households during

the most recent 2-year period, from 6.4 percent in 2017 to 6.3

percent in 2019, but remains higher than the prerecession level

of 5.3 percent in 2007.

Exhibit 1-4. Growth in Worst Case Needs

Among All U.S. Households

2009 2011 2013 2015 2017 2019

All households

(millions) 111.86 115.08 116.03 118. 29 121.56 124.14

Renters

with worst

case needs

(millions) 7.10 8.48 7.72 8.30 7.72 7.7 7

Worst case

needs as

percentage of

all households 6.34 7.3 6 6.65 7.0 2 6.35 6.26

Source: HUD-PD&R tabulations of American Housing Survey data

As reflected in the most recent year of exhibit 1-4, this report

captures housing need in the period immediately prior to the

onset of the COVID-19 pandemic and associated economic

recession early in 2020. The financial shock to the labor market

and household incomes has potential to cause substantial

increases in worst case needs when next measured with the

2021 American Housing Survey. The major federal legislative

response, however, complicates the expectation and

measurement of worst case needs. A Special Addendum in

this report discusses the recession, key features of the relief

legislation, impacts of the recession on housing outcomes, and

potential implications for future worst case needs estimates.

Because the problem of worst case needs is primarily one of

a scarcity of units with affordable rents relative to the number

of renters with very low incomes, the balance of section 1

examines the demographics of the renters who have those

problems. Section 2 explores the dimensions of the inadequate

supply of affordable rental units, and section 3 summarizes and

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

5WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

integrates supply and demand issues to shed light on the

root causes and shifting dimensions of this persistent national

problem.

Demographics of Worst Case

Needs

Worst case needs are an economic reality for many of the

nation’s VLI renter households. The severe housing problems

that trigger worst case needs are widespread for such

households, yet notable variations exist among subgroups of

the population.

Worst Case Needs by Race and

Ethnicity

Worst case needs were found across all types of

communities, racial groups, and ethnic lines. Both similarities

and differences emerged when examining the three largest

racial and ethnic groups: non-Hispanic White, non-Hispanic

Black, and Hispanic. This section also examines detailed

subgroups within the “all other races and ethnicities” group

to the extent supported by the AHS sample size. (See

exhibit 1-7.)

During 2019, non-Hispanic White renters accounted for

the largest number of households with worst case needs

(3.6 million) by race and ethnicity. Non-Hispanic Whites

accounted for 46.7 percent of worst case needs, followed by

Hispanics, with 24.7 percent; non-Hispanic Blacks, with 20.4

percent; and renters of all other races and ethnicities, with

8.2 percent. Together, the three largest race and ethnicity

groups accounted for 91.8 percent of worst case needs in

2019, and households headed by people of color accounted

for more than one-half—53.3 percent—of worst case

needs.

As suggested by exhibit 1-5, very low-income renters do not

experience worst case needs at a uniform rate. During 2019,

worst case needs affected 43.7 percent of VLI renters among

both non-Hispanic Whites and the other race and ethnicity

group—slightly less than the 45.1 percent prevalence among

In this section, race and ethnicity of households is based on the race and ethnicity of the householder as reported in the AHS data. People of color or households

of color refers to households that are not non-Hispanic White. “Other” is used in several ways. In the finest analysis that is consistently feasible with the AHS

data, “all other races and ethnicities” is the fifth of five main categories, comprising households of color in subgroups not otherwise listed or in a combination

of subgroups. In the more detailed breakout of exhibit 1-7, “Other race or ethnicity” has the same meaning but refers to a smaller residual category of households

because the exhibit provides additional categories by breaking down the AHS data to the full extent feasible. Finally, some portions of the narrative use “other” in

its plain sense of “ones not specified in the present context,” for example when groups of color are being compared to another group of color.

Similarly, the three largest race and ethnicity groups accounted for 92.1 percent of all VLI renter households nationally, and households of color accounted for

54.9 percent of VLI renter households.

See exhibit A-9.

Hispanics. Prevalence was lower among non-Hispanic

Blacks, with 36.1 percent having worst case needs. The lower

prevalence of worst case needs among Black households

reflects greater likelihood that Black households receive

housing assistance. Among Non-Hispanic Black renters with

very low incomes, 40.2 percent report housing assistance,

compared with only 24.3 percent for Non-Hispanic Whites,

23.2 percent for Hispanics, and 27.0 percent for all other

races and ethnicities (see exhibits A-9 and A-1A). Among

other factors contributing to this disparity, the geographic

distribution of housing assistance plays a prominent role, as

discussed in Section 2.

Variation in rates of housing assistance among VLI renter

households contributed to variation in the prevalence of

worst case needs and the likelihood that households avoided

severe housing problems unassisted in the private market.

Non-Hispanic White and Hispanic VLI renter households

had the best odds of avoiding severe housing problems in

the private market in 2019—32.5 percent of non-Hispanic

White VLI renters and 33.1 percent of Hispanic VLI renters

avoided severe problems without housing assistance. Only

about one-fourth of non-Hispanic Black and other VLI renter

households—23.8 and 29.2 percent, respectively—avoided

severe problems in the private market without housing

assistance.

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

6 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit 1-5. Very Low-Income Renters from All

Household Types and Racial and Ethnic Groups

Experienced Worst Case Needs in 2019

Source: HUD-PD&R tabulations of American Housing Survey data

Subgroup VLIR with worst case needs (percent)

Subgroup share of all very low-income renters (percent)

30%

10%

40%

50%

20% 30% 40%

50%

Other race/ethnicity

Hispanic

Non-Hispanic Black

Non-Hispanic White

VLIR = very low-income renters.

Source: HUD PD&R tabulations of American Housing Survey data

The position of the markers in exhibit 1-5 reflects each

group’s share of VLI renter households and the rate at which

they experience worst case needs. Groups account for a

greater share of worst case needs as their markers move

toward the upper-right quadrant. As a share of VLI renter

households, the subgroups based on race and ethnicity span

a range of 37.2 percentage points, but the prevalence of

worst case needs varied by only 9.0 percentage points. The

all other races/ethnicities group and Hispanic households are

particularly more likely to have worst case needs than other

subgroups, relative to their share of the VLI renter population.

Other race and ethnicity groups, not included in the three

main race and ethnicity categories (non-Hispanic White, non-

Hispanic Black, and Hispanic), represent a small proportion

of VLI households (7.9 percent), thus appearing in the

upper-left quadrant of the exhibit. About 44 percent of these

households, however, experience worst case housing needs.

Non-Hispanic White households account for a relatively large

share of the VLI renter household population (45.1 percent)

and have a relatively large prevalence of worst case needs

(43.7 percent), thus appearing in the upper-right quadrant

of the exhibit. Except for non-Hispanic Black households, all

subgroups have a larger share of their VLI renter household

population afflicted with worst case needs than the national

average of 42 percent.

Exhibit 1-6. Growth in Worst Case Needs Among

All Racial and Ethnic Groups, 2009-2019

Source: HUD PD&R tabulations of American Housing Survey data

Worst case needs (thousands)

Other Race/Ethnicity

Hispanic

Non-Hispanic Black

Non-Hispanic White

6,000

7,000

8,000

9,000

2,000

1,000

3,000

4,000

5,000

2009 2011 2013 2015 2017 2019

437

587

601

621

620 633

1,582

1,971

1,845

2,104

1,884 1,922

1,640

1,820

1,657

1,800

1,578 1,588

3,436

4,097

3,618

3,778

3,634 3,623

Exhibit 1-6 shows a slight increase in worst case needs

among households of color between 2017 and 2019 and

a small decrease among non-Hispanic Whites. Hispanic

households had the greatest increase of worst case

needs, 38,000 households, followed by an increase of

13,000 additional cases among renters of other races and

ethnicities, and about 10,000 cases more among non-

Hispanic Blacks. The proportion of VLI renters receiving

housing assistance decreased overall between 2017 and

2019. Renters in the all other races and ethnicities group

experienced the largest decrease in the housing assistance

rate, a reduction of 5.3 points from 32 percent in 2017 to 27

percent in 2019.

Despite those absolute changes in numbers, the prevalence

of worst case needs among VLI renters differed among racial

and ethnic groups. The rate of worst case needs modestly

improved for non-Hispanic Whites, decreasing 2 points from

46 percent in 2017 to 44 percent in 2019, and Hispanics,

decreasing 1 point from 46 to 45 percent. The prevalence

increased by 1.6 points among non-Hispanic Blacks, from

34.6 to 36.1 percent, and by 2 points among the all other

races and ethnicities group, from 42 to 44 percent.

Exhibit 1-6 also illustrates differences in the long-term growth

of worst case needs. Between 2009 and 2019, worst case

needs increased 9.5 percent overall but increased only 5.4

percent for the largest subgroup of VLI renter households:

non-Hispanic Whites. During the last 10 years, non-Hispanic

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

7WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Blacks saw a decrease of 3.2 percent in the number of worst

case needs. Worst case needs expanded much more rapidly

during these 10 years among other people of color, with

increases of 21.5 percent among Hispanics and 44.9 percent

among renters of all other races and ethnicities. In the most

recent biennial period, the population of non-Hispanic White

VLI renter households grew by 4.5 percent. The change

among minorities varied among subgroups; while Hispanic

VLI renters households increased by 4.3 percent, economic

recovery reduced the VLI renter populations of non-Hispanic

Blacks and all other races/ethnicities by 3.7 and 2.8 percent,

respectively.

The series of reports produced by the Native American Housing Needs Study are available at https://www.huduser.gov/portal/pdredge/pdr-edge-

research-022117.html.

Although renters of color who are not Hispanic or non-

Hispanic Black make up a small share (8.2 percent) of

households with worst case needs, the American Housing

Survey sample is large enough to provide detailed national

estimates for some subgroups within this category. Beginning

with the 2017 AHS, HUD has reported estimates of worst

case needs for Asian, American Indian or Alaska Native, and

Native Hawaiian or other Pacific Islander households. This

detail provides additional insight into the composition of the

small but growing group of “other race or ethnicity” renters

(exhibit 1-7).

Exhibit 1-7. Worst Case Needs Among Detailed Race and Ethnicity Subgroups in 2019

Non-

Hispanic

White

Non-

Hispanic

Black

Hispanic Asian American

Indian or

Alaska

Native

Hawaiian or

Other Pacic

Islander

Other

Race/

Ethnicity

0–30% AMI renter households

(thousands)

5,083 3,050 2,651 526 219 60 158

Worst case needs (thousands) 2,582 1,288 1,443 300 (D) (D) 71

Percent with worst case needs 50.8 42.2 54.4 57. 0 (D) (D) 44.9

>30%–50% AMI renter households

(thousands)

3,207 1,343 1,607 273 68 17 125

Worst case needs (thousands) 1,041 301 479 120 (D) (D) 32

Percent with worst case needs 32.5 22.4 29.8 44.0 (D) (D) 25.6

Total very low-income renter

households (thousands)

8,290 4,393 4,258 799 287 77 283

Worst case needs (thousands) 3,623 1,589 1,922 420 66 42 103

Percent with worst case needs 43.7 36.2 4 5.1 52.6 23.0 54.5 36.4

AMI = area median income. Other race / ethnicity = racial or ethnic group not listed individually or consisting of multiple races or ethnicities.

Source: HUD-PD&R tabulations of American Housing Survey data

Exhibit 1-7 shows that Asian households account for more

than one-half of worst-case households within the all

other races and ethnicities group presented in exhibit 1-6,

representing 4.3 percent of all households with worst case

needs. The prevalence of worst case needs among Asian VLI

renter households, 52.6 percent, was higher than among any

other racial or ethnic group except the small Native Hawaiian-

Pacific Islander group.

Together, American Indian or Alaska Native and Native

Hawaiian or other Pacific Islander households accounted

for 2.7 percent of all cases of worst case needs in 2019.

Although those estimates provide one indication of the

prevalence of severe housing affordability and quality

problems among those populations, HUD’s Native American

Housing Needs Study also found that overcrowding and

doubling up were far more common among Native American

households compared with other households in the United

States.

Thus, estimates of worst case housing needs should

be viewed as one component of a larger body of evidence on

housing problems among American Indian or Alaska Native

and Native Hawaiian or other Pacific Islander households in

tribal and urban areas.

Worst Case Needs by Household

Type

The composition of different households reflects variations

in their stage of life, income and resources, and housing

needs. Other nonfamily households (single adults, unmarried

couples, and roommates) constituted the largest share of

households experiencing worst case needs in 2019—32.6

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

8 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

percent—followed by families with children, with 29.2

percent; older adult households without children (hereafter,

older adult households), with 28.9 percent; and other family

households, with 9.3 percent (exhibit 1-8).

Exhibit 1-8. Very Low-Income Renters from All

Household Types and Racial and Ethnic Groups

Experienced Worst Case Needs in 2019

Source: HUD-PD&R tabulations of American Housing Survey data

Subgroup VLIR with worst case needs (percent)

Subgroup share of all very low-income renters (percent)

30%

10%

40%

50%

20% 30% 40%

50%

Other family

Other nonfamily

Older adults Families with children

VLIR = very low-income renters.

Source: HUD PD&R tabulations of American Housing Survey data

As a share of VLI renter households, the subgroups based

on household type span a range of 21.8 percentage

points, but the prevalence of worst case needs varied by

only 5.8 percentage points. Exhibit 1-8 shows that three

household types share very similar proportions of VLI renter

households—30.7 percent families with children, 30.3

percent older adult families, and 30.0 percent other nonfamily

households—while only 9.0 percent are other family

households. “Other” nonfamily households and “other” family

households are somewhat more likely to have worst case

needs than other subgroups. Compared with the average

prevalence of 42.2 percent, 43.7 percent of other family

households and 45.9 percent of other non-family households

have worst case needs.

The variations in prevalence among these subgroups,

although limited, may reflect the result of housing programs

prioritizing families with children, older adults, and veterans.

Families with Children. The largest VLI group by household

type, families with children, was the only household type that

saw a decrease in worst case needs cases between 2017

and 2019. The number of families with children having worst

case needs decreased by 300,000 during the 2017-to-2019

period, contributing to a total reduction of 965,000 cases

since their housing problems peaked in 2011.

See appendix E for more on the composition of household types. Families with children may include a parent with child and unmarried partner. Either family or

nonfamily households may include same-sex partners. The Household Demographics table for AHS 2019 in the AHS Table Creator is illustrative: https://www.

census.gov/programs-surveys/ahs/data/interactive/ahstablecreator.html.

Worst case needs decreased, in part, because VLI renter

households with children decreased by 545,000 between

2017 and 2019. This decrease took place wholly among

households with incomes at and below 30 percent of AMI.

Along with rising incomes, the decline in the number of renter

families with children appears to have had a causal role.

Nationally, the number of renter households with children was

reduced by 6.5 percent between 2017 and 2019.

Although progress is being made—partly attributable

to income gains among these households—the number

of families with children experiencing worst case needs

remained above prerecession levels. The share of VLI renter

households with children experiencing worst case needs

moderately decreased by 1.3 percentage points from the

2017 level to 40.2 percent in 2019, and the percentage

reporting housing assistance also declined from 26.8 to 25.6

percent.

Source: HUD-PD&R tabulations of American Housing Survey data

Worst case needs (thousands)

Other

nonfamily

Other

family

Families

with children

Older adults,

no children

6,000

7,000

8,000

9,000

2,000

1,000

3,000

4,000

5,000

2009 2011 2013 2015 2017 2019

2,401

2,969

2,676

2,758

2,497

2,535

633

801

741

805

716

720

2,734

3,236

2,833

2,888

2,571

2,271

1,328

1,470

1,471

1,853

1,932

2,241

Exhibit 1-9. Growth in Worst Case Needs

Among All Household Types, 2009–2019

Without housing assistance, substantially more cases of

worst case needs would occur. Among VLI renter households

with children, 1.45 million reported having rental assistance

in 2019 and, by definition, could not have worst case needs.

Only about one in four VLI renter households with children

received housing assistance, which helps account for the

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

9WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

fact that the greatest share of worst case needs occurred in

such families.

Older Adult Households. The number of older adult

households experiencing severe housing problems has

steadily climbed over the past decade. During 2019, 2.24

million older adult

renters had worst case needs, an

increase of 607,000 since 2017, even as 73,000 more of

these households reported receiving rental assistance in

2019. The increase is largely attributable to the growing

population of older adult VLI renter households. The

proportion of older adult VLI renter households with worst

case needs was 40.3 percent in 2019, marginally greater

than the rate for families with children and representing a

1.3-point surge since 2017. Although nearly 4 in 10 older

adult VLI households received housing assistance in 2019—

a 2.9-point decrease since 2017—aging baby boomers are

likely to continue to be a key demographic facing housing

problems in the years to come.

Other Family Households. After considering families with

children and older adult households, other households can

be divided into those that include multiple members of a

given family and those that do not. Other family households

include those such as married couples who are childless,

one or more parents with adult children at home, adult

siblings sharing an apartment, and householders boarding

an older adult parent.

Other family households constitute the smallest category

in exhibit 1-9, contributing 720,000 worst case households

in 2019. The rate of worst case needs among VLI renter

households in this group was 43.7 percent, exceeding the

prevalence for either families with children or older adult

Estimates of the number of rental households that reported receiving rental housing assistance are presented for various subgroups in the exhibits of appendix

A. AHS estimates of assisted very low-income renters in this report rely on self-reported data, which primarily include HUD-assisted households and may also

include households assisted through other federal, state, or local programs, such as U.S. Department of Agriculture rental housing subsidies. As expected,

HUD administrative data matching procedures suggest that excluding households assisted by non-HUD programs reduces the number of households classified

as receiving housing assistance. For the purposes of this report, however, households receiving assistance from a non-HUD program are not classified as

having worst case housing needs. Because administrative data matching across several federal, state, and local agencies is not feasible, AHS self-reported

assistance is the preferred measure of housing assistance for this report. The aggregate numbers of households served by HUD’s primary rental assistance

programs, based on administrative records, are outlined in appendix C.

HUD defines older adult households as those having a household head or spouse who is at least age 62 and including no children younger than age 18.

Harvard’s Joint Center for Housing Studies projects that aging baby boomers will swell the nation’s population aged 65 or older by 11.1 million over the next

decade, fueling both the housing remodeling market and demand for smaller, accessible homes (JCHS, 2019).

Among “other family” very low-income renter households, 41.3 percent include a married couple, 56.6 percent have a female householder, 63.5 percent have a

householder of color, and the mean household size is 2.47 persons. See exhibit A-6A.

Within HUD’s largest rental assistance program, the Housing Choice Voucher (HCV) program, the majority (70 percent) of households served are either families

with children or older adults (Picture of Subsidized households, 2020; https://www.huduser.gov/portal/datasets/assthsg.html). In addition to the data in this

report showing that only one-fourth of eligible VLI renter households receive housing assistance, the scarcity of HCV program resources is further evidenced by

long waiting lists. Administrative data indicate that in 2020, on average, eligible households who had applied for a voucher and received it had waited 2.3 years

(Picture of Subsidized households, 2020). Public housing authorities have the discretion to establish local preferences for choosing which households to assist

based on local housing needs and priorities within this constrained resource environment. See appendix C for additional information on HUD’s rental assistance

programs.

Among nonfamily VLI renter households, 83.6 percent were one-person households in 2017. See exhibit A-6A. The AHS does not include college students

living in institutional housing, but it may include students sharing off-campus housing and other households in which individuals double up to share housing

expenses. As the number of enrolled post-secondary students decreased by 222,000 between 2015 and 2017, a decrease in off-campus student households

might account for part of the reduction in worst case needs for this household type (NCES 2019.)

households. The high rate of worst case needs among

this group declined by 1.3 percentage points between

2017 and 2019, more than for any other household type.

A reduction of 58,000 VLI households in this subgroup

contributed to the reduction, although complex dynamics

within this small, diverse group are likely driving change.

As income trends improve nationally, fewer households of

adult relatives may choose to double up if other adequate

housing options are available. Still, some of those households

may be at a disadvantage in competing for limited available

housing. For example, people younger than 62, in families

without children, may be less likely to be prioritized among

households competing for limited housing assistance

resources.

Other Nonfamily Households. About 5.52 million VLI

renter households in 2019 were other nonfamily households,

making this category the second largest after families

with children. Like most household types, other nonfamily

households also saw an increase in those with very low

incomes between 2017 and 2019.

Worst case needs affected 2.54 million other nonfamily

households in 2019, an increase of 201,000 since 2017 and

more than the number found among any other household

type. The large number of VLI renters of this group continue

to be afflicted by the highest worst case needs prevalence,

even after the 1.0-point decline to 45.9 percent in 2019. Most

other nonfamily households are single individuals, and the

rest are unrelated people sharing a housing unit.

One-

person VLI households may be less well-equipped to handle

rent increases than those who share housing costs with a

roommate or, in family households, with a family member.

Income shocks may also affect one-person households

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

10 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

more severely than households in which two or more people

contribute resources to the household.

Households Including People with Disabilities. Having

worst case needs can be especially difficult for renter

households that include people with disabilities. Disabilities

can reduce employment options and create additional

difficulties in finding suitable housing at reasonable cost;

features such as elevators that are luxury amenities for some

households may be necessities for people with disabilities.

Additionally, Supplemental Security Income (SSI) benefits are

inadequate to cover housing costs in many markets.

DISABILITY AND ACCESSIBILITY IN THE

AMERICAN HOUSING SURVEY

Since 2009, the American Housing Survey (AHS)

has collected information about the following types

of disabilities:

• Deafness or serious difculty hearing.

• Blindness or difculty seeing, even when

wearing glasses.

• Serious difculty concentrating, remembering, or

making decisions because of a physical, mental,

or emotional condition.

• Serious difculty walking or climbing stairs.

• Serious difculty dressing or bathing.

• Difculty doing errands alone because of a

physical, mental, or emotional condition.

The 2015 AHS also included questions related to the

wheelchair accessibility of housing units and home

modications made to accommodate individuals

with physical disabilities. As in 2011, future AHS

surveys may periodically include topical modules of

questions related to housing accessibility features.

For further information, see the demographics

and accessibility sections of the AHS Codebook

interactive tool (Census-HUD, 2019).

In a similar vein, single adults, unaccompanied youth, or multiple-adult households are more prevalent within the population experiencing homelessness than

are families with children (HUD-CPD, 2018). Likewise, a recent study of community-level predictors of homelessness found that higher population rates of one-

person households were associated with higher homelessness rates (Nisar et al., 2019).

For 2021, the SSI monthly federal benefit rate for an individual living alone is $794 (SSA 2021).

The data about types of limitations are summarized in appendix A, exhibit A-15. Also see HUD-PD&R (2008).

The analysis is limited to people younger than 62 who have disabilities, because many older adults suffer from impairments and activity limitations in

consequence of aging. Note, however, that people younger than 62 who have disabilities may be found in older adult households, as exhibit 1-9 demonstrates.

Households headed by an older adult with disabilities are not excluded if they also include people younger than 62 who have disabilities.

Beginning with the 2009 AHS, respondents have been

asked directly whether household members have any of

six types of disabilities, including four basic functional

limitations—visual, hearing, cognitive, and ambulatory—and

two types of difficulties with activities of daily living—self-

care and independent living. Ambulatory limitations (walking

or climbing stairs) are the most frequently occurring type of

disability, affecting 42.3 percent of VLI renter households

that do not include an older adult with a disability. Cognitive

limitations (serious difficulties concentrating, remembering, or

making decisions) have a higher prevalence, affecting 48.9

percent of these households.

People with disabilities are

found among all four household types discussed previously.

As exhibit 1-10 shows, 2.89 million VLI renter households

(15.7 percent of VLI renter households) have people

younger than 62 reporting at least one of the six measures

of disability.

In 2019, 1.04 million (36.1 percent) of these

households experienced worst case needs, a modest

decrease from 1.30 million (39.8 percent) in 2017. Between

2017 and 2019, the number of VLI renter households with

people younger than 62 who have disabilities increased,

and the number of such households with worst case needs

decreased.

Exhibit 1-10 shows that the prevalence of worst case needs

among VLI renter households with people younger than

62 who have disabilities varied somewhat by household

type. Prevalence during 2019 ranged from 33.6 percent

for other nonfamily households to 40.7 percent for other

family households. Notwithstanding these differences in

prevalence, the largest household categories accounted

for most cases of worst case needs affecting people with

disabilities. Of the 1.04 million households with worst case

needs, 48.1 percent are other nonfamilies and 32.2 percent

are families with children.

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

11WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit 1-10. Worst Case Needs Were Common Among People Younger Than 62 Who Have Disabilities Across All

Household Types in 2019

Families

With

Children

Older Adult

Households

Other Family

Households

Other

Nonfamily

Households

Total

Very low-income renter households

(thousands)

5,654 5,567 1,649 5,518 18,388

Worst case needs (thousands)

2,271 2,241 720 2,535 7,76 6

Percentage with worst case needs

40.2 40.3 43.7 45.9 42.2

Percentage having people younger than 62 who

have disabilities

15.8 1.9 24.4 27.1 15.7

Very low-income renter households having

people younger than 62 who have disabilities

(thousands)

892 105 403 1,495 2,895

Worst case needs (thousands) 336 42 164 503 1,045

Percentage with worst case needs 37.7 40.0 40.7 33.6 36.1

Source: HUD PD&R tabulations of American Housing Survey data

Summary

Worst case needs for affordable rental housing remained a

serious national problem in 2019. Of the 18.39 million VLI

renter households susceptible to severe rent burdens and

severely inadequate housing in 2019, 7.77 million—42.2

percent—faced one or both problems without housing

assistance. Between 2017 and 2019, the number of

households with worst case needs increased by 0.6

percent, following a 7.1-percent decrease observed during

the 2015-to-2017 period. In 2019, the number of worst

case needs cases was higher than it was in 2009. The

data are a reminder of the enduring impact of the financial

crisis and recession that, a decade later, continue to affect

personal finances, credit histories, and affordable housing

opportunities.

Severely inadequate housing continues to be a relatively

minor cause of worst case needs. In 2019, severely

inadequate housing alone produced a mere 2.5 percent

of worst case needs, whereas 97.7 percent of households

with worst case needs had severe rent burdens, including

2.3 percent that had both types of housing problems.

Reflecting the importance of severe rent burdens as a cause

of worst case needs, 6 out of 10 households with worst case

needs (63.9 percent) had extremely low incomes during

2019.

Most racial or ethnic groups, and most household

compositions examined, experienced an increase in worst

case needs from 2017 to 2019. Among racial and ethnic

groups, Non-Hispanic Whites were the only group that had

fewer worst case needs in 2019. Renter households of color

experienced increases in worst case needs during 2017 to

2019. Worst case needs increased by 38,000 cases among

Hispanics, 14,000 cases among households of color who are

not Hispanic or Non-Hispanic Black (that is, “all other”), and

10,000 among Non-Hispanic Black households. Race and

ethnicity subgroup analysis suggests that the small group

of Native Hawaiian-Pacific Islanders, followed by the much

larger group of Asian households, had the highest rates of

worst case needs among the “all other” race-ethnicity group.

Asian households accounted for 5.4 percent of all worst case

needs in 2019.

Among VLI renter households, worst case needs affected

40.2 percent of families with children, 40.3 percent of older

adult households, 43.7 percent of other family households,

and 45.9 percent of other nonfamily households. The number

of worst case needs declined by 300,000 cases among

families with children between 2017 and 2019, but increased

for all other household types. The groups that saw the

most new cases in 2019 were older adults (309,000 cases)

and other non-family households (38,000). In 2019, other

nonfamilies (typically one-person households) accounted

for the greatest share of worst case needs—32.6 percent—

followed by households composed of families with children

(29.2 percent).

Worst case needs affected 36.1 percent of VLI renter

households reporting people younger than 62 who have

disabilities in 2019, moderately less than the 42.2-percent

prevalence among VLI renter households overall. Households

having people younger than 62 who have disabilities

accounted for 13.5 percent of worst case needs.

Section 2 examines how the broad problem of worst case

needs is caused by shortages of affordable housing and is

mitigated by assisted housing on a national basis and within

regional markets.

SECTION 1. EXTENT AND NATURE OF WORST CASE NEEDS

12 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

13WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

Shortage of Affordable

Housing

The United States faces a widespread shortage of rental units that are affordable

to very low-income (VLI) renter households. The supply of affordable units is

especially insufficient to meet the needs of extremely low-income households.

In 2019, only 70 affordable units (including assisted units) existed for every

100 extremely low-income (ELI) renter households nationwide. The presence

of higher-income renters in units that are affordable to ELI renter households

exacerbates this shortage. In 2019, only 40 of those 70 affordable units were

available for occupancy by ELI households. A final factor is that a significant

portion of the affordable and available stock continues to be physically

inadequate and may pose threats to occupants. In 2019, only one-half of the

affordable units (36 of 70 affordable units) were both physically adequate and

available for occupancy for every 100 ELI renter households. The geography of

worst case needs and housing assistance sets a foundation for understanding

the competition for affordable rental housing and its shortages.

Geography of Worst Case Needs

Housing markets are localized and often contain distinct submarkets. VLI and

ELI renter households are more likely than higher-income renters to find their

choice of housing units limited to communities and neighborhoods where

poverty is more common. Such market segmentation and supply restrictions

can manifest differently across market types in terms of renters’ likelihood of

experiencing worst case needs.

As a national survey of modest scale, the American Housing Survey (AHS)

does not support biennial estimates of worst case needs for most individual

metropolitan areas or for highly localized submarkets.

It does, however,

support select estimates of worst case needs for certain large metropolitan

HUD and the Census Bureau have traditionally conducted periodic AHS metropolitan surveys

to supplement the national AHS. In 2015, the AHS was redesigned with a new national and

metropolitan area longitudinal sample to account for changes in geography and attrition of

housing units over time. In 2019, as in 2017, a supplemental sample of housing units in select

metropolitan areas was combined with the national sample to produce metropolitan-level

estimates. Stand-alone surveys were also conducted in some additional metropolitan areas.

Section 2

14 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

areas included in the survey sample.

It also supports a

national examination of four types of metropolitan locations—

central cities, urban suburbs of central cities, rural suburbs

of central cities, and nonmetropolitan areas

—and of four

geographic regions—the Northeast, Midwest, South, and

West. Analysis of AHS data by region and metropolitan status

adds considerable depth to the national picture of worst case

needs.

Worst Case Needs and Housing

Assistance by Region and

Metropolitan Location

A key aspect of the definition of worst case needs is that it

can be understood as an indicator of need for affordable

housing. Because income-based rental assistance and other

deep public subsidies generally make housing affordable,

The redesigned AHS includes a longitudinal sample of the 15 largest metropolitan areas every 2 years and an additional 10 metropolitan areas surveyed on a

rotating basis every 4 years. Select estimates for the metropolitan areas sampled in 2019 are presented in exhibit 2-4 and exhibit A-11B. For more information on

the 2015 AHS redesign, see appendix E.

Both central cities and suburbs are located within metropolitan areas. A central city consists of the largest city within a metropolitan area. Suburbs are within

metropolitan counties but outside central cities. For the purposes of this report, suburban areas are further distinguished as urban or rural based on their

population density. Nonmetropolitan areas fall outside metropolitan counties and tend to be more rural in nature.

AHS questions about receipt of rental assistance are designed to focus on federal housing assistance programs. These data result in an estimate of 5.05 million

elf-reported VLI renter households with housing assistance, which is somewhat more than HUD’s program total. Other potential sources of housing assistance

include the U.S. Department of Agriculture’s Rural Housing Service, other federal agencies, or other state or local programs. Also affecting this comparison,

a small fraction of HUD-assisted renters may have incomes above the VLI threshold because they were admitted to programs under local policy preferences

or their incomes increased after program admission. See the discussion of HUD’s rental assistance programs in appendix C and housing assistance status in

appendix E.

Among suburban VLI renter households, most (78.4 percent) were concentrated in densely populated urban suburbs.

Changes in annual estimates of VLI renter households in nonmetropolitan areas should be viewed with caution because HUD assigns average income limits to

less populated areas to accommodate AHS data suppression. See the discussion of income cutoffs in association with AHS geography in appendix E.

the definition of worst case needs excludes renters with

housing assistance. Examining the spatial distribution

of housing assistance and of worst case needs together

provides information about the extent to which assistance is

mitigating severe housing problems.

Exhibit 2-1 shows the distribution of the nation’s 18.39 million

VLI renter households across the four census regions and

four metropolitan categories in 2019. On a regional basis,

more than one-third of VLI renter households—6.41 million—

lived in the South, 4.31 million lived in the West, 3.95 million

lived in the Northeast, and 3.71 million lived in the Midwest

in 2019. Central cities were home to most (8.94 million) VLI

renter households, followed by suburbs (7.18 million)

and

nonmetropolitan areas (2.28 million).

These geographic

patterns did not change substantially between 2017 and

2019.

Exhibit 2-1. Very Low-Income Renters Experienced Worst Case Needs Across Every Region and

Metropolitan Location in 2019

Metropolitan Location

Region Central Cities Suburbs, Urban Suburbs, Rural Nonmetro-

politan Areas

Total

Northeast (thousands)

2,119 1,342 239 250 3,950

Percentage with worst case needs 37. 3 40.9 33.1 35.5 38.2

Percentage with housing assistance 34.9 30.0 24.8 42.6 33 .1

Midwest (thousands) 1,626 1,075 311 703 3,715

Percentage with worst case needs 39.2 36.9 35.9 28.5 36.2

Percentage with housing assistance 31.3 25.9 19.5 36.7 29.8

South (thousands) 2,817 1,814 761 1,021 6,413

Percentage with worst case needs 44.2 50.0 40.4 33.0 43.6

Percentage with housing assistance 26.8 19.2 22.0 31.4 24.8

West (thousands) 2,374 1,394 240 302 4,310

Percentage with worst case needs 51.9 48.3 39.7 39.3 49.1

Percentage with housing assistance 23.1 24.9 23.2 31.7 24.3

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

(continued)

15WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Metropolitan Location

Region Central Cities Suburbs, Urban Suburbs, Rural Nonmetro-

politan Areas

Total

Total (thousands) 8,936 5,625 1,552 2,276 18,388

Percentage with worst case needs 43.7 44.9 38.3 32.7 42.2

Percentage with housing assistance 28.6 24.5 22.1 34.3 27. 5

Source: HUD PD&R tabulations of American Housing Survey data

Like VLI renter households, worst case needs were common

in every region and metropolitan category across the nation.

As a national average, 42.2 percent of VLI renter households

had worst case needs. The prevalence of worst case needs

among VLI renter households was greater than the national

average in the South and West and in central cities and

urban suburbs. The Midwest, Northeast, rural suburbs, and

nonmetropolitan areas had smaller-than-average shares of

VLI renter households with worst case needs. The national

total of 7.77 million worst case needs in 2019 consisted of

2.80 million households in the South, 2.12 million in the West,

1.51 million in the Northeast, and 1.35 million in the Midwest.

(See appendix exhibit A-10 for additional regional data.)

Exhibit 2-1 also demonstrates the important role that housing

assistance plays in reducing worst case needs. Nationwide,

5.05 million VLI renter households reported receiving housing

assistance in 2019, compared with the 7.77 million having

worst case needs. Thus, 1.5 VLI renter households had worst

case needs for every 1 that was assisted, the same ratio

as in 2017. Put differently, among VLI renter households,

about 28 percent of households had rental assistance, and

an additional 42 percent had worst case needs for assisted

or other affordable housing in 2019. The remaining minority

(30 percent) rented in the private market without housing

assistance and avoided severe housing problems. These

data suggest that in 2019, the private rental market was

working adequately for one in three VLI renter households

(exhibit 2-2).

Source: HUD-PD&R tabulations of American Housing Survey data

11.0%

21.5%

31.4%

19.3%

6.0%

10.8%

10%

15%

20%

25%

30%

35%

40%

45%

Unassisted without

Severe Problems

Assisted Unassisted with

Severe Problems

Percent of very low-income renters

Exhibit 2-2. Housing Problem Status of Very Low-

Income Renter Households by Relative Income

0-30% AMI >30-50% AMI

AMI = area median income.

Source: HUD PD&R tabulations of American Housing Survey data

Nationally, housing assistance is relatively less common

in the suburbs, where less than 25 percent of VLI renter

households were assisted. Newer central cities and

suburbs in the South and West had particularly low rates

of assistance. These regional disparities in the prevalence

of housing assistance for VLI renter households were also

evident nationally, ranging from 24.3 percent in the West to

33.1 percent in the Northeast. Another region with a low rate

of housing assistance, the West, has had the highest rate

of worst case needs for decades. Nearly one-half—49.1

percent—of VLI renter households in the West experienced

worst case needs in 2019. Areas that developed during an

earlier period continue to draw benefits from an established

but aging stock of public housing.

Exhibit 2-3 illustrates the vital role of housing assistance in

preventing households from falling into worst case needs.

In exhibit 2-3, central cities, suburbs, and nonmetropolitan

areas are represented by blue, purple, and green bubbles,

respectively. Larger bubbles represent a larger national

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

Exhibit 2-1. Very Low-Income Renters Experienced Worst Case Needs Across Every Region and Metropolitan Location

in 2019 (continued)

16 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

share of worst case needs households. Across regions

and metropolitan locations, housing assistance is inversely

related to worst case needs. Locations indicated in the

upper-left quadrant of the chart fared better than the national

average because of higher rates of housing assistance and

lower prevalence of worst case needs among VLI renter

households. The locations clustered in the middle of the

chart approximate average prevalence rates; the locations

in the lower-right quadrant fared worse than the national

average because of lower rates of housing assistance and

higher rates of severe housing problems.

Patterns in the suburbs tended to be worse than those in

nonmetropolitan areas nationally, whereas central cities

vary. Worst case needs affected a smaller share of very

low-income renters in nonmetropolitan areas, where housing

assistance was relatively more available. Central cities of the

Northeast and Midwest also fared better—with higher rates

of housing assistance and lower rates of worst case needs—

than their counterparts in the South and West.

Source: HUD-PD&R tabulations of American Housing Survey data

Exhibit 2-3. Prevalence of Worst Case Needs Was Inversely

Related to Prevalence of Housing Assistance in 2019

Percent of very low-income renters

with housing assistance

15%

40%

45%

35%

25%

30%

20%

25% 30% 35% 40% 45% 50% 55%

Percent of very low-income renters with worst case needs

Size of bubble is

proportional to share of

national worst case needs

(range 1% – 16%)

Midwest central city

Midwest

suburb, urban

Midwest

suburb,

rural

Midwest

nonmetro

Northeast central city

Northeast

suburb, urban

Northeast

suburb, rural

Northeast nonmetro

South central city

South

suburb, urban

South

suburb, rural

South

nonmetro

West

central city

West suburb, urban

West suburb,

rural

West

nonmetro

Worst case needs were more prevalent in the West and the

South, especially in the suburbs, where housing assistance

was scarcer—although high rents in the West also shape

this picture.

Several areas with a greater relative scarcity of

housing assistance and an abundance of worst case needs

account for substantial fractions of the national problem, as

shown by the size of the bubbles in the lower-right quadrant

of exhibit 2-3. The relative size and positioning of the bubbles

for central cities and urban suburbs also suggest that

High rents introduce the question of whether enough rental units are available at fair market rents (FMRs) to make housing vouchers an adequate policy

response to affordable housing shortfalls. Appendix B, exhibit B-3 addresses the extent of the supply of below-FMR housing on a regional basis. Also see

regional supply discussions later in this section.

denser urban areas contribute the largest shares of severe

housing affordability problems. Together, southern and

western-central cities and urban suburbs accounted for a

substantial share of the national picture in 2019, representing

52 percent of households with worst case needs nationally.

In recent decades, housing policy has not kept pace with

geographic shifts in the national population distribution and

housing needs. Policy enhancements to improve geographic

allocation of housing resources could reduce such spatial

disparities and their impacts on community well-being.

Compared with their urban counterparts, the small

populations of very low-income renters living in rural suburbs

represented a small share of worst-case households. Rural

suburbs of the West, however, do have low rates of housing

assistance coinciding with high rates of worst case needs.

Correspondingly, many Western rural suburbs experienced

high population rates of homelessness in 2017 (Nisar et al.,

2019).

Not shown in exhibit 2-3 are changes in rates of VLI renter

households with worst case needs between 2017 and 2019.

Slight improvements ranging from less than 1 percentage

point to 1.3 percentage points were observed in all regions

except for the West, where the prevalence rate worsened

slightly—by less than 1 percentage point (summarized

in exhibit A-10). During the same period, rates of worst

case needs increased in central cities by 1.3 percentage

points, with decreases of 6.7, 1.9, and 1.0 percentage

points observed in rural suburbs, urban suburbs, and

nonmetropolitan areas, respectively (summarized in exhibit

A-11A). The number of VLI renter households living in the

suburbs grew by 338,000 households in urban suburbs and

by 192,000 in rural suburbs, compared to a reduction of

149,000 households in nonmetropolitan areas and 59,000

fewer households in central cities. This variation may reflect

a combination of a slower rate of economic improvement in

nonsuburban areas during the 2-year period and some out-

migration of lower-income households from high-cost central

cities.

Variation in Worst Case Needs

Between Metropolitan Markets

An examination of VLI renter households’ distribution

and prevalence of worst case needs across the largest

metropolitan areas offers additional insight into the variation

of severe housing problems in central cities and suburbs.

With their densely populated urban cores connected to

surrounding counties through strong commuting ties,

metropolitan areas reflect groupings of central cities

and suburbs with a high degree of social and economic

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

17WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

integration. The redesigned AHS supports examining the

variation in worst case needs across some of the largest

metropolitan housing markets. Exhibit 2-4 shows the VLI

renter populations and the number and share experiencing

worst case needs in the nation’s 15 largest metropolitan

areas in 2017 and 2019.

Exhibit 2-4. Prevalence of Worst Case Needs Among Very Low-Income Renters Varied Across Metropolitan

Markets in 2019

Metropolitan Area 2017 2019 Metropolitan Area 2017 2019

New York-Newark-Jersey City, NY-NJ-PA Washington-Arlington-Alexandria, DC-VA-MD-WV

Very low-income renters (thousands) 1,712 1,769 Very low-income renters (thousands) 292 333

Worst case needs (thousands) 678 724 Worst case needs (thousands) 126 123

Percent with worst case needs 39.6 40.9 Percent with worst case needs 43.2 36.9

Los Angeles-Long Beach-Anaheim, CA San Francisco-Oakland-Hayward, CA

Very low-income renters (thousands) 968 976 Very low-income renters (thousands) 274 291

Worst case needs (thousands) 459 508 Worst case needs (thousands) 110 125

Percent with worst case needs 47. 4 52.0 Percent with worst case needs 40.1 43.0

Chicago-Naperville-Elgin, IL-IN-WI Atlanta-Sandy Springs-Roswell, GA

Very low-income renters (thousands) 509 476 Very low-income renters (thousands) 245 263

Worst case needs (thousands) 204 160 Worst case needs (thousands) 131 132

Percent with worst case needs 40.1 33.6 Percent with worst case needs 53.5 50.2

Dallas-Fort Worth-Arlington, TX Detroit-Warren-Dearborn, MI

Very low-income renters (thousands) 332 364 Very low-income renters (thousands) 243 235

Worst case needs (thousands) 159 174 Worst case needs (thousands) 105 111

Percent with worst case needs 47. 9 47.8 Percent with worst case needs 43.2 47.2

Miami-Fort Lauderdale-West Palm Beach, FL Seattle-Tacoma-Bellevue, WA

Very low-income renters (thousands) 384 356 Very low-income renters (thousands) 202 193

Worst case needs (thousands) 211 177 Worst case needs (thousands) 84 81

Percent with worst case needs 54.9 49.7 Percent with worst case needs 41.6 42.0

Boston-Cambridge-Newton, MA-NH Phoenix-Mesa-Scottsdale, AZ

Very low-income renters (thousands) 313 347 Very low-income renters (thousands) 189 185

Worst case needs (thousands) 100 99 Worst case needs (thousands) 97 99

Percent with worst case needs 31.9 28.5 Percent with worst case needs 51.3 53.5

Houston-The Woodlands-Sugar Land, TX Riverside-San Bernardino-Ontario, CA

Very low-income renters (thousands) 362 337 Very low-income renters (thousands) 158 149

Worst case needs (thousands) 177 179 Worst case needs (thousands) 91 98

Percent with worst case needs 48.9 5 3.1 Percent with worst case needs 57.6 65.8

Philadelphia-Camden-Wilmington, PA-NJ-DE-MD National

Very low-income renters (thousands) 336 306 Very low-income renters (thousands) 18,067 18,388

Worst case needs (thousands) 147 125 Worst case needs (thousands) 7,716 7,76 6

Percent with worst case needs 43.8 40.8 Percent with worst case needs 42.7 42.2

Notes: Estimates for the 15 largest metropolitan areas (by population ranking) are presented. The redesigned AHS samples these 15 metropolitan areas every 2

years. Estimates for 10 additional metropolitan areas surveyed in 2019 are presented in exhibit A-11B.

Source: HUD PD&R tabulations of American Housing Survey data

Although 42.2 percent of VLI renter households had worst

case needs nationally, local markets show a substantial

degree of variation beyond the macro-level trends observed

across regions and types of metropolitan locations. Worst

case needs affected substantial shares of VLI renter

households in each of the nation’s largest metropolitan

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

18 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

areas. Among the 15 metropolitan areas shown in exhibit 2-4,

44.3

percent of VLI renters had worst case needs in 2019,

compared with 44.2 percent in 2017, representing an

increase of 36,000 cases. Reflecting particularly severe local

conditions, more than one-half of the VLI renter households

residing in and around Riverside, Phoenix, Houston, Atlanta,

and Los Angeles experienced worst case needs in 2019.

The rates of worst case needs decreased in nearly one-half

of the large metropolitan areas between 2017 and 2019,

with Chicago, Washington, D.C., and Miami having the

largest decreases, a difference of 6.5, 6.2, and 5.2 points,

respectively. Local events, trends, and policies may account

for changing rates of housing problems within metropolitan

areas.

Factors Limiting Access to

Affordable Rental Housing

Even with slightly more than one-fourth of VLI renter

households receiving housing assistance, the private

market’s affordable rental housing supply falls far short

of need. Nationally, less than one-third of VLI renter

households could avoid severe housing problems in the

The method of assigning units to cost categories was modified in 2017 to also account for limited HUD administrative exceptions to program income limits.

Slight unit affordability adjustments were applied to outliner cases where AMI-determined affordability differed from administratively determined affordability

categories.

Note that renters whose incomes place them at the bottom of an income range would not be able to afford rents at the top of their range. More detailed

presentations of these data appear in appendixes A and B, where exhibit A-12 and exhibit B-2 show unit affordability and occupancy status using 10-point

income breaks.

unassisted private rental market in 2019. An examination of

the mismatches between the number of rental units needed

by renters of various income categories and the number of

affordable units provided by the market to those renters lends

considerable insight into private rental market dynamics

and the persistence of worst case needs during periods of

economic growth.

How the Market Allocates Affordable

Housing on a National Basis

The competition for good-quality, affordable housing

remains vigorous. Competition affects whether the neediest

households can live in the most affordable units, the vacancy

rate at different rent levels, and how quickly new units are

occupied. Exhibit 2-5 shows the distribution of rental units

and their occupancy by their rents’ affordability relative to

the area median income (AMI).

For this analysis, a unit is

considered affordable for a renter if the gross rent (rent plus

utilities) does not exceed 30 percent of the maximum income

of their income category. However, any given renter may live

in a unit renting for less than, the same as, or more than that

threshold.

Exhibit 2-5. Higher-Income Renters Occupied Many Affordable Units in 2019

Rental Units by Income Needed To Make the Rent Affordable (thousands)

Occupancy Status 0–30% of AMI >30–50% of AMI >50–80% of AMI >80% of AMI Total

Higher-income occupants 3,524 3,788 7,0 91 NA 14,403

Same-income or lower-income occupants 4,406 4,724 10,192 10,935 30,257

Vacant 326 881 1,830 1,638 4,675

Total 8,256 9,393 19,113 12,573 49,335

AMI = area median income. NA = not applicable.

Source: HUD PD&R tabulations of American Housing Survey data

The extent of competition for the most affordable housing

portrayed in exhibit 2-5 is striking. Higher-income renters

occupy 3.52 million, or 42.7 percent, of the units affordable

to ELI renter households. Similarly, higher-income renters

occupy 40.3 percent of units affordable at incomes of 30

to 50 percent of AMI and 37.1 percent of units affordable

at incomes of 50 to 80 percent of AMI, which is the largest

category of units. Rental units that are more affordable are

both rarer and more likely to be occupied by higher-income

renters.

Variations in vacancy rates across the affordability categories

further demonstrate the competition for affordable units.

The most affordable units are least likely to be vacant

(exhibit 2-6). Among the least costly units—those with rents

affordable at incomes of 0 to 30 percent of AMI—only 3.9

percent were vacant in 2019. Vacancy rates were much

greater at higher rent levels: 9.4 percent among units

affordable at incomes of 30 to 50 percent of AMI, 9.6 percent

at 50 to 80 percent of AMI, and 13.0 percent among the

highest rent units. Among higher rent levels, vacancy rates

have risen by 3.8 percentage points for 90 to 100 percent

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

19WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

of AMI, by 2.2 percentage points for 100 to 110 percent

of AMI, and by 1 percentage point for over 110 percent of

AMI, signaling construction of new luxury rental housing

since 2015. Overall, rental vacancy rates were consistently

below 10 percent in recent years—9.7 percent in 2015, 9.9

percent in 2017, and 9.5 percent in 2019—reflecting steady

absorption of unoccupied rental housing stock.

Affordability category

(percent of Area Median Income needed to afford the highest

rent in the category)

10%

15%

100

110

>110

Percent of rental units that are vacant

Vacancy rate, 2019

Vacancy rate, 2017

Vacancy rate, 2015

Exhibit 2-6. Vacancies Were Lowest Among Most

Affordable Rental Units, 2015–2019

Source: HUD PD&R tabulations of American Housing Survey data

The gradient in national vacancy rates seen in exhibit 2-6

remained relatively flat among units affordable to low-

income renters earning between 50 and 80 percent of AMI.

Nevertheless, the market for units affordable at ELI levels

remained very tight. The somewhat higher vacancy rate

for the units affordable at only 10 percent of AMI is often

ascribed to units with physical or locational challenges that

soon might be removed from the housing stock. Higher

vacancy rates continue to be found at the highest rent levels,

including numerous vacation homes,

and may reflect

developer preferences to construct higher-end rental units

in recent years. Regulatory barriers that make affordable

homebuilding difficult have exacerbated labor shortages

that constrain mid-range rental housing production needed

to cope with large tenure shifts and household formation. In

many areas, the production of housing for ELI renters is not

profitable.

Comparable estimates of the rental vacancy rate based on the Current Population Survey are slightly lower in recent years: 8.3 percent in 2013, 7.1 percent in

2015, 7.2 percent in 2017, and 6.8 percent in 2019. See U.S. Housing Market Conditions charting data, https://www.huduser.gov/portal/ushmc/hi_RentVac.html.

According to 2019 American Community Survey 1-Year Estimates, about one-third of vacant housing units in the United States are for seasonal, recreational, or

occasional use.

Regional variation in the availability of units within fair market rent (FMR) limits is further addressed in the “Rental Stock by Region” discussion later in this

section.

See exhibits A-12 and A-13.

Compared with the market for the most affordable units, the

availability of vacant units at higher rent levels shows that

in many markets, rental assistance in the form of vouchers

could reduce worst case needs to the extent that rents fall

within program limits and landlords are willing to participate.

The appendix exhibit B-3, which examines the availability

of units within HUD program rent limits (including all HUD-

assisted housing), shows that in 2019, about 75 affordable

and physically adequate rental units were available for every

100 households nationally.

Increasing landlord participation

in HUD’s voucher program could improve access to those

units among very low-income households while also

improving voucher utilization rates in places where vouchers

are available but difficult to lease up.

From 2017 to 2019, the rental stock grew by 515,000

units, or a little over 1 percent, yet the number of vacant

units decreased by 151,000, or 3.1 percent. Despite small

increases in vacant units while the overall rental stock

expanded in past years, strong rental demand nationwide

kept vacancy rates fairly constrained for renters with median

or lower incomes. The rental stock affordable to VLI renters

expanded by 1.3 million units, or 7.7 percent, between 2017

and 2019, whereas affordable vacant units declined 1.3

percent.

Although vacancy rates provide a valuable indication of

the balance between supply and demand, they do not

directly compare the number of affordable units with the

number of renters. The remainder of section 2 makes

such comparisons, employing three increasingly rigorous

concepts to assess the sufficiency of the rental housing stock

relative to need.

Affordability, Availability, and

Adequacy of the National Rental

Stock

The scarcity of affordable units is typically greatest for

the poorest renters, but because of the rapid increase in

renter households and greater competition since the Great

Recession, scarcity has reached higher up the income

scale. Although the renter population expansion slowed

somewhat in 2019 and slightly more renter households had

very low incomes, rental units largely remained out of reach

for households remaining at the lower end of local income

distributions. Exhibit 2-7 displays the rental housing stock

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

20 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

in 2019. These aggregate data portray how well the overall

stock could meet the need for affordable housing if location

did not matter.

Source: HUD PD&R tabulations of American Housing Survey data

Exhibit 2-7. The Supply of Affordable,

Available, and Adequate Rental Housing

Stock Was Insufficient in 2019

100

120

140

160

0 10 20 30 40 50 60 70 80 90 100 110 120

Rental units per 100 renters

Income as percent of Area Median Income

Affordable

Affordable & available

Affordable, available,

& adequate

The top (teal) line in exhibit 2-7 represents all affordable

units in 2019, regardless of whether higher-income

households occupied them or whether they were adequate.

The cumulative number of affordable units equaled the

cumulative number of renters (that is, 100 units per 100

renters) only for incomes exceeding 52 percent of AMI.

Beyond that threshold, more than 100 affordable units existed

per 100 renters—enough, with perfect allocation, to provide

affordable housing to every renter with a higher income. The

2019 threshold was 1 percentage point lower than the 2017

level but 7 points higher than the 2007 level, indicating that

many households recovering from the recession remained

renters for longer periods as the economy recovered.

The ratio of affordable units per renter peaked at 138 units

per 100 renters at the income level of 95 percent of AMI.

There was a substantial surplus of units affordable at higher

levels of household income on a cumulative basis. As income

increased, renters were increasingly likely to spend less than

30 percent of their incomes on housing.

The situation was completely different at the low end of

the income scale. Enough affordable units—both naturally

Measures of affordability, availability, and adequacy compare the entire housing stock with the entire renter population, and they do not reflect small-scale

geographic detail or the complexities of local housing markets.

Only 13.7 percent of renters with incomes above 80 percent of AMI had either moderate or severe rent burdens, compared with 68.6 percent of renters with

lower incomes. See exhibit A-1A.

The availability measure also removes units from consideration if they have artificially low rents because they are occupied as a benefit of employment (for

example, units provided for caretakers) or because relatives or friends of the occupants own the units. In 2019, 1.92 million renter households (4.3 percent)

occupied their units while paying no rent. The AHS does not provide estimates of the number of households paying a positive but less-than-market rent because

of employment or other reasons.

affordable and assisted—existed to house only 70 percent

of ELI renters in 2019, assuming those units somehow could

have been perfectly allocated. That shortage was substantial

and critical, with little improvement from the ratio of 69

percent observed in 2017.

MEASURING WHETHER AFFORDABLE

HOUSING STOCK IS SUFFICIENT FOR NEED

• Affordability measures the extent to which

enough rental housing units of different costs

can provide each renter household with a unit it

can afford (based on the 30-percent-of-income

standard). Affordability, which is the broadest

measure of the relative supply of the housing

stock, addresses whether sufcient housing

units would exist if allocated solely on the basis

of cost. The affordable stock includes both

vacant and occupied units.

• Availability measures the extent to which

affordable rental housing units are available

to renters within a particular income range.

Availability is a more restrictive concept because

units that meet the denition must be available

and affordable. Some renters choose to spend

less than 30 percent of their incomes on rent,

occupying housing that is affordable to renters of

lower incomes. Those units thus are not available

to lower-income renters. A unit is available at a

given level of income if (1) it is affordable at that

level, and (2) it is occupied by a renter either at

that income level or at a lower level or is vacant.

• Adequacy extends the concept of availability

by considering whether sufcient rental

units are physically adequate (based on unit

characteristics described in appendix E),

affordable, and available. Adequacy thus is the

most restrictive of the three measures.

The second (lavender) line in exhibit 2-7 represents all

affordable and available rental units in 2019, meaning that it

considers whether higher-income renters currently occupy

affordable units.

Availability poses an important additional

constraint on renters seeking affordable units. The exhibit

shows that, at best, only 40 percent of ELI renters could find

an affordable and available unit, even if location were not a

factor.

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

21WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

The paucity of affordable and available units is worsened

by the occupancy of a considerable proportion of the most

affordable housing stock by renters who could afford to

spend more but do not choose to do so (as shown previously

in exhibit 2-5). Such renters may be cautious about their

finances because of income instability, a desire to reduce

debt burdens, or saving for a downpayment to buy a house.

In 2019, the affordable stock was nominally sufficient to

house every renter with an income greater than 52 percent of

AMI, yet the affordable and available stock did not match the

number of renters until household incomes reached about 85

percent of AMI.

The third (plum) line in exhibit 2-7 adds a third criterion—that

units should be physically adequate—which further reduces

the supply of the rental housing stock. Even for renters with

low incomes (up to 80 percent of AMI), only 87 adequate

units were available for every 100 renters. The physically

adequate stock did not fully match the need until it included

units affordable only to renters with incomes exceeding 124

percent of AMI, similar to 2017 levels.

Rental Stock by Income

Thus far, the analysis has shown that relatively few rental

units were affordable in 2019, and—because of occupancy

by higher-income renters and limited vacancies—even fewer

were available to renters with the lowest incomes. Exhibit 2-8

summarizes the availability of rental units for the standard

income groups used in this report.

A severe mismatch existed between the number of ELI renter

households and the number of affordable units available. For

every 100 ELI renter households, only 70 affordable units

existed. Only 40 of those units were affordable and available,

and only 36 were affordable, available, and physically

adequate.

About 11 percent of affordable and available

units for ELI renters had severe quality deficiencies.

Previous research based on the Residential Finance Survey indicated that 12 percent of units with gross rents of $400 or less produced negative net operating

income, suggesting they were headed for demolition or conversion to nonresidential use (JCHS, 2006). More recent research based on the Housing Vacancy

Survey suggests that more than 10 percent of vacant units held off-market are in need of repair, abandoned, condemned, or to be demolished (JCHS, 2016).

Exhibit 2-8. Rental Housing Stock Was Scarcest for

Extremely Low-Income Renters in 2019

Income Category Affordable

Rental

Units per

100 Renter

Households

Affordable

and Available

Rental

Units per

100 Renter

Households

Affordable,

Available,

and Adequate

Rental Units

per 100 Renter

Households

Extremely low-

income renter

households (0–30%

AMI)

70.3 40.3 35.7

Very low-income

renter households

(0–50% AMI)

96.0 62.2 54.7

Low-income renter

households

(0–80% AMI)

135.3 97. 3 87. 2

AMI = area median income.

Source: HUD-PD&R tabulations of American Housing Survey data

Renters with very low incomes found 96 affordable units,

62 affordable and available units, and only 55 affordable,

available, and physically adequate units per 100 renters.

About 12 percent of the affordable and available units for this

larger group had severe quality deficiencies.

Renters with low incomes found that the affordable and

available rental stock was nearly sufficient to house them all,

although about 10 percent of such units had severe quality

deficiencies.

Overall, affordable housing supply improved slightly for

extremely low-income renters between 2017 and 2019, an

expansion of 273,000 units. Exhibit 2-9 shows that the supply

of affordable housing stock for ELI renters increased by 1 unit

per 100 households, from 69 in 2017 to 70 in 2019, following

a three-unit gain during the previous 2 years. The ratio of

affordable and available units stayed at about 40 units per

100 ELI renter households between 2017 and 2019.

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

22 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit 2-9. Fewer Affordable Units Were Available to Very Low-Income Renters in 2019

Income Category 2015

Rental Units per

100 Renters

2017

Rental Units per

100 Renters

2019

Rental Units per

100 Renters

2015 to 2017

Change

2017 to 2019

Change

Extremely low-income renter households (0–30% AMI)

Affordable

66.0 69.1 70.3 3.1 1.1

Affordable and available

37.7 39.8 40.3 2.1 0.5

Very low-income renter households (0–50% AMI)

Affordable

92.9 90.7 96.0 – 2.2 5.3

Affordable and available 62.0 59.0 62.2 – 3.0 3.2

AMI = area median income.

Source: HUD-PD&R tabulations of American Housing Survey data

For renters who could afford rents no greater than the FMR, appendix B, exhibit B-3 reveals that although enough affordable units existed in each region, the

number of available units in each region was sufficient to house only 76 to 85 percent of those renters. For renters who attempt to find a unit with a housing

choice voucher, the housing quality standards of that program imply that their success will depend on the prevalence of physically adequate units in their area—

not merely affordable and available units. Across regions, there were only enough physically adequate, affordable, and available units to house 71 to 78 percent

of renters who could not afford rents higher than FMR.

For very low-income renters, the housing supply grew

more substantially between 2017 and 2019. The supply

of affordable units for VLI households increased by more

than 5 units per 100 renters, and the supply of affordable

and available units increased by more than 3 units per 100

renters. Comparing the negligible gain in unit availability for

ELI renters to the more than 5-unit expansion for VLI renters

suggests that renters with incomes between 30 and 50

percent of AMI benefited the most from the expanded stock

during the 2017-to-2019 period.

Geography of Supply

The preceding discussion shows that worst case needs

in 2019 were dispersed across the nation, although their

concentration varied across geography. It further shows that

spatial variation in worst case needs was affected in part by

the availability and utilization of housing assistance.

Affordable rental housing includes both units that receive

public rent assistance and units that for-profit and nonprofit

housing providers offer at modest rents. The examination of

affordable housing supply on a national basis revealed that

the supply of rental units that are affordable to very low-

income and poorer households remained deeply insufficient

in 2019 and that this shortage was worsened by the

preference of higher-income renters for more affordable units

and by the physical inadequacy of some of the stock.

The following discussion sharpens that picture by showing

how shortages vary by geography.

Rental Stock by Region

Rental markets are constrained for ELI renters across the four

census regions despite substantial variation in the availability

of affordable rental units.

Exhibit 2-10 illustrates that the

Midwest had the best availability in 2019, with 80 units per

100 VLI renter households. The West was worst off, with

fewer than 44 units per 100 VLI renter households, compared

with 64 units for the South and 63 for the Northeast. For

ELI renters, the availability of affordable units was far from

adequate in any region. Even low-income renters with

incomes up to 80 percent of AMI found that not enough

affordable units were available in the West and Northeast.

Exhibit 2-10. Rental Housing Stock Was Insufcient for Extremely Low-Income Renters Across All Regions in 2019

Region and

Renter Income Category

Affordable Housing

Units per 100 Renters

Affordable and

Available

Housing Units per 100

Renters

Affordable, Available,

and Adequate

Housing Units per 100

Renters

Northeast

Extremely low-income (0–30% AMI) 75.5 46.1 40.9

Very low-income (0–50% AMI) 91.9 62.9 56.2

Low-income (0–80% AMI) 125.8 92.8 82.4

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

(continued)

23WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Region and

Renter Income Category

Affordable Housing

Units per 100 Renters

Affordable and

Available

Housing Units per 100

Renters

Affordable, Available,

and Adequate

Housing Units per 100

Renters

Midwest

Extremely low-income (0–30% AMI) 77. 9 44.2 40.0

Very low-income (0–50% AMI) 128.7 79.7 71.4

Low-income (0–80% AMI) 15 7.0 110.6 100.6

South

Extremely low-income (0–30% AMI) 73.0 39.9 34.9

Very low-income (0–50% AMI) 99.0 63.6 54.6

Low-income (0–80% AMI) 141.1 101.2 89.8

West

Extremely low-income (0–30% AMI) 54.2 31.9 28.3

Very low-income (0–50% AMI) 66.9 44.3 38.9

Low-income (0–80% AMI) 115.9 83.8 75.9

AMI = area median income.

Source: HUD-PD&R tabulations of American Housing Survey data

On a regional basis, adding the adequacy test restricted the

estimated supply for VLI renters less in the West, eliminating

5 units, than in the other regions, which lost 7 to 9 units

per 100 VLI renter households. Even so, the West retains

its regional disadvantage for such renters across all three

measures of sufficiency.

The primary point in exhibit 2-10 is that ELI renter households

continued to face severely constrained markets across

all four regions. The Northeast, Midwest, and South had

affordable units available for only two in five ELI renter

households, and the West for only one in three.

Rental Stock by Metropolitan

Location

Similar analysis of affordable housing supply based on

metropolitan status showed market variation in 2019.

Exhibit 2-11 demonstrates the primacy of urban areas

in terms of severe shortages of affordable units for VLI

renter households. As shown in exhibit 2-11, measures of

affordability, availability, and adequacy for each income

group in central cities and urban suburbs were generally

lower than the national summary values presented in exhibit

2-8.

Exhibit 2-11. Rental Housing Stock Was Insufcient for Extremely Low-Income Renters Across All Metropolitan

Locations in 2019

Metropolitan Location and

Income Category

Affordable Housing

Units per 100 Renter

Households

Affordable and

Available

Housing Units per 100

Renter Households

Affordable, Available,

and Adequate

Housing Units per 100

Renter Households

Central cities

Extremely low income renters (0–30% AMI) 58.8 37. 2 32.6

Very low income renters (0–50% AMI) 86.4 59.8 52.2

Low-income renters (0–80% AMI) 128.3 95.4 85.3

Suburbs, urban

Extremely low income renters (0–30% AMI) 63.7 36.4 33.0

Very low income renters (0–50% AMI) 83.4 52.9 47.8

Low-income renters (0–80% AMI) 134.4 9 3.1 85.1

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

Exhibit 2-10. Rental Housing Stock Was Insufcient for Extremely Low-Income Renters Across All Regions in 2019

(continued)

24 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Metropolitan Location and

Income Category

Affordable Housing

Units per 100 Renter

Households

Affordable and

Available

Housing Units per 100

Renter Households

Affordable, Available,

and Adequate

Housing Units per 100

Renter Households

Suburbs, rural

Extremely low income renters (0–30% AMI) 102.2 45.5 40.8

Very low income renters (0–50% AMI) 134.7 80.6 70.6

Low-income renters (0–80% AMI) 143.9 104.8 92.9

Nonmetropolitan areas

Extremely low income renters (0–30% AMI) 108.0 56.7 49.6

Very low income renters (0–50% AMI) 138.4 81.6 70.3

Low-income renters (0–80% AMI) 158.7 110.5 96.0

AMI = area median income.

Source: HUD-PD&R tabulations of American Housing Survey data

That is, 26 of the 86 units affordable for every 100 VLI renter households in central cities are not available; the same is true for 30 of 83 affordable units in urban

suburbs.

Likewise, Divringi et al. (2019) found that aggregate repair costs were particularly high among single-family rental units, especially older units occupied by

renters with incomes at or below the poverty line. Repair needs among those units accounted for about 20 percent of the aggregate estimated repair costs of all

renter households in 2018. Those units are disproportionately clustered in nonmetropolitan areas.

Stock in rural suburbs and nonmetropolitan areas was not

as constrained as the stock for the nation as a whole. Exhibit

2-11 also highlights severe deficiencies in the availability

and adequacy of affordable units in rural areas. Among

affordable units to VLI renter households in urban areas,

31 to 36 percent are occupied by higher-income renters.

In rural areas, that figure ranges from 40 to 41 percent,

suggesting that higher-income renters consume a larger

share of the affordable housing stock in rural areas than

those who live closer to city centers. This evidence disrupts

the notion that the affordable housing crisis could be

resolved simply by lower-income renters moving away from

cities, and represents a mobility barrier for people who want

to move for job opportunities or other reasons. Likewise, a

greater share of units had severe quality deficiencies in rural

areas, where 12 to 14 percent of affordable units available to

very low-income renters are inadequate.

These problems

are less prevalent in urban areas—affecting 10 to 13 percent

of units affordable and available to very low-income renters.

Summary

Worst case needs are commonplace in every region and

metropolitan category across the United States. The national

total of 7.77 million worst case needs in 2019 is distributed

on a regional basis, with 2.79 million households in the

South, followed by 2.12 million in the West, 1.51 million in the

Northeast, and 1.35 million in the Midwest. Nationwide, 42.2

percent of very low-income renters had worst case needs

in 2019, a rate higher than in 2017. Prevalence increased in

the West but decreased modestly in the other three regions

since 2017

. Both the South and West maintained greater-

than-average rates of worst case needs in 2019. Urban

areas (urban suburbs and central cities) also had greater-

than-average prevalence rates and were home to about 83

percent of worst case needs households.

Housing assistance, including HUD-provided assistance, is

an important preventer of worst case needs among very low-

income renters. Nationwide, 27.5 percent of very low-income

renters, or 5.05 million households, reported receiving

housing assistance. For every VLI renter household assisted,

another 1.5 renter households had worst case needs that

could have been mitigated with such assistance.

Steady absorption of unoccupied rental housing stock has

reduced overall vacancy rates to consistently less than

10 percent since 2011. With 96 rental units affordable for

every 100 VLI renter households nationally, not all such

households could find an affordable unit in 2019, even if

allocations were perfect among households across the nation

(that is, if the lowest rent units were allocated to the lowest

income households first). Many fewer affordable units were

actually available to renters with the lowest incomes because

vacancy rates were lowest for the lowest rent units, and

many affordable units were rented to higher-income families.

In 2019, the vacancy rate was only 3.9 percent for units

affordable at extremely low incomes, compared with 13.0

percent for units affordable at more than 80 percent of AMI.

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

Exhibit 2-11. Rental Housing Stock Was Insufcient for Extremely Low-Income Renters Across All Metropolitan

Locations in 2019 (continued)

25WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

The slight expansion of rental stock to meet rental demand

between 2017 and 2019 mostly benefited higher-income

households, with fewer new units affordable to VLI renter

households.

Because of competition for affordable units, when a simple

ratio of affordable units per 100 VLI renter households is

made more stringent by adding availability as a constraint,

the ratio decreases from 96 affordable units to only 62

affordable and available units per 100 VLI renter households,

and it decreases from 70 to 40 per 100 ELI renter

households. Higher-income families occupied 42.7 percent of

units affordable to ELI renter households.

In addition, a substantial proportion of available units

are not in adequate physical condition. The number of

affordable, available, and adequate units in 2019 was 55 per

100 VLI renter households and only 36 per 100 ELI renter

households.

Given the scarcity of affordable, available, and adequate

units for the renters with the lowest incomes, the efficacy

of housing assistance in preventing worst case needs, and

the surplus of units available at higher rent levels, housing

vouchers continue to offer an important policy option for

addressing the growing problem of worst case needs

using the existing housing stock. Provided that physically

adequate units with rents within program limits are available

on the market, vouchers could reduce worst case needs to

the extent that landlords are willing to participate in HUD’s

voucher program. Increasing landlord participation could

improve access to those units among VLI households

while also improving voucher utilization rates in places

where vouchers are available but difficult to lease up. HUD

continues to reach out to landlords and conduct program

demonstrations to test incentives for greater landlord

participation in HUD’s primary rent subsidy program with

the aim of making voucher-eligible units more accessible,

especially in higher-opportunity neighborhoods.

See https://www.huduser.gov/portal/periodicals/em/winter19/highlight1.html.

SECTION 2. SHORTAGE OF AFFORDABLE HOUSING

26 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

27WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Understanding

the Trend in Worst Case

Needs

Section 2 demonstrated that worst case needs are prevalent across the nation

because of the limited availability of adequate, affordable rental units relative to

the number of very low-income (VLI) renter households who need them. Section

3 elaborates how the changes in the number of rental units, the number of renter

households, and rental costs during the 2017-to-2019 period contributed to the

prevalence of worst case needs.

In 2019, worst case needs had increased by an estimated 50,000 cases from

2017 levels. The analysis in this section attributes the marginal increase in

worst case needs to the ongoing formation of new households while housing

assistance underwent a net decline. What could have been a substantial

increase was mostly offset by easing the competition for affordable units. The

improvement in market conditions experienced by VLI renters fully offset the

role of household formation and helped mitigate other demographic factors that

tended to increase the number of unassisted VLI renter households. Household

incomes among most very low-income renters did not rise enough to exceed the

very low-income threshold and shrink the VLI population. Homeownership rates,

which have been increasing since 2017, have slightly offset household formation

as a force for increasing worst case needs.

Section 3

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

28 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Drivers of Affordable Housing

Demand

The previous sections of this report have shown that the

increase in the number of households with worst case

needs reflects both changes in the population vulnerable

to worst case needs—unassisted VLI renter households—

and changes in the share of those renters experiencing

the severe problems that trigger worst case needs. The

population of vulnerable renters is primarily affected by

demographic factors (including their incomes and, to a

small extent, HUD’s categorization of their incomes). This

population, in turn, substantially determines the demand

for affordable housing. The current rate of worst case

needs among these vulnerable renters, by contrast, reflects

the economic response of the housing market to these

demographic changes.

The following analysis sorts out the factors driving the most

recent change in worst case needs. First, we distinguish

between the effects of population change and the prevalence

of worst case needs to estimate their relative importance.

Then we identify how much various demographic factors

affected the population change.

The population of unassisted VLI renter households

increased 3.6 percent between 2017 and 2019, from 12.88

million to 13.33 million. The rate of worst case needs in this

population slightly decreased from 42.7 to 42.2 percent

during this same period due to income gains among renters

with incomes between 30 and 50 percent of AMI.

Based solely on the demographic increase of unassisted VLI

renters, we might expect to have recorded a net increase of

268,000 cases of worst case needs. That potential increase

was muted, however, by a decrease of 218,000 cases

attributable to changes in the prevalence of severe problems

(the prevalence effect). The prevalence effect reduced the

total potential increase in worst case needs over the 2-year

period by about 81 percent. The combined demographic

and prevalence effects explain the 50,000 additional cases

of worst case needs observed in the AHS in 2019 compared

with the number of cases observed in 2017.

The 268,000 increase in worst case needs resulting from

demographic shifts can be further broken down into four

components, illustrated by the first four columns of exhibit

3-1 and discussed below.

The columns of this chart are

Any analysis of survey data faces limitations from both sampling error and non-sampling error. Such errors are compounded when multiple survey years are

compared. This analysis takes the AHS estimates at face value, but the reader should recognize that multiple sources of potential error exist.

The demographic effect equals the new prevalence rate times the numerical increase (or decrease) in renters, and the prevalence effect is the increase (or

decrease) in the prevalence rate times the baseline number of renters.

Demographic components shown in the chart sum to 269,000 rather than 268,000 because of rounding.

These are AHS estimates. Annual homeownership estimates based on the Current Population Survey/Housing Vacancy Survey were 63.9 percent for 2017,

64.6 percent for 2019, and 66.6 percent for 2020.

cascading in the sense that each column begins where the

previous one ends.

Source: HUD PD&R tabulations of American Housing Survey data

159,000

increase from

population

growth

45,000

decrease from

homeowner

increases

relative to

renter share

20,000 increase

from renter

income changes

adding to VLI

numbers

218,000 decrease

from more

affordable and

available units for

VLI renters

Household

formation

Renter share Renter income

changes

Rental

assistance

gap

Affordable unit

competition

135,000 increase

from reduction in VLI

renters receiving

assistance

Net change:

50,000

increase

Exhibit 3-1. Contributions to Worst Case Needs

from Household Formation and the Rental

Assistance Gap were Offset by Less Competition

for Affordable Units from 2017 to 2019

1. Household formation. The nation added 2.6 million

new households between 2017 and 2019, to which we

attribute a proportional increase of 159,000 cases of

worst case needs. The household formation growth rate

was 2.1 percent during this 2-year period, exceeding the

average biennial increase of 1.9 percent since 2007 as

measured by the AHS.

2. Renter share of households. A decline in renters’

share of households accounts for a reduction of worst

case needs by 45,000 cases, diminishing the effect

of new household formation. The homeownership

rate increased slightly from 63.8 percent in 2017 to

64.0 percent in 2019.

Such growth contrasts sharply

with the 9.1-percent increase in renter households

between 2013 and 2015 and biennial increases in renter

households averaging 4.5 percent since 2007.

3. Renter income changes. Income losses and shifts in

the income distribution affecting the very low-income

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

29WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

category account for a small 20,000-case rise of

worst case needs. The number of renters with very

low incomes increased by 321,000, or 1.8 percent,

in 2019. That biennial increment limits the average

biennial change since 2009 to a 2.6-percent increase.

Simultaneously, the population of renters with higher

incomes grew by 1.3 percent between 2017 and 2019.

In other words, a growing population of higher-income

households was competing with

the population of VLI

renter households for available rental units in 2019.

4. Rental assistance gap. Moderation of the rental

assistance gap accounts for an increase of worst case

needs by 135,000 cases. The number of unassisted

VLI renter households increased by 459,000 during the

2017-to-2019 period; those who reported assistance

decreased by 138,000 even as numbers of VLI renter

households expanded. The 3.6-percent increase in the

number of unassisted VLI renter households between

2017 and 2019 was greater than the average biennial

increases of 2.6 percent over the past decade (2009 to

2019).

This analysis shows that the offsetting demographic

factors that resulted in the net increase in the population of

unassisted very low-income renters would have accounted

for an increase of worst case needs by 268,000 between

2017 and 2019 absent other factors. The housing market,

however, blunted the potential increase in severe housing

problems. The fifth column of exhibit 3-1, labeled “Affordable

unit competition,” represents the extent to which the market

responded to quantitative changes in demand for VLI-

affordable rental units. The column’s size and downward

direction show that a nearly proportionate market response

(218,000 units) offset much of the 268,000 incremental

mismatches between affordable units and unassisted VLI

Methodological factors are summarized in the sidebar, “Changes in Income Limits and Worst Case Needs.”

Higher-income renters have accounted for a growing share of renters in recent years. As a percentage of renter households, VLI renters decreased from a high

of 49.6 percent in 2011 to 41.2 percent in 2019.

See “Unfurnished Rental Apartments–Completed.” https://www.huduser.gov/portal/ushmc/hd_mul_fam.html.

renters who need them. As a result, net additions to worst

case needs between 2017 and 2019 were limited to only

50,000, as unassisted renter households faced a lower

prevalence of severe housing problems—42.2 percent in

2019 compared with 42.7 percent in 2017—because of

modestly improved availability of affordable units. Beneficial

market responses to growing quantitative demand for

affordable units result primarily from the construction of

affordably priced units and reductions in rents (known as

filtering down) of surplus or aging higher-rent units. Indeed,

an average of 286,000 completed, unfurnished rental

apartments were absorbed into the market during 2017 and

2018, a pace not seen since 1988.

The next section further

explores such market factors.

Other Factors Affecting

Affordable Housing Supply

and Demand

Exhibit 2-9 showed that the availability of affordable rental

units slightly eased during the 2017-to-2019 period. Such

affordability metrics are affected by multiple demographic

and market factors. Some additional data, including key

numbers underlying the changes in available unit ratios, will

shed light on the issue.

Exhibit 3-2 examines the factors responsible for the change

in the availability of affordable units. AHS data show an

increase of 515,000 (1.1 percent) in the total number of rental

units between 2017 and 2019 (exhibit A-13). This increase

slightly lagged the 664,000 new renter households that were

added during the same period as the various income groups

expanded at similar rates.

Exhibit 3-2. Changes in Affordable Rental Housing Availability Were Driven by Income Gains Among Renters That

Outpaced Rising Costs, 2017 to 2019

Extremely Low-

Income (0–30%

AMI)

Very Low-Income

(0–50% AMI)

Low-Income

(0–80% AMI)

Total

Cumulative affordable & available rental

units (thousands)

2017 4,595 10,661 26,014 48,820

2019 4,732 11,4 32 26,441 49,335

Percent change +3.0 +7.2 +1. 6 +1.1

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

(continued)

30 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Extremely Low-

Income (0–30%

AMI)

Very Low-Income

(0–50% AMI)

Low-Income

(0–80% AMI)

Total

Cumulative households (thousands)

2017 11,54 8 18,067 26,704 43,993

2019 11,74 8 18,388 27,174 44,660

Percent change +1.7 +1.8 +1. 8 +1. 5

Income limit (median, current dollars)

2017 19,800 29,400 47,000 —

2019 21,330 32,250 51,600 —

Percent change +7.7 +9.7 +9.8 —

Median household income (all renters,

current dollars)

2017 — — — 36,10 0

2019 — — — 40,000

Percent change — — — +10. 8

Median monthly housing cost (all renters,

current dollars)

2017 — — — 991

2019 — — — 1,071

Percent change — — — +8.1

AMI = area median income.

Total represents all units or renters, not the sum of the cumulative income categories.

Source: HUD PD&R tabulations of American Housing Survey data

The median renter household’s income placed it in the VLI category in 2011 but in the low-income category in subsequent years.

See “National Housing Market Summary: 4th Quarter 2019.” https://www.huduser.gov/portal/sites/default/files/pdf/NationalSummary_4Q19.pdf.

In 2019, as in most years, both supply and demand

factors were influential in the extent of worst case needs.

Some of those trends are promising, and others reflect

ongoing challenges. Median renter incomes increased by

10.8 percent between 2017 and 2019, as a strengthening

economy sustained the trend of income gains following the

recession.

Rising incomes among renter households could

translate into increased ability to bid for housing of greater

quality. Higher incomes also provide greater resources to

consume non-housing necessities. Either effect can mitigate

the risk and consequences of severe housing problems—

provided that growing housing costs do not consume income

gains.

Rising renter incomes may occur due to more renters

deferring home purchases. However, home purchase is

typically quite difficult for VLI renters. By the end of 2019,

house prices had fully recovered to pre-Great Recession

levels. There was limited availability of entry-level homes

for purchase compared to homes at higher price points.

Nevertheless, the increase of 771,000 rental units affordable

and available to VLI renters exceeded the increase of

321,000 VLI renter households. As incomes increased and

were perceived as stable during the economic recovery,

higher-income renters were increasingly able to move out

of VLI-affordable units in favor of better-quality units, thus

increasing availability ratios. For ELI renters, however, the

increase of affordable and available units by 137,000 was

mostly offset by an increase of 200,000 renter households in

this group.

Renters overall benefited from modestly improved housing

supply and growing incomes during 2017 to 2019, so that

rents did not increase as much as renter incomes. Yet

housing problems worsened for ELI renters, who faced

disadvantages in rates of both income growth and rent

growth during this period. This disparity is revealed by

examining data from exhibits 3-2 and A-14:

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

Exhibit 3-2. Changes in Affordable Rental Housing Availability Were Driven by Income Gains Among Renters That

Outpaced Rising Costs, 2017 to 2019 (continued)

31WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

- The median change in renter income of 10.8 percent

between 2017 and 2019 was exceeded by the mean

change of 13 percent.

Households with incomes

exceeding 120 percent of AMI experienced gains of

17 percent, but ELI renters experienced gains of only 4

percent—which was less than one-half that of any other

income group.

- Median housing costs for renters increased by 8.1

percent between 2017 and 2019, building on a similar

increase in the prior period.

The mean change

in housing costs was 9.1 percent among all renter

households. For the ELI renter subgroup, however,

housing costs increased by 12 percent during the 2-year

period.

As a result, ELI renters’ housing costs increased almost

three times faster than their incomes from 2017 to 2019. This

growing financial challenge explains why the prevalence

of severe problems among ELI renters increased from 48.1

percent in 2017 to 49.2 percent in 2019, approaching half

of all cases. As a result, ELI renters increased their share of

worst case needs to 74 percent in 2019—a proportion not

seen since 2005.

CHANGES IN INCOME LIMITS AND WORST

CASE NEEDS

A minimal portion of the population change in

renters with extremely low and very low incomes

between 2017 and 2019, and of those with worst

case needs, is explained by a shift in income limits.

HUD calculates income limits on the basis of area

median family incomes, which include both owners

and renters, and then uses the income limits to

dene the boundaries of the extremely low-, very

low-, and low-income categories.

Exhibit 3-2 shows that, across the nation, the

income limits dening each income category

increased roughly in proportion to increases in

AMI between 2017 and 2019. The greatest income

qualifying as extremely low income increased by

$1,530. The greatest income qualifying as very

low income increased by $2,850. As a result of

the higher thresholds, additional households were

captured within the extremely low-income and very

low-income categories in 2019.

Median values are less likely than mean or average values to show the effect of extremes at the end of the distribution. Such extremes, or disparities, affect both

the income and the housing cost distributions.

Those housing costs include rent, utilities, property insurance, land rent, and association fees but exclude any separate security deposit or parking fees.

In addition to experiencing housing problems with growing

frequency, ELI renter households increased in number

from 2017 to 2019. These trends suggest that ELI renter

households may have been isolated from the benefits

of a growing economy—for example, because of fixed

incomes associated with disability or advanced age. For

such households, few paths of escape exist from the severe

shortage of affordable and available housing.

Another critical element to improvement in worst case needs

over time, however, is improving the access VLI renter

households have to an adequate supply of affordable rental

units. Exhibit 3-3 presents how the market for rental units

affordable to VLI households has responded to demand

trends over the past 10 years.

Source: HUD PD&R tabulations of American Housing Survey data

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

20.0

60.0

40.0

100.0

80.0

2009 2011 2013 2015 2017 2019

Worst case needs (thousads)

Units per 100 renter houseolds

Exhibit 3-3. Trends in Housing Supply Mismatch and Worst

Case Needs, 2009–2019

Affordable and available units for ELI renters (per 100 households)

Worst Case Needs (thousands)

Affordable and available units for VLI renters (pre 100 households)

ELI = extremely low-income. VLI = very low-income.

Source: HUD PD&R tabulations of American Housing Survey data

In most instances, increases in worst case needs were

also accompanied by declines in the national supply of

units affordable to VLI renters. Despite a modest increase

in demand between 2017 and 2019, VLI renter households

found more affordable units available. The potential benefits

of this more abundant supply for worst case needs were

constrained by the lack of significant improvement in the

affordable-and-available ratio for ELI renters, who account for

most instances of worst case needs.

In short, the effect of weak growth in the rental supply and

strong competition for available rental units from higher-

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

32 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

income renters is displayed most acutely in the availability

of units affordable to renters with incomes at and below

30 percent of AMI. Although higher-income renters may

be unlikely to compete for the units with the very cheapest

rents because of their quality deficiencies, the competition

for marginally higher tiers of units both reduces availability

directly and causes rents to increase. If renters with incomes

at and below 30 percent of AMI had access to an adequate

supply of affordable rental units during the biennial period,

we might have seen a substantial decrease in worst case

needs nationally. Therefore, supplying a range of rental and

homeownership options to households with both lower and

higher incomes is important to sustaining the downward

trend in worst case needs seen between 2017 and 2019.

Concluding Summary

Amid the strong U.S. economy of 2019, the number of worst

case needs modestly worsened, adding 50,000 cases to the

2017 number to reach 7.76 million. An analysis decomposing

demographic and market factors indicates that the

demographic factors affecting the number of unassisted VLI

renter households had the potential to exacerbate worst case

needs by 268,000—had the market supply of affordable

rental units not slightly improved, causing the prevalence

rate of worst case needs to decrease among renters with

incomes between 30 and 50 percent of AMI.

Of the four demographic factors, the national household

formation would have been expected to increase worst

case needs by 159,000 cases. A shift from renting to

homeownership diminished that effect by an estimated

45,000 cases. Deterioration in renter incomes modestly

increased worst case needs by 20,000 cases. A widening

of the rental assistance gap accounted for an increase of

135,000.

However, the market response to those unfavorable

demographic trends undercut the national increase in worst

case needs. The total supply of rental units increased slightly

more between 2017 and 2019 than during the previous

biennial period, adding 515,000 units, or 1.1 percent. The

total renter population also grew more slowly between 2017

and 2019, moderated somewhat by a slight improvement in

homeownership rates. With vacancy rates highest among

the most expensive units and in some segments continuing

to increase, the number of rental units affordable and

available to VLI renter households increased by 771,000

units (more than 7 percent) as the VLI renter household

population increased by 321,000 households (almost

2 percent). This pattern in supply growth was mimicked in

renter income changes, with median income growing more

sharply for renter households overall (10.8 percent) than

the median income threshold for VLI renter households

(9.7 percent). Further, renters’ housing costs continued

to rapidly escalate—the 8.1-percent increase in median

monthly housing costs for renter households outpaced the

7.5-percent increase in costs during the previous 2 years.

Changes in renters’ incomes and housing costs were not

uniform across the income distribution. Although the mean

change in renter income during 2017 to 2019 was 13 percent,

incomes increased by 17 percent among those with incomes

exceeding 120 percent of AMI, but only by 4 percent among

those with incomes below 30 percent of AMI. Simultaneously,

mean housing costs of renters increased by 9.1 percent as

the lowest-income subgroup, ELI renters, saw 12 percent

increases in housing costs—triple their average income

increase. These dual disparities in income changes and rent

changes explain why the prevalence of worst case needs

among ELI renters increased from 48.1 percent in 2017 to

49.2 percent in 2019.

Worst case housing needs are a national problem with

variations in severity across both demographic and

geographic dimensions. Worst case needs expanded

dramatically as a result of the Great Recession and

associated collapse of the housing market, which reduced

homeownership through foreclosures and increased demand

for renting. During the 10-year economic recovery, 2009 to

2019, worst case needs continued to persist at high levels.

Renter income gains in recent years have been offset by

rent increases because of limited production of affordable

rental units. Even with public rental assistance, 6 of 10 ELI

renter households and 4 of 10 VLI renter households do not

have access to affordable and available housing units. In

2019, there were 1.5 VLI renter households with worst case

needs for every VLI renter household with rental assistance.

Based on the most recent evidence of the 2019 AHS, modest

easing of the shortage of affordable homes offset otherwise

unfavorable demographic trends and helped limit increases

in national levels of severe housing problems as the economy

improved.

Within a growing economy that substantially benefits low-

income households, a broad strategy at the federal, state,

and local levels is needed to continue to support market

production and access to affordable homes and provide

assistance to those families most in need. With the expected

impact of the COVID-19 pandemic and associated economic

difficulties in 2020 and 2021, worst case housing needs have

the potential to increase substantially before HUD’s next

report, as explored in the following Special Addendum. A

comprehensive approach to housing policy is sorely needed

to address the long-standing and evolving challenge of worst

case housing needs.

SECTION 3. UNDERSTANDING THE TREND IN WORST CASE NEEDS

33WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

The Potential Effect of

the COVID Pandemic on

Worst Case Needs

This report on the nation’s worst case housing needs is based primarily on

American Housing Survey (AHS) data collected in 2019 before the novel

coronavirus set off a major COVID-19 pandemic early in 2020. The pandemic

caused extensive economic disruption, recessionary contraction,

and job

losses that ordinarily would be expected to increase worst case needs estimates

beyond these 2019 levels. HUD anticipates that future estimates of worst case

needs captured in the 2021 AHS will reflect the pandemic’s impact.

There is

cause to question, however, both how extensively worst case needs will change

and how reliably the AHS will capture such needs. This section discusses

factors that influence the answers.

Economic Implications of the Pandemic

Economic downturns can increase worst case needs simply by making

very low incomes more prevalent among renter households or through other

mechanisms. The Great Recession of 2007 to 2009 was followed by a lengthy

recovery that saw large increases in worst case needs. In the years leading up

to the recession, relaxed housing finance terms, aggressive and risky subprime

mortgage lending, and rapidly growing home prices had induced many

lower-income renters to finance home purchases under terms they could not

sustain. Such factors contributed to a near-collapse of financial markets, falling

house prices, sudden restrictions on mortgage capital, unemployment, and

extremely high levels of mortgage defaults and foreclosures that forced many

homeowners to become renters (HUD-PD&R, 2010). This tenure shift away from

homeownership greatly increased market demand pressure on the relatively

fixed rental stock and drove up market rents, even as economic conditions

further depressed the incomes of renter households and slowed housing

The National Bureau of Economic Research dates the recession from the peak of the business

cycle in February 2020 and has not yet noted a trough at the time of this writing in March 2021

(https://www.nber.org/research/data/us-business-cycle-expansions-and-contractions).

The 2021 AHS data collection will occur from May to September, 2021 (https://www.census.gov/

programs-surveys/ahs/about/respondent-information.html).

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

Special

Addendum

34 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

construction. Under the influence of these multiple forces,

the number of renter households increased by 11 percent,

and the number of very low-income renters increased by 21

percent between 2007 and 2011. Worst case housing needs

reached their historical peak in the same year, 2011, at 43

percent above their 2007 level.

The COVID-19 pandemic, however, has very different causes

and likely effects. Evidence suggests that social distancing

during the pandemic and record low mortgage interest rates

have motivated strong demand for homeownership. Home

sales volumes and prices are substantially higher at the

beginning of 2021 than they were a year earlier (HUD-PD&R,

2021). The move toward homeownership among renter

households with income stability is the opposite of consumer

reactions during the Great Recession and can dampen future

rent increases. Homeownership, however, is difficult for very

low-income renters to attain even with record-low mortgage

interest rates, so negative impacts to their incomes may

make their rental costs less affordable.

The labor market received its greatest shock from the

pandemic when mandatory social distancing restrictions

imposed during April 2020 increased the unemployment rate

to 14.3 percent, more than four times the 3.5 percent rate 2

months earlier in February (BLS, 2021a). Aggregate wage

and salary income in the second quarter of 2020 was down

7 percent from the first quarter, representing $617 billion of

lost household income (BEA, 2021a). The sharpest declines

in employment were recorded in the least telework-friendly

industries: leisure and hospitality, retail trade, construction,

transportation and warehousing, and manufacturing (Dalton,

2020).

After social distancing restrictions eased, the unemployment

picture improved substantially in the second half of 2020. By

January 2021, the official unemployment rate had improved

to 6.3 percent (still almost double pre-pandemic levels).

More than 4 million people, however, had left the labor force

during the previous 12 months; if they remained in the labor

force despite being unemployed in January 2021, then the

unemployment rate would be 8.8 percent. In January 2021,

5.7 percent of the workforce was unable to work at some

point in the last 4 weeks because their employers closed

or lost business due to the coronavirus pandemic, and 87.3

percent of these workers received no pay for the time not

working (BLS, 2021b).

Although the generosity of unemployment insurance benefits

varies substantially across states, in the first quarter of 2020,

unemployment insurance replaced about 45 percent of

wages on average or about $374 (38 percent) of $972 weekly

earnings.

Weekly earnings are based on 40 hours at the normal hourly wage (DOL, ETA, 2020).

Such shocks to household finances leave many renter

households unable to pay their rent and utilities. Lease

violations for non-payment of rent become grounds for

eviction and possible homelessness. Additionally, major

drops in rent revenues prevent many landlords from servicing

their mortgages and can lead to foreclosure and later

eviction.

Federal Pandemic Relief Legislation

Congress passed three major rounds of federal legislation

in response to the pandemic that provides direct or indirect

support for housing stability associated with income losses

or housing cost difficulties. The fiscal provisions highlighted

below can affect growth in the number of very low-income

households and the prevalence of severe housing problems

that cause worst case housing needs.

Coronavirus Aid, Relief, and Economic Security Act

(CARES Act, Pub.L. 116–136). Enacted in March 2020, the

act provided $2.2 trillion of economic relief and stimulus,

including these household benefits that could affect housing

affordability:

- Provided direct Economic Impact Payments of up to

$1,200 per income-eligible adult and $500 per child.

- Increased unemployment benefits by an additional $600

per week for 4 months.

Consolidated Appropriations Act of 2021 (Pub.L.

116–260). Enacted in December 2020, the act included a

COVID-19 relief bill totaling $868 billion that:

- Provided direct payments as Recovery Rebates of up to

$600 per income-eligible adult or child.

- Restored the increase of unemployment benefits at the

level of $300 per week and extended the maximum

period for collecting unemployment benefits to 50

weeks.

- Established a $25 billion Emergency Rental Assistance

(ERA) program for states administered by the U.S.

Department of Treasury.

American Rescue Plan Act (Pub.L. 117–2). Enacted on

March 11, 2021, the act provided an additional $1.9 trillion of

federal relief which:

- Provided direct payments as Recovery Rebates of up to

$1,400 per income-eligible adult or child.

- Extended increased unemployment benefits, including

the additional $300 weekly, through September 6, 2021.

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

35WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

- Expanded the Child Tax Credit for the 2021 tax year by

increasing the maximum amount from $2,000 to $3,000,

or $3,600 for children under age 6, made children

aged 17 eligible for the tax credit, made the tax credit

refundable, and provided for distribution through monthly

advance payments on the basis of 2019 or 2020 tax

returns.

- Expanded the Child and Dependent Care Credit for

the 2021 tax year by increasing the amount of eligible

expenses that may be used to calculate it, increased the

credit rate from 35 to 50 percent, made the tax credit

refundable, and increased the income phase-out.

- Expanded the Earned Income Tax Credit for the 2021

tax year by basing the credit on the higher of 2019 or

2021 income, raised the income phase-out level, and

increased the benefit for childless households from $543

to $1,502.

- Provided $21.55 billion of funding for Emergency

Rental Assistance and additional amounts for mortgage

assistance and homelessness prevention.

The various subsidies or transfers provided through the

pandemic relief acts have had a substantial impact on

household incomes. Aggregate personal income in the U.S.

reached $21.45 trillion in January 2021, up $1.14 trillion from

a year earlier. Transfer income accounted for $5.78 trillion of

the January total, including pandemic relief transfers of $1.96

trillion in economic impact payments to individuals and $570

billion in other relief (BEA, 2021b).

Eviction Moratoria

In addition to providing financial relief, the CARES Act

established a moratorium on eviction, dating from March 17,

2020, from any property with a federally backed mortgage

loan. The law did not prevent eviction, however, for violating

lease terms other than nonpayment of rent or other fees,

penalties, and charges. Numerous federal agencies,

including HUD’s Federal Housing Administration, issued

moratoria to implement the law. The Urban Institute estimated

that these federal eviction moratoria cover 12.4 percent of

rental units in single-family (one-to-four-unit) properties and

48.9 percent of rental units in multifamily properties, totaling

12.3 million units (Urban Institute, 2020).

Numerous states also issued eviction moratoria in March

or April 2020 to cover properties not covered by federal

mortgages; many state moratoria expired within a few months

(Benfer et al., 2020).

These personal transfer receipts do not include proprietor income such as the Paycheck Protection program.

On September 4, 2020, the Centers for Disease Control and

Prevention (CDC) issued a national moratorium on eviction as

a public health measure. The moratorium was subsequently

extended through June 30, 2021 (CDC, 2021).

What Percentage of Renter

Households Are Behind on Rent

During the Pandemic?

In addition to direct Economic Impact Payments, some

households have received short-term rental assistance

through state and local governments to mitigate income lost

through the pandemic. The Emergency Rental Assistance

program will provide significant additional funding to

state and local governments to fully or partially reimburse

landlords who face rent arrears.

To track the rapidly evolving situation during the pandemic,

the Census Bureau developed a frequently administered

survey (Pulse Survey) with the guidance of HUD and

other federal agencies. The Pulse data show that housing

problems have grown among renter households during the

pandemic. In exhibit SA-1, the number of renter households

who were behind on housing payments was already 6.5

million by late August 2020 and increased another 1.3 million

by the end of January 2021.

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

36 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit SA-1. Circumstances of Renter Households During the COVID-19 Pandemic

Renter households reporting specied

circumstances

(millions)

Week 13

(Aug. 19–31,

2020)

Week 17

(Oct. 14–26,

2020)

Week 21

(Dec. 9–21,

2020)

Week 23

(Jan. 20–Feb. 1,

2021)

Week 25

(Feb. 17–Mar. 1,

2021)

Behind on payment 6.48 6.84 8.24 7.75 7.73

Not at all confident in ability to pay rent on time 4.83 4.74 6.48 5.41 5.40

Behind on payments and eviction is “very likely”

in the next 2 months

1.15 1.03 1.46 1.22 1.58

Behind on payments and eviction is “somewhat

likely” in the next 2 months

1.91 1.96 2.83 2.38 2.26

Source: HUD PD&R tabulations of Phase 2 Pulse Survey data

As of late February 2021, an estimated 3.84 million renter

households report that eviction was either “very likely”

(1.58 million) or “somewhat likely” (2.26 million) in the next

2 months. The total figure is equivalent to almost half (49

percent) of the number of households with worst case needs

in 2019.

The Pulse data also suggest that rental housing difficulties

are not distributed uniformly. Point estimates of the mean

percentage of renters behind on rent range from a low of 12.0

percent for non-Hispanic Whites to 29.8 percent for non-

Hispanic Blacks. Where the 90-percent confidence intervals

overlap, however, such differences do not meet the threshold

for statistical significance.

Percentage of Renter Households

All Renters

17.3

20.1

18.8

29.8

12.0

Hispanic Asian, Non

Hispanic

Black, Non

Hispanic

White, Non

Hispanic

Exhibit SA-2. Percentage of Renters Behind on

Rent by Race and Ethnicity, with 90 Percent

Confidence Intervals, Jan. 20–Feb. 17, 2021

Source: HUD PD&R tabulations of Pulse Survey, weeks 23–25,

Jan. 20–Feb. 17, 2021

The Pulse data also show (exhibit SA-3) that renters who live

in single-family homes—about 50 percent of the occupied

rental stock—are more likely to be behind on rent than those

who live in multifamily buildings.

Source: U.S. Census Bureau, Pulse Survey, weeks 13–26 (pooled)

Behind on Rent (percentage of Renters)

Exhibit SA-3. Percentage of Renter Households Behind on

Rent, by Structure Type of Housing, August 2020 to March 2021

All Renters

Renters in Single-Family

Buildings (1-4 Units)

Renters in Multifamily

Buildings (5+ Units)

25%

Aug. 19–

Sep. 14

Sep. 16–

Oct. 12

Oct. 14–

Nov. 9

Nov. 11–

Dec. 7

Dec. 9–

Jan. 18

Jan. 20–

Feb. 15

Feb. 17–

Mar. 15

20%

15%

10%

In addition to being less likely to be behind on rent, rent

problems among households in multifamily properties

increased more slowly through the end of the year and were

resolved more quickly after the Department of Treasury

issued stimulus payments in early January. The reason for

such differences by structure type is not clear. One

possibility is that individual owners of single-family properties

might be more tolerant of delinquent rent than professional

managers of larger multifamily properties. AHS data do not

show clear patterns of either higher rents or more prevalent

severe rent burdens among renters in single-family versus

multifamily properties that might explain the disparity (U.S.

Census Bureau, n.d.).

Data from another national survey conducted by the

Consumer Finance Institute at the Federal Reserve Bank of

Philadelphia, however, suggest that landlord characteristics

do indeed play a significant role. Among renters who had

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

37WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

missed payments as of early January 2021, 79.9 percent

reported reaching an agreement with their landlord to either

pay a lump sum at a later date (37.6 percent), pay the back

rent over time (40.3 percent), or have the back rent forgiven

(6.7 percent). Only 34.2 percent of those who negotiated with

their landlord, however, made the agreement in writing. The

last 20.1 percent of renters missed payments without even a

verbal agreement, placing them at risk of future legal action

(Akana, 2021).

Federal pandemic relief payments made a significant

contribution to household budgets and to rent. In the Pulse

survey of late February 2021, of an estimated 31.4 million

individuals who received a stimulus payment during the

preceding 7 days and mostly spent it, 7.7 million spent it on

rent (U.S. Census Bureau, 2021). Another 13.8 million spent

it on utilities and telecommunications, although this total

includes homeowners.

How Well Will the AHS Measure

Pandemic Effects in 2021?

HUD and the Census Bureau are implementing eviction

questions in the 2021 AHS similar to those used in the Pulse

Survey. The AHS longitudinal design, which revisits the

same housing units every 2 years, is not ideal for measuring

eviction problems that may develop over shorter intervals.

The AHS data may prove useful, however, in assessing

the extent of housing instability associated with the severe

housing problems that compose worst case needs or with

household characteristics.

The federal eviction moratoria, as well as various local

moratoria, enabled numerous renters to skip rent payments

for a period that extends into 2021, but at present, such

moratoria have not canceled obligations to pay back rent.

Provided they can avoid foreclosure triggered by diminished

rental income, landlords may have three main options:

- Landlords might evict the tenants with unpaid rent and

lease the unit to a new tenant.

- Landlords might be reimbursed in part or in full

for unpaid rent revenues by the Emergency Rental

Assistance program or other sources.

- Landlords might work out formal or informal

arrangements with existing tenants to repay back rent

over time.

It is not clear whether these unusual income sources,

deferred or cancelled rent obligations, and subsequent

evictions or other housing disruptions will be captured

accurately by the AHS or result in major increases in worst

case needs for 2021.

The impacts of employment and

potential eviction may be substantially offset by pandemic

relief that is not fully measured by the AHS.

The AHS measures income for the 12 months preceding

the interview for everyone age 16 and older who currently

lives in the housing unit. “Money income” is the income

received on a regular basis before paying personal income

taxes, social security, union dues, Medicare deductions, etc.

(Census-HUD 2019). Because of this definition of income,

lump-sum transfer payments such as the economic impact

payments provided by the three relief bills will not be counted

as income in the AHS questionnaire. AHS pre-tax income

also would not capture increases to household income

from tax expenditures such as the child tax credit or tax

code changes that increase the household’s tax deduction.

Although the AHS questionnaire does not capture such

income sources, they probably could be approximated fairly

accurately afterward in regard to these pandemic responses.

Exhibit SA-4 examines how the money income of a

hypothetical four-person household with extremely low

income might be measured by the AHS, and how post-tax

income differs (shown in brackets).

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

38 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit SA-4. How Pandemic Relief Benets Could End the Worst Case Needs Status of a Four-Person Extremely Low-

Income Renter Household Participating in a mid-2021 American Housing Survey Interview

Sources of Money Income

and Transfers

Without Pandemic

Conditions or Relief

With Pandemic

Conditions and Relief

Earned income (two workers with full-time minimum wage) $29,580 $0

Unemployment Insurance Payments

(38% of base earnings plus $300/week per worker through Sep.6, 2021)

$0 $35,240

Direct Payments, rounds 2 and 3:

$600+$1400 per person times 4

$0 [$8,000]

Earned Income Tax Credit

[$5,980] [$5,980]

Child Tax Credit (two children, one <6 yrs) [$4,000] [$6,600]

Child and Dependent Care Credit (35% of $20,000; assume childcare needed

when working)

[$7,000] N/A

Pre-tax money income counted by AHS $29,580 $35,240

Post-tax income [$46,560] [$ 47, 82 0]

Housing Cost Burden of $1,250 gross rent relative to pre-tax money income

measured by AHS

50.7% 42.6%

Housing Cost Burden of $1,250 gross rent relative to post-tax income 32.2% 31.4%

AHS = American Housing Survey. NA = not applicable.

Note: Values in brackets [ ] are not counted as money income in the American Housing Survey.

EITC information is found at https://www.irs.gov/credits-deductions/individuals/earned-income-tax-credit/earned-income-and-earned-income-tax-credit-eitc-

tables#EITC%20Tables.

Because disbursements for direct payments and child tax

credits are post-tax income and are not regularly received,

the AHS will not include them as income. Nevertheless,

enhanced unemployment insurance alone might be sufficient

to increase household income enough to eliminate a severe

cost burden or to raise an extremely low- or very low-income

household to a higher income category. On the other hand,

households might lose substantial income if workers become

unemployed in states with less generous unemployment

benefits. Either way, numbers of worst case needs can

change readily because the number of very low-income

renters changes or because their housing cost burdens may

change relative to the 50 percent-of-income threshold for

severe burden.

Summary

The impacts of the COVID-19 pandemic on the economy

and housing market have rivaled those of major recessions.

Had Congress not passed multiple stimulus acts in 2020 and

early 2021, major increases in worst case housing needs

during 2021 clearly would be unavoidable. Increased risk of

foreclosures and evictions, increased numbers of renters with

incomes below the very low-income threshold, and increased

financial stresses on landlords almost certainly would worsen

every component of the worst case needs measure.

As discussed in this analysis, however, the legislative

response has been remarkable in its scope and scale. The

numerous direct and indirect federal benefits for low-income

households, for the most part, are not benefits that are

readily measured by a survey such as AHS that focuses on

pre-tax income. It is possible, although not likely, that some

of the direct benefits such as enhanced unemployment

insurance could improve the financial conditions of some

low-wage working households enough that they could even

escape very low-income status. Additionally, households who

become new recipients of the emergency housing vouchers

funded through 2023 would not be recorded as having worst

case needs should they appear in the AHS sample.

All told, there is little clarity about how the pandemic and the

policy response will play out for vulnerable rental households

and the measurement of worst case needs. Nevertheless,

the housing crisis is as real as the health crisis for those

most immediately affected. HUD and partners of all types

must remain diligent to avoid housing calamity in the coming

months and address the underlying affordable housing crisis

over the longer term.

SPECIAL ADDENDUM. THE POTENTIAL EFFECT OF THE COVID PANDEMIC ON WORST CASE NEEDS

39WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Detailed Data on

Housing Problems and

Supply of Affordable

Housing

Exhibit A-1A. Housing Conditions of Renter Households by Relative Income,

2017 and 2019

Exhibit A-1B. Housing Conditions of Owner Households by Relative Income,

2017 and 2019

Exhibit A-2A. Housing Conditions of Renters and Owners, 2001–2019—

Number of Households

Exhibit A-2B. Housing Conditions of Renters and Owners, 2001–2019—

Percentage of Households

Exhibit A-3. Housing Conditions of Unassisted Renter Households by Relative

Income, 2017 and 2019

Exhibit A-4. Prevalence of Housing Problems Among Renters by Relative

Income, 2017 and 2019

Exhibit A-5A. Prevalence of Housing Problems Among Very Low-Income

Renters by Household Type, 2017 and 2019

Exhibit A-5B. Prevalence of Housing Problems Among Very Low-Income

Renter Households Containing People with Disabilities by Household Type,

2017 and 2019

Exhibit A-6A. Housing Problems and Characteristics of Very Low-Income

Renters by Household Type, 2019

Exhibit A-6B. Housing Problems and Characteristics of Extremely Low-Income

Renters by Household Type, 2019

Exhibit A-7. Housing Problems and Characteristics of Very Low-Income Worst

Case Renters by Household Type, 2019

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

APPENDIX A

40 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-8. Housing Problems and Characteristics of

Extremely Low-Income Worst Case Renters by Household

Type, 2019

Exhibit A-9. Prevalence of Housing Problems Among

Very Low-Income Renters by Race and Ethnicity, 2017 and

2019—Number and Percentage

Exhibit A-10. Prevalence of Housing Problems Among Very

Low-Income Renters by Region, 2017 and 2019—Number

and Percentage

Exhibit A-11A. Prevalence of Housing Problems Among Very

Low-Income Renters by Metropolitan Location, 2017 and

2019—Number and Percentage

Exhibit A-11B. Housing Conditions of Renter Households

by Relative Income, Sampled Metropolitan Areas, 2017 and

2019

Exhibit A-12. Households Occupying Rental Units by

Affordability of Rent and Income of Occupants, 2017 and

2019

Exhibit A-13. Renters and Rental Units Affordable and

Available to Them by Relative Income, 2001–2019

Exhibit A-14. Average Income and Average Gross Rent of

Renter Households by Relative Income, 2017 and 2019

Exhibit A-15. Housing Conditions of Households Having

People Younger Than 62 Who Have Disabilities by Disability

Type, 2017 and 2019

The U.S. Census Bureau reviewed this data product for

unauthorized disclosure of confidential information and

approved the disclosure avoidance practices applied to this

release.

CBDRB-FY21-POP001-0037.

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

41WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-1A. Housing Conditions of Renter Households by Relative Income, 2017 and 2019

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2019 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 11,748 6,640 8,786 7,583 9,902 44,659

Unassisted with severe problems 5,780 1,986 1,013 372 200 9,351

Unassisted with nonsevere problems only 955 2,642 3,805 1,737 1,086 10,225

Unassisted with no problems 1,064 909 3,378 5,206 8,363 18,920

Assisted 3,950 1,10 3 590 268 253 6 ,16 4

Any with severe problems 7,537 2,157 1,041 380 206 11,321

Rent burden >50% of income

7,3 72 2,064 880 289 133 10,738

Severely inadequate housing

353 131 169 90 73 816

Any with nonsevere problems only

2,087 3,153 4,029 1,787 1,118 12,174

Rent burden >30–50% of income

1,791 3,016 3,510 1,358 616 10,291

Moderately inadequate housing

336 274 479 289 351 1,729

Crowded housing

247 196 314 168 181 1,10 6

Any with no problems

2,124 1,329 3,715 5,417 8,578 21,163

2017 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 11,548 6,519 8,637 7,30 6 9,983 43,993

Unassisted with severe problems

5,555 2,161 973 277 232 9,198

Unassisted with nonsevere problems only

1,049 2,381 3,804 1,768 1,18 0 10,181

Unassisted with no problems

908 823 3,220 5,003 8,272 18,226

Assisted

4,037 1,15 4 641 259 298 6,388

Any with severe problems 7,362 2,411 1,010 286 243 11,312

Rent burden >50% of income

7,198 2,300 879 236 144 10,757

Severely inadequate housing

378 156 134 54 104 826

Any with nonsevere problems only 2,127 2,890 4,066 1,810 1,219 12,113

Rent burden >30–50% of income

1,848 2,756 3,541 1,463 607 10,215

Moderately inadequate housing

321 256 433 258 439 1,708

Crowded housing

256 241 376 149 223 1,245

Any with no problems 2,060 1,217 3,562 5,210 8,520 20,568

Source: HUD PD&R tabulations of the American Housing Survey data

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

42 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-1B. Housing Conditions of Owner Households by Relative Income, 2017 and 2019

Household Income as Percentage of HUD-Adjusted Area Median Family

Income

2019 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 8,265 6,355 11,741 14,516 38,599 79,476

Unassisted with severe problems

5,082 1,772 1,410 764 598 9,626

Unassisted with nonsevere problems only

1,439 2,124 3,581 3,016 3,209 13,369

Unassisted with no problems

1,744 2,459 6,750 10,736 34,793 56,482

Any with severe problems 5,082 1,772 1,410 764 598 9,626

Cost burden >50% of income

4,974 1,722 1,322 626 467 9,111

Severely inadequate housing

215 58 98 138 133 642

Any with nonsevere problems only 1,439 2,124 3,581 3,016 3,209 13,369

Cost burden >30–50% of income

1,260 1,949 3,15 5 2,586 2,435 11,38 5

Moderately inadequate housing

196 188 372 353 610 1,719

Crowded housing

88 78 201 148 219 734

Any with no problems 1,744 2,459 6,750 10,736 34,793 56,482

2017 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 7,883 6,172 10,959 13,736 38,817 77,567

Unassisted with severe problems

4,829 1,756 1,400 74 4 667 9,396

Unassisted with nonsevere problems only

1,365 2,12 5 3,481 3,12 8 3,353 13,452

Unassisted with no problems

1,689 2,291 6,078 9,864 34,797 54,719

Any with severe problems 4,829 1,756 1,400 744 667

9,396

Cost burden >50% of income

4,742 1,692 1,347 658 527 8,967

Severely inadequate housing

146 78 66 87 141 517

Any with nonsevere problems only 1,365 2,125 3,481 3,128 3,353 13,452

Cost burden >30–50% of income

1,174 1,952 3,13 6 2,674 2,454 11,391

Moderately inadequate housing

236 220 308 346 750 1,860

Crowded housing

75 64 195 187 186 706

Any with no problems 1,689 2,291 6,078 9,864 34,797 54,719

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

43WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-2A. Housing Conditions of Renters and Owners, 2001–2019—Number of Households

2001 2003 2005 2007 2009 2011 2013 2015 2017 2019

Total households

(thousands)

105,435 105,868 108,901 110,719 111,861 115,076 116,032 118,290 121,560 124,135

Unassisted with severe

problems 13,494 13,398 16 ,142 16,944 19,259 20,717 18,553 18,000 18,594 18,978

Unassisted with nonsevere

problems only 19,217 19,790 20,849 22,752 23,225 24,079 22,15 3 21,672 23,633 23,593

Unassisted with no problems 66,445 66,468 65,362 65,862 64,506 64,983 69,796 73,059 72,945 75,400

Assisted 6,279 6,211 6,547 5 ,161 4,871 5,298 5,530 5,559 6,388 6,16 4

Cost burden >50% of income 13,330 13,188 16,433 17,14 0 19,458 20,781 18,810 18,799 19,724 19,849

Cost burden >30–50% of

income 16,923 17,8 56 19,403 21,15 3 21,818 22,369 20,884 19,252 21,606 21,676

Severely inadequate housing 2,10 8 1,971 2,023 1,805 1,866 2,12 6 1,942 1,500 1,343 1,458

Moderately inadequate

housing 4,504 4,311 4,177 3,954 3,884 3 ,133 3,946 3,907 3,568 3,449

Crowded housing 2,631 2,559 2,621 2,529 2,509 1,923 2,509 1,803 1,951 1,840

Renter households

(thousands)

33,727 33,614 33,951 35,054 35,396 38,867 40,273

43,930 43,993 44,660

Unassisted with severe

problems 5,758 5,887 6,860 6,993 8,085 9,548 8,874 9,651 9,198 9,352

Unassisted with nonsevere

problems only 7,2 83 7,557 7, 30 3 8,445 8,229 9,194 9,233 10,455 10,181 10,225

Unassisted with no problems 14,407 13,958 13,240 14,455 14, 211 14,828 16,636 18,265 18,226 18,919

Assisted 6,279 6,211 6,547 5 ,161 4,871 5,298 5,530 5,559 6,388 6 ,16 4

Cost burden >50% of income 6,412 6,477 7,891 7,79 3 9,000 10,391 9,74 4 10,988 10,757 10,738

Cost burden >30–50% of

income 6,916 7,468 7,5 02 8,340 8,240 9,124 9,292 10,118 10,215 10,291

Severely inadequate housing 1,16 8 1,038 1,10 0 1,073 998 1,204 1,155 828 826 816

Moderately inadequate

housing 2,508 2,525 2,542 2,400 2,264 1,830 2,508 2,027 1,708 1,730

Crowded housing 1,658 1,615 1,635 1,511 1,499 1,072 1,652 1,120 1,245 1,10 6

Owner households

(thousands)

71,708 72,254 74,950 75,665 76,465 76,209 75,759 74,360 7 7,567 79,475

Unassisted with severe

problems 7,73 6 7, 511 9,282 9,951 11,174 11,16 9 9,679 8,349 9,396 9,626

Unassisted with nonsevere

problems only 11,934 12,233

13,546 14,307 14,996 14,885 12,920 11,217 13,452 13,368

Unassisted with no problems 52,038 52,510 52,12 2 51,407 50,295 50,155 53,16 0 54,794 54,719 56,481

Cost burden >50% of income 6,918 6,711 8,542 9,347 10,458 10,390 9,066 7, 811 8,967 9,111

Cost burden >30–50% of

income 10,007 10,388 11,9 01 12,813 13,578 13,245 11,592 9,13 5 11,391 11,3 85

Severely inadequate housing 940 933 923 732 868 922 787 673 517 642

Moderately inadequate

housing 1,996 1,786 1,635 1,554 1,620 1,303 1,438 1,881 1,860 1,719

Crowded housing 973 944 986 1,018 1,010 851 857 683 706 734

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

44 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-2B. Housing Conditions of Renters and Owners, 2001–2019—Percentage of Households

2001 2003 2005 2007 2009 2011 2013 2015 2017 2019

Total households 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Unassisted with severe problems 12.8 12.7 14.8 15.3 17.2 18.0 16.0 15.2 15.3 15.3

Unassisted with nonsevere problems only 18.2 18.7 19.1 20.5 20.8 20.9 19.1 18.3 19.4 19.0

Unassisted with no problems 63.0 62.8 60.0 59.5 57.7 56.5 60.2 61.8 60.0 60.7

Assisted 6.0 5.9 6.0 4.7 4.4 4.6 4.8 4.7 5.3 5.0

Cost burden >50% of income 12.6 12.5 15.1 15.5 17.4 18.1 16.2 15.9 16.2 16.0

Cost burden >30–50% of income 16 .1 16.9 17.8 19.1 19.5 19.4 18.0 16.3 17. 8 17.5

Severely inadequate housing 2.0 1.9 1.9 1.6 1.7 1.8 1.7 1.3 1.1 1.2

Moderately inadequate housing 4.3 4.1 3.8 3.6 3.5 2.7 3.4 3.3 2.9 2.8

Crowded housing 2.5 2.4 2.4 2.3 2.2 1.7 2.2 1.5 1.6 1.5

Renter households 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Unassisted with severe problems 17.1 17.5 20.2 19.9 22.8 24.6 22.0 22.0 20.9 20.9

Unassisted with nonsevere problems only 21.6 22.5 21.5 24.1 23.2 23.7 22.9 23.8 2 3.1 22.9

Unassisted with no problems 42.7 41.5 39.0 41.2 4 0.1 38.2 41.3 41.6 41.4 42.4

Assisted 18.6 18.5 19.3 14.7 13.8 13.6 13.7 12.7 14.5 13.8

Cost burden >50% of income 19.0 19.3 23.2 22.2 25.4 26.7 24.2 25.0 24.5 24.0

Cost burden >30–50% of income 20.5 22.2 22.1 23.8 23.3 23.5 2 3.1 23.0 23.2 23.0

Severely inadequate housing 3.5 3 .1 3.2 3 .1 2.8 3.1 2.9 1.9 1.9 1.8

Moderately inadequate housing 7. 4 7. 5 7. 5 6.8 6.4 4.7 6.2 4.6 3.9 3.9

Crowded housing 4.9 4.8 4.8 4.3 4.2 2.8 4.1 2.6 2.8 2.5

Owner households 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Unassisted with severe problems 10.8 10.4 12.4 13.2 14.6 14.7 12.8 11.2 12.1 12.1

Unassisted with nonsevere problems only 16.6 16.9 18.1 18.9 19.6 19.5 17.1 15.1 17. 3 16.8

Unassisted with no problems 72.6 72.7 69.5 67.9 65.8 65.8 70.2 73.7 70.5 71.1

Cost burden >50% of income 9.6 9.3 11.4 12.4 13.7 13.6 12.0 10.5 11.6 11.5

Cost burden >30–50% of income 14.0 14.4 15.9 16.9 17.8 17.4 15.3 12.3 14.7 14.3

Severely inadequate housing 1.3 1.3 1.2 1.0 1.1 1.2 1.0 0.9 0.7 0.8

Moderately inadequate housing 2.8 2.5 2.2 2.1 2.1 1.7 1.9 2.5 2.4 2.2

Crowded housing 1.4 1.3 1.3 1.3 1.3 1.1 1.1 0.9 0.9 0.9

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

45WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-3. Housing Conditions of Unassisted Renter Households by Relative Income, 2017 and 2019

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2019 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total unassisted households (thousands) 7,799 5,537 8,196 7,315 9,650 38,497

Any with severe problems 5,780 1,986 1,013 372 200 9,351

Rent burden >50% of income 5,672 1,896 863 283 128 8,842

[Rent above FMR] 2,065 1,047 658 280 128 4,178

Severely inadequate housing 247 127 157 89 72 692

Any with nonsevere problems only 955 2,642 3,805 1,737 1,086 10,225

Rent burden >30–50% of income 764 2,541 3,327 1,322 610 8,564

Moderately inadequate housing 213 216 427 278 335 1,469

Crowded housing 175 178 306 165 172 996

Any with no problems 1,064 909 3,378 5,206 8,363 18,920

2017 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total unassisted households (thousands) 7,511 5,365 7,997 7,048 9,685 37,605

Any with severe problems 5,555 2,161 973 277 232 9,198

Rent burden >50% of income 5,453 2,068 856 230 144 8,750

[Rent above FMR] 1,898 1,197 779 223 143 4,240

Severely inadequate housing 266 131 120 52 93 662

Any with nonsevere problems only 1,049 2,381 3,804 1,768 1,180 10,181

Rent burden >30–50% of income 852 2,278 3,333 1,432 598 8,494

Moderately inadequate housing 208 200 406 255 421 1,489

Crowded housing 200 217 336 140 211 1,10 3

Any with no problems 908 823 3,220 5,003 8,272 18,226

FMR = Fair Market Rent.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

46 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-4. Prevalence of Housing Problems Among Renters by Relative Income, 2017 and 2019

Number Percentage

2017 2019 2017 2019

Renter households (thousands) 43,990 44,659 100.0 100.0

Unassisted with severe problems 9,19 8 9,351 20.9 20.9

Unassisted with nonsevere problems only 10,18 0 10,225 2 3.1 22.9

Unassisted with no problems 18,230 18,920 41.4 42.4

Assisted 6,388 6,16 4 14.5 13.8

Any with severe problems 11,310 11,321 25.7 25.3

Rent burden >50% of income 10,760 10,738 24.5 24.0

Severely inadequate housing 826 816 1.9 1.8

[Rent burden only] 9,748 9,743 22.2 21.8

Any with nonsevere problems only 12 ,110 12,174 27.5 27.3

Rent burden >30–50% of income 10,220 10,291 23.2 23.0

Moderately inadequate housing 1,708 1,729 3.9 3.9

Crowded housing 1,245 1,10 6 2.8 2.5

[Rent burden only] 9,254 9,402 21.0 21.1

Any with no problems 20,570 21,163 46.8 47.4

Income 0-30% HAMFI (thousands) 11,550 11,748 100.0 100.0

Unassisted with severe problems 5,555 5,780 4 8.1 49.2

Unassisted with nonsevere problems only 1,049 955 9.1 8.1

Unassisted with no problems 908 1,064 7. 9 9.1

Assisted 4,037 3,950 35.0 33.6

Any with severe problems 7,362 7,537 63.7 64.2

Rent burden >50% of income 7,19 8 7,3 72 62.3 62.8

Severely inadequate housing 378 353 3.3 3.0

[Rent burden only] 6,406 6,601 55.5 56.2

Any with nonsevere problems only 2,127 2,087 18.4 17.8

Rent burden >30–50% of income 1,848 1,791

16.0 15.2

Moderately inadequate housing 321 336 2.8 2.9

Crowded housing 256 247 2.2 2.1

[Rent burden only] 1,570 1,523 13.6 13.0

Any with no problems 2,060 2,124 17.8 18.1

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

47WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

Income >30-50% HAMFI (thousands) 6,519 6,640 100.0 100.0

Unassisted with severe problems 2,161 1,986 3 3.1 29.9

Unassisted with nonsevere problems only 2,381 2,642 36.5 39.8

Unassisted with no problems 823 909 12.6 13.7

Assisted 1,15 4 1,10 3 17.7 16.6

Any with severe problems 2,411 2,157 37.0 32.5

Rent burden >50% of income 2,300 2,064 35.3 31.1

Severely inadequate housing 156 131 2.4 2.0

[Rent burden only] 2,136 1,900 32.8 28.6

Any with nonsevere problems only 2,890 3,153 44.3 47.5

Rent burden >30–50% of income 2,756 3,016 42.3 45.4

Moderately inadequate housing 256 274 3.9 4.1

Crowded housing 241 196 3.7 3.0

[Rent burden only] 2,415 2,692 37.0 40.5

Any with no problems 1,217 1,329 18.7 20.0

Income >50-80% HAMFI (thousands) 8,637 8,786 100.0 100.0

Unassisted with severe problems 973 1,013 11.3 11.5

Unassisted with nonsevere problems only 3,804 3,805 44.0 43.3

Unassisted with no problems 3,220 3,378 37. 3 38.4

Assisted 641 590 7. 4 6.7

Any with severe problems 1,010 1,041 11.7 11.8

Rent burden >50% of income 879 880 10.2 10.0

Severely inadequate housing 134 169 1.6 1.9

[Rent burden only] 849 837 9.8 9.5

Any with nonsevere problems only 4,066 4,029 47.1 45.9

Rent burden >30–50% of income 3,541 3,510 41.0 39.9

Moderately inadequate housing 433 479 5.0 5.5

Crowded housing 376 314 4.4 3.6

[Rent burden only] 3,278 3,248 38.0 37.0

Any with no problems 3,562 3,715 41.2 42.3

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-4. Prevalence of Housing Problems Among Renters by Relative Income, 2017 and 2019 (continued)

(continued)

48 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

Income >80–120% HAMFI (thousands) 7,306 7,583 100.0 100.0

Unassisted with severe problems 277 372 3.8 4.9

Unassisted with nonsevere problems only 1,768 1,737 24.2 22.9

Unassisted with no problems 5,003 5,206 68.5 68.7

Assisted 259 268 3.5 3.5

Any with severe problems 286 380 3.9 5.0

Rent burden >50% of income 236 289 3.2 3.8

Severely inadequate housing 54 90 0.7 1.2

[Rent burden only] 226 274 3.1 3.6

Any with nonsevere problems only 1,810 1,787 24.8 23.6

Rent burden >30–50% of income 1,463 1,358 20.0 17.9

Moderately inadequate housing 258 289 3.5 3.8

Crowded housing 149 168 2.0 2.2

[Rent burden only] 1,407 1,335 19.3 17. 6

Any with no problems 5,210 5,417 71.3 71.4

Income >120% HAMFI (thousands) 9,983 9,902 100.0 100.0

Unassisted with severe problems 232 200 2.3 2.0

Unassisted with nonsevere problems only 1,18 0 1,086 11.8 11.0

Unassisted with no problems 8,272 8,363 82.9 84.5

Assisted 298 253 3.0 2.6

Any with severe problems 243 206 2.4 2.1

Rent burden >50% of income 144 133 1.4 1.3

Severely inadequate housing 104 73 1.0 0.7

[Rent burden only] 132 131 1.3 1.3

Any with nonsevere problems only 1,219 1,118 12.2 11.3

Rent burden >30–50% of income 607 616 6 .1 6.2

Moderately inadequate housing 439 351 4.4 3.5

Crowded housing 223 181 2.2 1.8

[Rent burden only] 582 604 5.8 6.1

Any with no problems 8,520 8,578 85.3 86.6

HAMFI = HUD-Adjusted Median Family Income.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-4. Prevalence of Housing Problems Among Renters by Relative Income, 2017 and 2019 (continued)

49WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-5A. Prevalence of Housing Problems Among Very Low-Income Renters by Household Type, 2017 and 2019

Number Percentage

Household type

2017 2019 2017 2019

All household types (thousands) 18,067 18,388 100.0 100.0

Older adults without children (thousands) 4,960 5,567 100.0 100.0

Unassisted with severe problems 1,932 2,241 39.0 40.3

Unassisted with nonsevere problems only 641 74 3 12.9 13.3

Unassisted with no problems 467 590 9.4 10.6

Assisted 1,920 1,993 38.7 35.8

Any with severe problems 2,600 3,002 52.4 53.9

Rent burden >50% of income 2,571 2,930 51.8 52.6

Severely inadequate housing 75 131 1.5 2.4

[Rent burden only] 2,372 2,636 47.8 47.4

Any with nonsevere problems only 1,279 1,394 25.8 25.0

Rent burden >30–50% of income 1,19 0 1,320 24.0 23.7

Moderately inadequate housing 155 142 3.1 2.6

Crowded housing (s) (D) (s) (D)

[Rent burden only] 1,12 0 1,252 22.6 22.5

Any with no problems 1,081 1,172 21.8 21.1

Families with children (thousands) 6,199 5,654 100.0 100.0

Unassisted with severe problems 2,571 2,271 41.5 40.2

Unassisted with nonsevere problems only 1,528 1,463 24.6 25.9

Unassisted with no problems 441 470 7.1 8.3

Assisted 1,659 1,450 26.8 25.6

Any with severe problems 3,333 2,865 53.8 50.7

Rent burden >50% of income 3,228 2,797 52.1 49.5

Severely inadequate housing 223 131 3.6 2.3

[Rent burden only] 2,869 2,509 46.3 44.4

Any with nonsevere problems only 2,020

1,975 32.6 34.9

Rent burden >30–50% of income 1,799 1,760 29.0 31.1

Moderately inadequate housing 230 194 3.7 3.4

Crowded housing (s) 420 (s) 7.4

[Rent burden only] 1,377 1,383 22.2 24.5

Any with no problems 847 813 13.7 14.4

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

50 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

Household type

2017 2019 2017 2019

Other family households (thousands) 1,591 1,649 100.0 100.0

Unassisted with severe problems 716 720 45.0 43.7

Unassisted with nonsevere problems only 335 432 21.1 26.2

Unassisted with no problems 212 196 13.3 11.9

Assisted 327 301 20.6 18.3

Any with severe problems 843 821 53.0 49.8

Rent burden >50% of income 826 795 51.9 48.2

Severely inadequate housing 55 45 3.5 2.7

[Rent burden only] 710 719 44.6 43.6

Any with nonsevere problems only 423 520 26.6 31.5

Rent burden >30–50% of income 385 489 24.2 29.7

Moderately inadequate housing 41 70 2.6 4.2

Crowded housing (s) 15 (s) 0.9

[Rent burden only] 347 436 21.8 26.4

Any with no problems 324 308 20.4 18.7

Other nonfamily households (thousands) 5,317 5,518 100.0 100.0

Unassisted with severe problems 2,497 2,535 47.0 45.9

Unassisted with nonsevere problems only 925 958 17.4 17.4

Unassisted with no problems 611 716 11.5 13.0

Assisted 1,284 1,309 24.1 23.7

Any with severe problems 2,997 3,006 56.4 54.5

Rent burden >50% of income 2,872 2,913 54.0 52.8

Severely inadequate housing 181 177 3.4 3.2

[Rent burden only] 2,589 2,636 48.7 47. 8

Any with nonsevere problems only 1,295 1,351 24.4 24.5

Rent burden >30–50% of income 1,229 1,238 2 3.1 22.4

Moderately inadequate housing 151 206 2.8 3.7

Crowded housing (s) (D) (s) (D)

[Rent burden only] 1,142 1,145 21.5 20.8

Any with no problems 1,025 1,161 19.3 21.0

(s) = Unweighted counts of 5 or fewer suppressed. (D) = value suppressed in accord with Census Bureau disclosure prevention requirements.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-5A. Prevalence of Housing Problems Among Very Low-Income Renters by Household Type,

2017 and 2019 (continued)

51WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-5B. Prevalence of Housing Problems Among Very Low-Income Renter Households Containing People with

Disabilities* by Household Type, 2017 and 2019

Number Percentage

Household type 2017 2019 2017 2019

All household types (thousands) 3,276 2,895 100.0 100.0

Older adults without children (thousands) 154 105 100.0 100.0

Unassisted with severe problems 47 42 30.5 40.0

Unassisted with nonsevere problems only 49 20 31.8 19.0

Unassisted with no problems 4 8 2.6 7.6

Assisted 54 36 3 5.1 34.3

Any with severe problems 72 57 46.8 54.3

Rent burden >50% of income 72 49 46.8 46.7

Severely inadequate housing (s) 10 (s) 9.5

[Rent burden only] 68 41 44.2 39.0

Any with nonsevere problems only 70 33 45.5 31.4

Rent burden >30–50% of income 59 32 38.3 30.5

Moderately inadequate housing 15 3 9.7 2.9

Crowded housing (s) (D) (s) (D)

[Rent burden only] 56 30 36.4 28.6

Any with no problems 12 15 7.8 14.3

Families with children (thousands) 1,175 892 100.0 100.0

Unassisted with severe problems 485 336 41.3 37.7

Unassisted with nonsevere problems only 236 228 2 0.1 25.6

Unassisted with no problems 53 33 4.5 3.7

Assisted 401 295 3 4.1 33.1

Any with severe problems 690 437 58.7 49.0

Rent burden >50% of income 669 416 56.9 46.6

Severely inadequate housing (s) 45 (s) 5.0

[Rent burden only] 541 346 46.0 38.8

Any with nonsevere problems only 347

340 29.5 38.1

Rent burden >30–50% of income 302 298 25.7 33.4

Moderately inadequate housing 69 71 5.9 8.0

Crowded housing (s) 51 (s) 5.7

[Rent burden only] 213 219 18 .1 24.6

Any with no problems 138 116 11.7 13.0

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

52 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

Household type 2017 2019 2017 2019

Other family households (thousands) 436 403 100.0 100.0

Unassisted with severe problems 200 164 45.9 40.7

Unassisted with nonsevere problems only 76 80 17. 4 19.9

Unassisted with no problems 36 46 8.3 11.4

Assisted 124 114 28.4 28.3

Any with severe problems 243 204 55.7 50.6

Rent burden >50% of income 228 195 52.3 48.4

Severely inadequate housing (s) 13 (s) 3.2

[Rent burden only] 189 166 43.3 41.2

Any with nonsevere problems only 111 113 25.5 28.0

Rent burden >30–50% of income 104 95 23.9 23.6

Moderately inadequate housing 17 32 3.9 7. 9

Crowded housing (s) (D) (s) (D)

[Rent burden only] 94 80 21.6 19.9

Any with no problems 82 86 18.8 21.3

Other nonfamily households (thousands) 1,511 1,495 100.0 100.0

Unassisted with severe problems 572 503 37. 9 33.6

Unassisted with nonsevere problems only 174 238 11.5 15.9

Unassisted with no problems 90 110 6.0 7.4

Assisted 677 644 44.8 4 3.1

Any with severe problems 797 730 52.7 48.8

Rent burden >50% of income 739 703 48.9 47.0

Severely inadequate housing (s) 64 (s) 4.3

[Rent burden only] 620 577 41.0 38.6

Any with nonsevere problems only 374 439 24.8 29.4

Rent burden >30–50% of income 355 391 23.5 26.2

Moderately inadequate housing 47 102 3.1 6.8

Crowded housing (s) (D) (s) (D)

[Rent burden only] 326 337 21.6 22.5

Any with no problems 341 326 22.6 21.8

(s) = Unweighted counts of 5 or fewer suppressed. (D) = value suppressed in accord with Census Bureau disclosure prevention requirements.

* Older adults with disabilities were excluded.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-5B. Prevalence of Housing Problems Among Very Low-Income Renter Households Containing

People with Disabilities* by Household Type, 2017 and 2019 (continued)

53WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-6A. Housing Problems and Characteristics of Very Low-Income Renters by Household Type, 2019

Total Older Adults,

No Children

Families with

Children

Other

Families

Other

Nonfamily

Households

Renter households (thousands) 18,388 5,567 5,654 1,649 5,518

Number of children 11,427 NA 11,427 NA NA

Number of persons 39,704 7,075 21,744 4,076 6,809

Children/household 2.02 NA 2.02 NA NA

Persons/household 2.16 1.27 3.85 2.47 1.23

Unassisted with severe problems 7,767 2,241 2,271 720 2,535

Unassisted with nonsevere problems only 3,596 743 1,463 432 958

Unassisted with no problems 1,972 590 470 196 716

Assisted 5,053 1,993 1,450 301 1,309

Any with severe problems 9,694 3,002 2,865 821 3,006

Rent burden >50% of income 9,435 2,930 2,797 795 2,913

Severely inadequate housing 484 131 131 45 177

[Rent burden only] 8,500 2,636 2,509 719 2,636

Any with nonsevere problems only 5,240 1,394 1,975 520 1,351

Rent burden >30–50% of income 4,807 1,320 1,760 489 1,238

Moderately inadequate housing 612 142 194 70 206

Crowded housing 444 (D) 420 15 (D)

[Rent burden only] 4,216 1,252 1,383 436 1,145

Any with no problems 3,454 1,172 813 308 1,161

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

54 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Total Older Adults,

No Children

Families with

Children

Other

Families

Other

Nonfamily

Households

Other characteristics

One person in household 8,845 4,299 NA NA 4,546

Two-spouse household 3,437 776 1,980 681 NA

Female householder 11,54 0 3,584 4,022 933 3,001

Householder of color 10,097 2,391 3,937 1,047 2,722

Welfare/SSI income 3,606 1,208 1,118 290 990

Social Security income 5,308 4,14 3 477 228 460

Income below 50% poverty 4,509 1,284 1,447 280 1,498

Income below poverty 9,882 2,846 3,242 712 3,082

Income below 150% of poverty 14,394 4,204 4,865 1,19 9 4,126

High school graduate 13,959 4,124 4,051 1,219 4,565

2+ years post-high school 4,13 6 1,301 868 361 1,606

Earnings at minimum wage: At least half time 7,9 59 538 3,794 1,063 2,564

Earnings at minimum wage: At least full time 6,250 338 3,207 852 1,853

Earnings main source of income 8,608 487 4,014 1,125 2,982

Housing rated poor 1,098 190 459 99 350

Housing rated good+ 13,387 4,482 3,864 1,207 3,834

Neighborhood rated poor 1,283 198 539 102 444

Neighborhood rated good+ 13,500 4,462 3,926 1,236 3,876

In central cities 8,935 2,440 2,625 844 3,026

Suburbs, urban 5,624 1,844 1,911 510 1,359

Suburbs, rural 1,552 555 436 134 427

Nonmetropolitan 2,275 728 681 161 705

Northeast 3,949 1,363 1,13 8 329 1,119

Midwest 3,715 1,15 4 1,008 263 1,290

South 6,413 1,797 2,089 595 1,932

West 4,311 1,253 1,419 462 1,177

NA = Not applicable. (D) = value suppressed in accord with Census Bureau disclosure prevention requirements.

SSI = Supplemental Security Income.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-6A. Housing Problems and Characteristics of Very Low-Income Renters by Household Type, 2019

(continued)

55WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-6B. Housing Problems and Characteristics of Extremely Low-Income Renters by Household Type, 2019

Total Older Adults,

No Children

Families

with

Children

Other

Families

Other

Nonfamily

Households

Renter households (thousands) 11,749 3,727 3,569 879 3,574

Number of children 7,72 6 NA 7,72 6 NA NA

Number of persons 25,202 4,593 14,026 2,185 4,398

Children/household 2.16 NA 2.16 NA NA

Persons/household 2.15 1.23 3.93 2.49 1.23

Unassisted with severe problems 5,780 1,651 1,741 515 1,873

Unassisted with nonsevere problems only 955 202 463 60 230

Unassisted with no problems 1,064 301 211 90 462

Assisted 3,949 1,572 1,154 215 1,008

Any with severe problems 7,537 2,343 2,295 601 2,298

Rent burden >50% of income 7,373 2,302 2,256 585 2,230

Severely inadequate housing 352 90 99 27 136

[Rent burden only] 6,602 2,075 1,996 521 2,010

Any with nonsevere problems only 2,088 648 834 122 484

Rent burden >30–50% of income 1,790 598 686 108 398

Moderately inadequate housing 337 86 99 27 125

Crowded housing 247 (D) 236 4 (D)

[Rent burden only] 1,523 562 512 91 358

Any with no problems 2,124 736 440 156 792

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

56 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Total Older Adults,

No Children

Families

with

Children

Other

Families

Other

Nonfamily

Households

Other characteristics

One person in household 5,975 2,979 NA NA 2,996

Two-spouse household 1,847 446 1,041 360 NA

Female householder 7,675 2,504 2,694 514 1,963

Householder of color 6,665 1,774 2,528 541 1,822

Welfare/SSI income 2,964 1,047 872 177 868

Social Security income 3,495 2,657 324 141 373

Income below 50% poverty 4,509 1,284 1,447 280 1,498

Income below poverty 9,828 2,835 3,220 712 3,061

Income below 150% of poverty 11,475 3,574 3,560 863 3,478

High school graduate 8,516 2,619 2,438 603 2,856

2+ years post-high school 2,431 805 463 206 957

Earnings at minimum wage: At least half time 3,420 164 1,899 387 970

Earnings at minimum wage: At least full time 1,983 64 1,369 209 341

Earnings main source of income 4,257 191 2,175 477 1,414

Housing rated poor 774 148 291 61 274

Housing rated good+ 8,329 2,934 2,362 604 2,429

Neighborhood rated poor 914 158 393 64 299

Neighborhood rated good+ 8,397 2,902 2,383 644 2,468

In central cities 5,869 1,756 1,687 428 1,998

Suburbs, urban 3,365 1,18 9 1,10 8 261 807

Suburbs, rural 905 318 255 73 259

Nonmetropolitan 1,608 464 518 117 509

Northeast 2,518 996 707 145 670

Midwest 2,309 699 619 130 861

South 4,281 1,215 1,417 347 1,302

West 2,641 817 827 257 74 0

NA = Not applicable. SSI = Supplemental Security Income. (D) = value suppressed in accord with Census Bureau disclosure prevention requirements.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-6B. Housing Problems and Characteristics of Extremely Low-Income Renters by Household Type, 2019

(continued)

57WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-7. Housing Problems and Characteristics of Very Low-Income Worst Case Renters by Household Type, 2019

Total

Older Adults,

No Children

Families

with

Children

Other

Families

Other

Nonfamily

Households

Renter households (thousands) 7,767 2,241 2,271 720 2,535

Number of children 4,524 NA 4,524 NA NA

Number of persons 16,768 2,878 8,735 1,808 3,347

Children/household 1.99 NA 1.99 NA NA

Persons/household 2.16 1.28 3.85 2.51 1.32

Unassisted with severe problems 7,767 2,241 2,271 720 2,535

Unassisted with nonsevere problems only — — — — —

Unassisted with no problems — — — — —

Assisted — — — — —

Any with severe problems 7,767 2,241 2,271 720 2,535

Rent burden >50% of income 7, 56 9 2,18 9 2,222 696 2,462

Severely inadequate housing 374 102 91 40 141

[Rent burden only] 6,802 1,948 1,991 629 2,234

Any with nonsevere problems only — — — — —

Rent burden >30–50% of income — — — — —

Moderately inadequate housing — — — — —

Crowded housing — — — — —

[Rent burden only] — — — — —

Any with no problems — — — — —

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

58 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Total

Older Adults,

No Children

Families

with

Children

Other

Families

Other

Nonfamily

Households

Other characteristics

One person in household 3,672 1,711 NA NA 1,961

Two-spouse household 1,518 340 850 328 NA

Female householder 4,750 1,446 1,564 373 1,367

Householder of color 4,14 3 842 1,589 471 1,241

Welfare/SSI income 1,241 354 426 110 351

Social Security income 2,208 1,773 180 109 146

Income below 50% poverty 2,319 710 715 157 737

Income below poverty 4,893 1,316 1,548 427 1,602

Income below 150% of poverty 6,531 1,827 2,070 607 2,027

High school graduate 6,041 1,767 1,592 516 2,16 6

2+ years post-high school 2,147 616 423 209 899

Earnings at minimum wage: At least half time 3,281 159 1,424 422 1,276

Earnings at minimum wage: At least full time 2,264 85 1,108 292 779

Earnings main source of income 3,890 168 1,600 483 1,639

Housing rated poor 431 76 153 35 167

Housing rated good+ 5,641 1,775 1,572 525 1,769

Neighborhood rated poor 435 78 164 39 154

Neighborhood rated good+ 5,761 1,804 1,619 542 1,796

In central cities 3,904 984 1,035 407 1,478

Suburbs, urban 2,524 771 887 240 626

Suburbs, rural 594 258 122 39 175

Nonmetropolitan 74 4 227 227 34 256

Northeast 1,507 522 426 129 430

Midwest 1,346 413 297 97 539

South 2,797 735 867 265 930

West 2,119 572 681 229 637

NA = Not applicable. SSI = Supplemental Security Income.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-7. Housing Problems and Characteristics of Very Low-Income Worst Case Renters by Household Type, 2019

(continued)

59WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-8. Housing Problems and Characteristics of Extremely Low-Income Worst Case Renters by Household

Type, 2019

Total Older Adults,

No Children

Families

with

Children

Other

Families

Other Nonfamily

Households

Renter households (thousands) 5,780 1,651 1,741 515 1,873

Number of children 3,631 NA 3,631 NA NA

Number of persons 12,734 2,107 6,846 1,308 2,473

Children/household 2.09 NA 2.09 NA NA

Persons/household 2.20 1.28 3.93 2.54 1.32

Unassisted with severe problems 5,780 1,651 1,741 515 1,873

Unassisted with nonsevere problems only — — — — —

Unassisted with no problems — — — — —

Assisted — — — — —

Any with severe problems 3,975 1,132 1,224 398 1,221

Rent burden >50% of income 3,902 1,118 1,209 390 1,18 5

Severely inadequate housing 165 40 40 14 71

[Rent burden only] 3,502 1,022 1,061 351 1,068

Any with nonsevere problems only

— — — — —

Rent burden >30–50% of income — — — — —

Moderately inadequate housing — — — — —

Crowded housing — — — — —

[Rent burden only] — — — — —

Any with no problems

— — — — —

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

60 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Total Older Adults,

No Children

Families

with

Children

Other

Families

Other Nonfamily

Households

Other characteristics

One person in household 2,727 1,264 NA NA 1,463

Two-spouse household 1,059 234 597 228 NA

Female householder 3,577 1,088 1,232 265 992

Householder of color 3,198 699 1,235 343 921

Welfare/SSI income 1,086 312 376 75 323

Social Security income 1,686 1,288 164 96 138

Income below 50% poverty 2,319 710 715 157 737

Income below poverty 4,880 1,316 1,540 427 1,597

Income below 150% of poverty 5,646 1,593 1,737 510 1,806

High school graduate 4,346 1,252 1,164 352 1,578

2+ years post-high school 1,441 423 259 140 619

Earnings at minimum wage: At least half time 1,989 70 965 251 703

Earnings at minimum wage: At least full time 1,044 24 667 130 223

Earnings main source of income 2,641 95 1,15 0 324 1,072

Housing rated poor 357 64 112 34 147

Housing rated good+ 4,070 1,282 1,167 356 1,265

Neighborhood rated poor 357 65 137 31 124

Neighborhood rated good+ 4,16 0 1,295 1,201 383 1,281

In central cities 2,968 766 803 303 1,096

Suburbs, urban 1,790 541 655 161 433

Suburbs, rural 427 175 86 25 141

Nonmetropolitan 595 170 196 26 203

Northeast 1,137 401 326 88 322

Midwest 1,049 301 252 73 423

South 2,15 4 565 694 194 701

West 1,441 385 469 160 427

NA = Not applicable. SSI = Supplemental Security Income.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-8. Housing Problems and Characteristics of Extremely Low-Income Worst Case Renters by Household

Type, 2019 (continued)

61WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-9. Prevalence of Housing Problems Among Very Low-Income Renters by Race and Ethnicity, 2017 and 2019—

Number and Percentage

Number Percentage

2017 2019 2017 2019

Non-Hispanic White (thousands) 7,934 8,290 100.0 100.0

Unassisted with severe problems 3,634 3,623 45.8 43.7

Unassisted with nonsevere problems only 1,461 1,628 18.4 19.6

Unassisted with no problems 911 1,063 11.5 12.8

Assisted 1,927 1,977 24.3 23.8

Any with severe problems 4,291 4,263 54.1 51.4

Rent burden >50% of income 4,18 6 4,158 52.8 50.2

Severely inadequate housing 212 161 2.7 1.9

[Rent burden only] 3,764 3,752 47.4 45.3

Any with nonsevere problems only 2,075 2,346 26.2 28.3

Rent burden >30–50% of income 1,929 2,170 24.3 26.2

Moderately inadequate housing 244 285 3 .1 3.4

Crowded housing 92 127 1.2 1.5

[Rent burden only] 1,74 6 1,940 22.0 23.4

Any with no problems 1,568 1,682 19.8 20.3

Non-Hispanic Black (thousands) 4,561 4,393 100.0 100.0

Unassisted with severe problems 1,578 1,588 34.6 36.1

Unassisted with nonsevere problems only 810 671 17. 8 15.3

Unassisted with no problems 339 373 7.4 8.5

Assisted 1,835 1,761 40.2 4 0.1

Any with severe problems 2,365 2,341 51.9 53.3

Rent burden >50% of income 2,310 2,279 50.6 51.9

Severely inadequate housing 106 134 2.3 3.1

[Rent burden only] 2,091 2,046 45.8 46.6

Any with nonsevere problems only 1,307 1,182 28.7 26.9

Rent burden >30–50% of income 1,182 1,054

25.9 24.0

Moderately inadequate housing 171 174 3.7 4.0

Crowded housing 83 76 1.8 1.7

[Rent burden only] 1,063 942 23.3 21.4

Any with no problems 890 870 19.5 19.8

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

62 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

Hispanic (thousands) 4,083 4,258 100.0 100.0

Unassisted with severe problems 1,884 1,922 46.1 4 5.1

Unassisted with nonsevere problems only 916 1,029 22.4 24.2

Unassisted with no problems 336 382 8.2 9.0

Assisted 947 924 23.2 21.7

Any with severe problems 2,295 2,333 56.2 54.8

Rent burden >50% of income 2,213 2,275 54.2 53.4

Severely inadequate housing 161 133 3.9 3.1

[Rent burden only] 1,965 2,061 4 8.1 48.4

Any with nonsevere problems only 1,246 1,314 30.5 30.9

Rent burden >30–50% of income 1,140 1,208 27. 9 28.4

Moderately inadequate housing 120 103 2.9 2.4

Crowded housing 252 212 6.2 5.0

[Rent burden only] 898 1,008 22.0 23.7

Any with no problems 542 611 13.3 14.3

Source: HUD-PD&R tabulations of the American Housing Survey

Exhibit A-9. Prevalence of Housing Problems Among Very Low-Income Renters by Race and Ethnicity, 2017 and 2019—

Number and Percentage (continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

63WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-10. Prevalence of Housing Problems Among Very Low-Income Renters by Region, 2017 and 2019—Number

and Percentage

Number Percentage

2017 2019 2017 2019

Northeast (thousands) 3,961 3,950 100.0 100.0

Unassisted with severe problems 1,526 1,507 38.5 38.2

Unassisted with nonsevere problems only 683 720 17. 2 18.2

Unassisted with no problems 376 416 9.5 10.5

Assisted 1,377 1,307 34.8 3 3.1

Any with severe problems 2,088 2,029 52.7 51.4

Rent burden >50% of income 1,990 1,958 50.2 49.6

Severely inadequate housing 198 166 5.0 4.2

[Rent burden only] 1,709 1,726 43.1 43.7

Any with nonsevere problems only 1,085 1,141 27.4 28.9

Rent burden >30–50% of income 1,015 1,080 25.6 2 7.3

Moderately inadequate housing 94 112 2.4 2.8

Crowded housing 98 75 2.5 1.9

[Rent burden only] 896 969 22.6 24.5

Any with no problems 789 779 19.9 19.7

Midwest (thousands) 3,670 3,715 100.0 100.0

Unassisted with severe problems 1,378 1,346 37.5 36.2

Unassisted with nonsevere problems only 767 877 20.9 23.6

Unassisted with no problems 400 387 10.9 10.4

Assisted 1,12 6 1,10 6 30.7 29.8

Any with severe problems 1,757 1,690 47.9 45.5

Rent burden >50% of income 1,697 1,656 46.2 44.6

Severely inadequate housing 93 66 2.5 1.8

[Rent burden only] 1,552 1,521 42.3 40.9

Any with nonsevere problems only 1,118 1,239 30.5 33.4

Rent burden >30–50% of income 1,048 1,145

28.6 30.8

Moderately inadequate housing 127 124 3.5 3.3

Crowded housing 65 81 1.8 2.2

[Rent burden only] 934 1,042 25.4 28.0

Any with no problems 795 786 21.7 21.2

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

64 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

South (thousands) 6,358 6,413 100.0 100.0

Unassisted with severe problems 2,844 2,796 44.7 43.6

Unassisted with nonsevere problems only 1,278 1,246 20 .1 19.4

Unassisted with no problems 643 780 10.1 12.2

Assisted 1,593 1,592 25 .1 24.8

Any with severe problems 3,475 3,428 54.7 53.5

Rent burden >50% of income 3,438 3,331 5 4.1 51.9

Severely inadequate housing 115 152 1.8 2.4

[Rent burden only] 3,112 2,978 48.9 46.4

Any with nonsevere problems only 1,767 1,727 27.8 26.9

Rent burden >30–50% of income 1,590 1,549 25.0 24.2

Moderately inadequate housing 261 270 4.1 4.2

Crowded housing 137 150 2.2 2.3

[Rent burden only] 1,386 1,310 21.8 20.4

Any with no problems 1,116 1,258 17.6 19.6

West (thousands) 4,078 4,310 100.0 100.0

Unassisted with severe problems 1,968 2,118 48.3 49.1

Unassisted with nonsevere problems only 702 754 17.2 17. 5

Unassisted with no problems 313 390 7.7 9.0

Assisted 1,095 1,048 26.9 24.3

Any with severe problems 2,453 2,548 60.2 59.1

Rent burden >50% of income 2,373 2,491 58.2 57. 8

Severely inadequate housing 128 100 3 .1 2.3

[Rent burden only] 2,16 9 2,275 53.2 52.8

Any with nonsevere problems only 1,047 1,133 25.7 26.3

Rent burden >30–50% of income 951 1,033 23.3 24.0

Moderately inadequate housing 95 105 2.3 2.4

Crowded housing 197 137 4.8 3.2

[Rent burden only] 770 894 18.9 20.7

Any with no problems 577 630 14.1 14.6

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-10. Prevalence of Housing Problems Among Very Low-Income Renters by Region, 2017 and 2019—Number and

Percentage (continued)

65WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-11A. Prevalence of Housing Problems Among Very Low-Income Renters by Metropolitan Location,

2017 and 2019—Number and Percentage

Number Percentage

2017 2019 2017 2019

Central cities (thousands) 8,995 8,936 100.0 100.0

Unassisted with severe problems 3,816 3,904 42.4 43.7

Unassisted with nonsevere problems only 1,613 1,629 17.9 18.2

Unassisted with no problems 783 850 8.7 9.5

Assisted 2,783 2,553 30.9 28.6

Any with severe problems 4,989 4,996 55.5 55.9

Rent burden >50% of income 4,828 4,867 53.7 54.5

Severely inadequate housing 313 273 3.5 3 .1

[Rent burden only] 4,302 4,347 47. 8 48.6

Any with nonsevere problems only 2,421 2,386 26.9 26.7

Rent burden >30–50% of income 2,226 2,172 24.7 24.3

Moderately inadequate housing 261 296 2.9 3.3

Crowded housing 284 240 3.2 2.7

[Rent burden only] 1,902 1,873 21.1 21.0

Any with no problems 1,585 1,554 17.6 17.4

Suburbs, urban (thousands) 5,287 5,625 100.0 100.0

Unassisted with severe problems 2,472 2,524 46.8 44.9

Unassisted with nonsevere problems only 1,059 1,213 20.0 21.6

Unassisted with no problems 468 511 8.9 9.1

Assisted 1,287 1,377 24.3 24.5

Any with severe problems 2,976 3,048 56.3 54.2

Rent burden >50% of income 2,921 3,023 55.2 53.7

Severely inadequate housing 98 77 1.9 1.4

[Rent burden only] 2,691 2,768 50.9 49.2

Any with nonsevere problems only 1,460 1,684 27.6 29.9

Rent burden >30–50% of income 1,374 1,565

26.0 2 7.8

Moderately inadequate housing 135 163 2.6 2.9

Crowded housing 142 123 2.7 2.2

[Rent burden only] 1,19 2 1,403 22.5 24.9

Any with no problems 851 893 16.1 15.9

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

66 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

Suburbs, rural (thousands) 1,360 1,552 100.0 100.0

Unassisted with severe problems 611 594 44.9 38.3

Unassisted with nonsevere problems only 255 327 18.8 21.1

Unassisted with no problems 178 288 13 .1 18.6

Assisted 317 343 23.3 22.1

Any with severe problems 733 716 53.9 46.1

Rent burden >50% of income 710 675 52.2 43.5

Severely inadequate housing 56 51 4.1 3.3

[Rent burden only] 623 615 45.8 39.6

Any with nonsevere problems only 352 452 25.9 29.1

Rent burden >30–50% of income 311 416 22.9 26.8

Moderately inadequate housing 55 44 4.0 2.8

Crowded housing 29 33 2.1 2.1

[Rent burden only] 272 375 20.0 24.2

Any with no problems 275 384 20.2 24.7

Nonmetropolitan (thousands) 2,425 2,276 100.0 100.0

Unassisted with severe problems 816 74 4 33.6 32.7

Unassisted with nonsevere problems only 503 427 20.7 18.8

Unassisted with no problems 302 324 12.5 14.2

Assisted 803 781 3 3.1 34.3

Any with severe problems 1,074 935 44.3 41.1

Rent burden >50% of income 1,039 870 42.8 38.2

Severely inadequate housing 66 82 2.7 3.6

[Rent burden only] 925 770 38 .1 33.8

Any with nonsevere problems only 784 718 32.3 31.5

Rent burden >30–50% of income 694 654 28.6 28.7

Moderately inadequate housing 127 107 5.2 4.7

Crowded housing 42 47 1.7 2.1

[Rent burden only] 619 564 25.5 24.8

Any with no problems 566 623 23.3 27.4

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-11A. Prevalence of Housing Problems Among Very Low-Income Renters by Metropolitan

Location, 2017 and 2019—Number and Percentage (continued)

(continued)

67WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Number Percentage

2017 2019 2017 2019

U.S. Total (thousands) 18,067 18,388 100.0 100.0

Unassisted with severe problems 7,715 7,767 42.7 42.2

Unassisted with nonsevere problems only 3,430 3,596 19.0 19.6

Unassisted with no problems 1,731 1,972 9.6 10.7

Assisted 5,19 0 5,053 28.7 27. 5

Any with severe problems 9,772 9,694 54.1 52.7

Rent burden >50% of income 9,498 9,435 52.6 51.3

Severely inadequate housing 533 484 3.0 2.6

[Rent burden only] 8,541 8,500 47. 3 46.2

Any with nonsevere problems only 5,017 5,240 27.8 28.5

Rent burden >30–50% of income 4,605 4,807 25.5 26 .1

Moderately inadequate housing 578 612 3.2 3.3

Crowded housing 497 444 2.8 2.4

[Rent burden only] 3,985 4,216 22.1 22.9

Any with no problems 3,277 3,454 18.1 18.8

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-11A. Prevalence of Housing Problems Among Very Low-Income Renters by Metropolitan

Location, 2017 and 2019—Number and Percentage (continued)

68 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan Areas,

2017 and 2019

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

Atlanta-Sandy Springs-Roswell, GA

Total households (thousands) 245 794 263 795

Unassisted with severe problems 131 132

Unassisted with nonsevere problems only 39 56

Unassisted with no problems 22 25

Assisted 53 50

Boston-Cambridge-Newton, MA-NH

Total households (thousands) 313 715 347 754

Unassisted with severe problems 100 99

Unassisted with nonsevere problems only 28 53

Unassisted with no problems 29 42

Assisted 156 153

Chicago-Naperville-Elgin, IL-IN-WI

Total households (thousands) 509 1,238 476 1,237

Unassisted with severe problems 204 160

Unassisted with nonsevere problems only 92 118

Unassisted with no problems 58 62

Assisted 155 136

Dallas-Fort Worth-Arlington, TX

Total households (thousands) 332 1,060 364 1,076

Unassisted with severe problems 159 174

Unassisted with nonsevere problems only 83 79

Unassisted with no problems 35 54

Assisted 54 56

Detroit-Warren-Dearborn, MI

Total households (thousands) 243 527 235 494

Unassisted with severe problems 105 111

Unassisted with nonsevere problems only 49 38

Unassisted with no problems 23 26

Assisted 65 60

Houston-The Woodlands-Sugar Land, TX

Total households (thousands) 362 897 337 898

Unassisted with severe problems 177 179

Unassisted with nonsevere problems only 91 85

Unassisted with no problems 43 29

Assisted 51 44

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

69WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

Los Angeles-Long Beach-Anaheim, CA

Total households (thousands) 968 2,281 976 2,310

Unassisted with severe problems 459 508

Unassisted with nonsevere problems only 210 200

Unassisted with no problems 86 88

Assisted 213 180

Miami-Fort Lauderdale-West Palm Beach, FL

Total households (thousands) 384 854 356 874

Unassisted with severe problems 211 177

Unassisted with nonsevere problems only 54 60

Unassisted with no problems 39 36

Assisted 80 82

New York-Newark-Jersey City, NY-NJ-PA

Total households (thousands) 1,712 3,644 1,769 3,732

Unassisted with severe problems 678 724

Unassisted with nonsevere problems only 269 261

Unassisted with no problems 168 215

Assisted 596 569

Philadelphia-Camden-Wilmington, PA-NJ-DE-MD

Total households (thousands) 336 734 306 718

Unassisted with severe problems 147 125

Unassisted with nonsevere problems only 61 53

Unassisted with no problems 29 42

Assisted 98 86

Phoenix-Mesa-Scottsdale, AZ

Total households (thousands) 189 593 185 607

Unassisted with severe problems 97 99

Unassisted with nonsevere problems only 37 26

Unassisted with no problems 20 20

Assisted 35 40

Riverside-San Bernardino-Ontario, CA

Total households (thousands) 158 479 149 478

Unassisted with severe problems 91 98

Unassisted with nonsevere problems only 19 10

Unassisted with no problems 19 12

Assisted 30 29

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan

Areas, 2017 and 2019 (continued)

70 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

San Francisco-Oakland-Hayward, CA

Total households (thousands) 274 774 291 778

Unassisted with severe problems 110 125

Unassisted with nonsevere problems only 48 43

Unassisted with no problems 26 36

Assisted 89 87

Seattle-Tacoma-Bellevue, WA

Total households (thousands) 202 614 193 596

Unassisted with severe problems 84 81

Unassisted with nonsevere problems only 32 24

Unassisted with no problems 21 21

Assisted 65 68

Washington-Arlington-Alexandria, DC-VA-MD-WV

Total households (thousands) 292 802 333 853

Unassisted with severe problems 126 123

Unassisted with nonsevere problems only 58 78

Unassisted with no problems 25 40

Assisted 82 92

Baltimore-Columbia-Towson, MD

Total households (thousands) 291 673 — —

Unassisted with severe problems 124 —

Unassisted with nonsevere problems only 57 —

Unassisted with no problems 23 —

Assisted 87 —

Birmingham-Hoover, AL

Total households (thousands) 120 279 — —

Unassisted with severe problems 52 —

Unassisted with nonsevere problems only 15 —

Unassisted with no problems 6 —

Assisted 47 —

Cincinnati, OH-KY-IN

Total households (thousands) — — — —

Unassisted with severe problems

— —

Unassisted with nonsevere problems only

— —

Unassisted with no problems

— —

Assisted

— —

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan

Areas, 2017 and 2019 (continued)

71WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

Cleveland-Elyria, OH

Total households (thousands)

— — — —

Unassisted with severe problems

— —

Unassisted with nonsevere problems only

— —

Unassisted with no problems

— —

Assisted

— —

Denver-Aurora-Lakewood, CO

Total households (thousands)

— — — —

Unassisted with severe problems

— —

Unassisted with nonsevere problems only

— —

Unassisted with no problems

— —

Assisted

— —

Kansas City, MO-KS

Total households (thousands)

— — — —

Unassisted with severe problems

— —

Unassisted with nonsevere problems only

— —

Unassisted with no problems

— —

Assisted

— —

Las Vegas-Henderson-Paradise, NV

Total households (thousands) 242 784 — —

Unassisted with severe problems 132 —

Unassisted with nonsevere problems only 50 —

Unassisted with no problems 21 —

Assisted 39 —

Memphis, TN-MS-AR

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

Milwaukee-Waukesha-West Allis, WI

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

(continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan

Areas, 2017 and 2019 (continued)

72 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

Minneapolis-St. Paul-Bloomington, MN-WI

Total households (thousands) 392 866 — —

Unassisted with severe problems 138 —

Unassisted with nonsevere problems only 79 —

Unassisted with no problems 28 —

Assisted 146 —

New Orleans-Metairie, LA

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

Oklahoma City, OK

Total households (thousands) 131 331 — —

Unassisted with severe problems 53 —

Unassisted with nonsevere problems only 35 —

Unassisted with no problems 19 —

Assisted 23 —

Pittsburgh, PA

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

Portland-Vancouver-Hillsboro, OR-WA

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

Raleigh, NC

Total households (thousands) — — — —

Unassisted with severe problems — —

Unassisted with nonsevere problems only — —

Unassisted with no problems — —

Assisted — —

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan

Areas, 2017 and 2019 (continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

73WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Household Income as Percentage of HUD-Adjusted Area Median Family Income

2017 2019

0–50% All Incomes 0–50% All Incomes

Richmond, VA

Total households (thousands) 131 303 — —

Unassisted with severe problems 55 —

Unassisted with nonsevere problems only 23 —

Unassisted with no problems 24 —

Assisted 29 —

Rochester, NY

Total households (thousands) 142 280 — —

Unassisted with severe problems 50 —

Unassisted with nonsevere problems only 15 —

Unassisted with no problems 5 —

Assisted 72 —

San Antonio-New Braunfels, TX

Total households (thousands) 207 631 — —

Unassisted with severe problems 105 —

Unassisted with nonsevere problems only 46 —

Unassisted with no problems 9 —

Assisted 47 —

San Jose-Sunnyvale-Santa Clara, CA

Total households (thousands) 182 647 — —

Unassisted with severe problems 85 —

Unassisted with nonsevere problems only 33 —

Unassisted with no problems 21 —

Assisted 44 —

Tampa-St. Petersburg-Clearwater, FL

Total households (thousands) 278 825 — —

Unassisted with severe problems 165 —

Unassisted with nonsevere problems only 43 —

Unassisted with no problems 27 —

Assisted 44 —

Notes: Each of the 15 largest metropolitan areas, listed first, are part of the American Housing Survey longitudinal panel surveyed every 2 years. The remaining 10

metropolitan areas represent a subset of the 16th to 50th largest metropolitan areas surveyed on a rotating basis every 4 years.

Source: HUD-PD&R tabulations of the American Housing Survey

Exhibit A-11B. Housing Conditions of Renter Households by Relative Income, Sampled Metropolitan

Areas, 2017 and 2019 (continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

74 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-12. Households Occupying Rental Units by Affordability of Rent and Income of Occupants, 2017 and 2019

Relative Income of

Households

Occupied and Vacant Rental Units (thousands) by Unit Affordability Category (percent of HAMFI needed to

afford the highest rent in the category)

2019 10

20 30 40 50 60 70 80 90 100 110 >110 Total

Extremely low income

(<30% HAMFI)

735 1,647 2,024 922 1,810 1,654 818 904 342 144 231 517 11,74 8

Very low income

(30 –50%)

185 369 542 521 1,471 1,288 657 776 212 105 196 317 6,639

Low income (50–80%) 192 331 485 499 1,479 1,578 1,19 5 1,322 507 291 395 512 8,786

Middle income or

higher (>80%)

299 475 646 388 1,422 2,325 2,030 2,736 1,775 1,308 1,724 2,359 17,487

Vacant units for rent 91 81 154 246 635 757 490 583 325 294 372 647 4,675

Total units vacant and

occupied

1,502 2,903 3,851 2,576 6,817 7,6 02 5,19 0 6,321 3 ,161 2,142 2,918 4,352 49,335

2017 10* 20 30 40 50 60 70 80 90 100 110 120+ Total

Extremely low income

(<30% HAMFI)

801 1,583 1,839 703 1,782 1,822 890 955 321

165 205 483 11,54 8

Very low income

(30 –50%)

145 326 511 505 1,242 1,304 700 820 273 123 218 351 6,519

Low income (50–80%) 243 280 476 364 1,315 1,698 1,288 1,246 531 301 305 590 8,637

Middle income or

higher (>80%)

304 478 624 330 1,311 2,193 2,200 2,604 1,751 1,381 1,678 2,433 17,289

Vacant units for rent 122 89 160 236 616 770 579 603 371 261 355 664 4,827

Total units vacant and

occupied

1,616 2,757 3,610 2,13 8 6,265 7,78 8 5,657 6,228 3,246 2,231 2,761 4,522 48,820

HAMFI = HUD-Adjusted Median Family Income.

* The 10-percent-of-HAMFI category includes units occupied with no cash rent.

Notes: The method of assigning units to cost categories was modified in 2017 to account for limited HUD administrative exceptions to program income limits.

Slight unit affordability adjustments were applied to outlier cases for which Area Median Income-determined affordability differed from administratively determined

affordability categories.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

75WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-13. Renters and Rental Units Affordable and Available to Them by Relative Income, 2001–2019

2001 2003 2005 2007 2009 2011 2013 2015 2017 2019

Renter households (thousands) 34,042 33,614 33,951 35,054 35,396 38,867 40,294 43,930 43,993 44,660

Extremely low-income (<30%

HAMFI)

8,739 9,077 9,729 9,243 9,961 11,7 74 11,16 3 11,2 90 11,54 8 11,748

Very low-income (30–50%) 6,315 6,581 6,342 6,697 7,157 7,4 92 7,375 7,9 45 6,519 6,640

Low-income (50–80%) 7, 251 7,4 6 0 7,4 8 8 7,65 0 7,16 8 7,75 0 7,79 5 8,696 8,637 8,786

Middle-income or higher (>80%) 11,737 10,496 10,392 11,46 4 11,110 11,85 0 13,961 15,999 17,289 17,4 8 6

Affordable units 37,197 37,577 37,924 39,330 39,744 43,075 43,992 48,670 48,820 49,335

Extremely low-income (<30%

HAMFI)

6,870 7,0 98 6,747 7, 28 0 6,265 6,854 7, 29 4 7,117 7,9 82 8,256

Very low-income (30–50%) 12,366 12,863 12,368 11,071 10,938 10,947 10,727 9,643 8,404 9,393

Low-income (50–80%) 13,634 13,518 14,044 15,063 16,228 17, 99 5 17,9 0 4 19,326 19,674 19,112

Middle-income or higher (>80%) 4,328 4,099 4,765 5,916 6,313 7, 279 8,067 12,584 12,760 12,574

Affordable and available units 37,197 37,577 37,924 39,330 39,744 43,075 43,992 48,670 48,820 49,335

Extremely low-income (<30%

HAMFI)

3,803 3,996 3,982 4,224 3,665 4,220 4,354 4,278 4,595 4,732

Very low-income (30–50%) 8,132 8,744 8,549 7,78 6 8,045 8,225 7,73 4 7, 576 6,066 6,700

Low-income (50–80%) 11,66 5 12,396 12,865 13 ,19 6 14,004 15,361 14,529 15,862 15,353 15,009

Middle-income or higher (>80%) 13,597 12,441 12,528 14,12 3 14,029 15,270 17,375 20,955 22,806 22,894

HAMFI = HUD-Adjusted Median Family Income.

Notes: Income categories in this exhibit do not overlap and therefore differ from the standard definitions. The method of assigning units to cost categories was

modified in 2017 to account for limited HUD administrative exceptions to program income limits. Slight unit affordability adjustments were applied to outlier cases for

which Area Median Income-determined affordability differed from administratively determined affordability categories.

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

76 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-14. Average Income and Average Gross Rent of Renter Households by Relative Income, 2017 and 2019

Household Income as Percentage of HUD-Adjusted Area Median Family

Income

2019 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 11,748 6,640 8,786 7,583 9,902 44,660

Unassisted with severe problems 5,780 1,986 1,013 372 200 9,352

Unassisted with nonsevere problems only 955 2,642 3,805 1,737 1,086 10,225

Unassisted with no problems 1,064 909 3,378 5,206 8,363 18,919

Assisted 3,950 1,103 590 268 253 6 ,16 4

Average monthly income $887 $2,359 $3,636 $5,235 $12,100 $4,871

Unassisted with severe problems $904 $2,275 $3,502 $4,527 $10,500 $1,826

Unassisted with nonsevere problems only $1,414 $2,508 $3,674 $ 5,18 4 $10,520 $ 4,146

Unassisted with no problems $557 $2,397 $3,651 $5,334 $12,380 $7,73 9

Assisted $823 $2,118 $3,536 $4,625 $10,860 $1,891

Average gross rent $887 $1,078 $1,218 $1,393 $1,731 $1,254

Unassisted with severe problems $1,216 $1,669 $2,611 $4,071 $5,558 $1,670

Unassisted with nonsevere problems only $648 $981 $1,299 $1,695 $2,293 $1,329

Unassisted with no problems $536 $462 $766 $1,103 $1,578 $1,19 0

Assisted $559 $755 $898 $1,347 $1,355 $693

2017 0–30% >30–50% >50–80% >80–120% >120% All Incomes

Total households (thousands) 11,548 6,519 8,637 7,306 9,983 43,993

Unassisted with severe problems 5,555 2,161 973 277 232 9,198

Unassisted with nonsevere problems only 1,049 2,381 3,804 1,768 1,180 10 ,181

Unassisted with no problems 908 823 3,220 5,003 8,272 18,226

Assisted 4,037 1,15 4 641 259 298 6,388

Average monthly income $850 $2,136 $3,306 $4,726 $10,300 $4,310

Unassisted with severe problems $868 $2,067 $ 3,16 2 $3,383 $11,06 0 $1,725

Unassisted with nonsevere problems only $1,355 $2,273 $3,343 $4,716 $9,047 $3,787

Unassisted with no problems $533 $ 2,14 4 $3,335 $4,818 $10,510 $6,805

Assisted $765 $1,974 $3,170 $4,455 $8,791 $1,749

Average gross rent $790 $1,010 $1,109 $1,281 $1,594 $1,149

Unassisted with severe problems $1,101 $1,491 $2,358 $4,464 $5,344 $1,534

Unassisted with nonsevere problems only $569 $887 $1,16 9 $1,622 $2,087 $1,226

Unassisted with no problems $492 $399 $708 $1,000 $1,434 $1,093

Assisted $487 $800 $869 $965 $1,179 $633

Source: HUD-PD&R tabulations of the American Housing Survey

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

77WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Exhibit A-15. Housing Conditions of Households Having People Younger than 62 Who Have Disabilities by Disability

Type, 2017 and 2019

Functional Limitations ADL/IADL Limitations

2019

Any

Limitation

Hearing Visual Cognitive Ambulatory Self-Care

Independent

Living

Households (thousands) 10,947 2,927 2,238 4,533 4,709 1,570 3,029

Renter households 5,092 1,13 3 1,13 0 2,337 2,377 733 1,562

Owner households 5,855 1,794 1,10 8 2,19 6 2,332 836 1,466

Renters (thousands) 5,092 1,133 1,130 2,337 2,377 733 1,562

Unassisted with severe

problems 1,145 243 241 527 549 202 353

Unassisted with nonsevere

problems only 1,208 288 331 539 504 147 320

Unassisted with no problems 1,501 365 314 590 601 145 342

Assisted 1,237 237 245 681 723 239 547

Very low-income renters

(thousands)

1,100 207 252 538 475 156 305

Unassisted with severe

problems 212 40 40 93 89 34 56

Unassisted with nonsevere

problems only 483 89 136 231 202 68 118

Unassisted with no problems 174 38 36 77 74 20 39

Assisted 231 40 40 136 110 33 93

Any with severe problems 247 45 49 115 111 43 73

Rent burden >50% of income 221 38 46

106 102 43 73

Severely inadequate housing 29 10 (D) 9 9 (D) (D)

[Rent burden only] 206 33 44 104 94 37 66

Any with nonsevere problems

only

580 109 153 285 248 84 154

Rent burden >30–50% of

income 523 89 135 248 228 75 147

Moderately inadequate housing 97 28 25 52 43 10 23

Crowded housing 39 (D) 17 30 (D) (D) (D)

[Rent burden only] 446 78 112 204 200 66 126

Any with no problems 273 53 49 138 116 29 77

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

(continued)

78 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Functional Limitations ADL/IADL Limitations

2017

Any

Limitation

Hearing Visual Cognitive Ambulatory Self-Care

Independent

Living

Households (thousands) 12,360 3,071 2,558 5,202 5,423 1,885 3,633

Renter households 5,750 1,146 1,300 2,710 2,669 923 1,819

Owner households 6,613 1,925 1,258 2,492 2,754 962 1,815

Renters (thousands) 5,750 1,146 1,300 2,710 2,669 923 1,819

Unassisted with severe

problems 1,442 234 334 735 721 280 524

Unassisted with nonsevere

problems only 1,293 284 325 562 542 159 325

Unassisted with no problems 1,604 378 343 653 580 184 377

Assisted 1,412 250 298 760 827 299 594

Very low-income renters

(thousands)

1,260 215 269 560 562 199 385

Unassisted with severe

problems 344 71 56 180 162 63 130

Unassisted with nonsevere

problems only 493 87 157 179 192 59 120

Unassisted with no problems 192 24 24 92 82 36 37

Assisted 231 33 33 109 126 41 98

Any with severe problems

377 81 63 194 185 68 146

Rent burden >50% of income 332 72 53 162 156 62 129

Severely inadequate housing 45 8 11 32 28 6 17

[Rent burden only, adequate

housing] 299 71 46 149 134 56 110

Any with nonsevere problems

only

601 101 175 239 248 84 165

Rent burden >30–50% of

income 551 93 156 220 234 80 155

Moderately inadequate housing 90 8 17 39 36 18 22

Crowded housing 57 12 24 23 17 11 17

[Rent burden only] 459 84 136 178 199 57 128

Any with no problems

282 33 32 127 130 46 74

ADL = Activities of Daily Living. IADL = Instrumental Activities of Daily Living

(D) = value suppressed in accord with Census Bureau disclosure prevention requirements.

Source: HUD-PD&R tabulations of the American Housing Survey

Exhibit A-15. Housing Conditions of Households Having People Younger than 62 Who Have Disabilities by Disability

Type, 2017 and 2019 (continued)

APPENDIX A. DETAILED DATA ON HOUSING PROBLEMS AND SUPPLY OF AFFORDABLE HOUSING

79WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX B

Supplemental Exhibits

Exhibit B-1. Bird’s-Eye View of Worst Case

Housing Needs in 2019

0.20

million

Worst case needs,

7.77 million

Renters with severely

inadequate housing,

0.82 million

Unassisted very

low-income renters

13.34 million

Renters with

severe

rent burden,

10.74 million

Other renters,

31.32 million

0.18

million

0.38

million

0.06

million

7.39

million

3.11

million

Note: Not to scale.

Source: HUD-PD&R tabulations of American Housing Survey data

APPENDIX B. SUPPLEMENTAL EXHIBITS

80 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

10 20 30 40

Source: HUD-PD&R tabulations of American Housing Survey data

Rental units (millions)

Unit affordability: percent of Area Median Income needed to afford the

highest rent in the category

50 60 70 80 90 100 110 >110

0.19

0.37

0.33

0.15

0.54

0.49

0.25

0.52

0.50

0.64

1.47

1.48

0.76

1.29

1.58

0.49

0.66

1.20

0.58

0.78

1.32

0.33

0.51

0.65

0.51

0.74

0.30

0.09 0.08

1.65

0.48

0.65

0.39

1.42

2.33

2.03

2.74

1.78

1.31

1.72

2.36

2.02

0.92

1.81

1.65

0.82

0.90

0.52

Renters with incomes of >80% AMI

Renters with incomes of 50–80% AMI

Renters with incomes of 30–50% AMI

Renters with incomes of 0–30% AMI

Vacant units for rent

0.21

0.34

0.32

0.29

0.37

0.11

0.14

0.29

0.23

0.20

0.40

Exhibit B-2. Affordable Rental Units

Occupied by Higher Income Renters, 2019

AMI = area median income.

Source: HUD PD&R tabulations of American Housing Survey data

Exhibit B-3. Rental Stock of Below-Fair Market Rent Units by Region and Metropolitan Location, 2019

Households

(thousands)

Affordable

Housing

Units

(thousands)

Affordable

and

Available

Housing

Units

(thousands)

Affordable,

Available, and

Adequate

Housing Units

(thousands)

Affordable

Housing

Units

per 100

Households

Affordable and

Available

Housing Units

per 100

Households

Affordable,

Available, and

Adequate

Housing Units

per 100

Households

All 26,521 29,925 21,657 19,880 112.8 81.7 75.0

Northeast 5,487 5,766 4,444 4,003 10 5.1 81.0 73.0

Midwest 4,504 5,576 3,716 3,462 123.8 82.5 76.9

South 9,704 11,384 8,298 7,577 117. 3 85.5 78 .1

West 6,826 7, 2 00 5,200 4,838 105.5 76.2 70.9

Central cities 12,669 13,360 10,053 9,118 105.5 79.4 72.0

Suburbs, urban 8,824 9,877 7, 02 6 6,590 111.9 79.6 74.7

Suburbs, rural 2,166 2,798 2,034 1,884 129.2 93.9 87.0

Nonmetropolitan

areas

2,862 3,890 2,544 2,287 135.9 88.9 79.9

Source: HUD PD&R tabulations of American Housing Survey data

APPENDIX B. SUPPLEMENTAL EXHIBITS

81WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX C

Federal Housing

Assistance and

Affordable Housing

Programs

HUD provides rental housing assistance through three key programs.

1. Public housing. This program provides affordable housing to

approximately 1.0 million households through units owned and managed by

local public housing agencies. Families are required to pay 30 percent of

their incomes for rent.

2. Project-based assisted housing. This program provides assistance to 1.2

million households living in privately owned rental housing. The assistance

is attached to the units reserved for low-income families who are required to

pay 30 percent of their incomes for rent.

3. Tenant-based rental assistance. The Housing Choice Voucher Program

supplements the rent payments of 2.3 million households in the private

rental market. The program is administered through state and local housing

agencies. Although 30 percent of income is the rent baseline, families

often pay more and use these portable subsidies to locate housing of their

choice.

Several other federal housing programs produce affordable housing, typically

with shallower subsidies. Although these units are often more affordable than

market-rate units, without additional rent subsidies (such as vouchers), extremely

low-income families would often have to pay much more than 30 percent of their

incomes under these programs.

The number of households assisted by key programs based on HUD administrative records are

available through the Picture of Subsidized Households query tool at https://www.huduser.gov/

portal/datasets/assthsg.html.

APPENDIX C. FEDERAL HOUSING ASSISTANCE AND AFFORDABLE HOUSING PROGRAMS

82 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

- Low-Income Housing Tax Credit (LIHTC) Program. Ta x

credits offered to investors by the U.S. Department of the

Treasury subsidize the capital costs of units that have rents

affordable to households with incomes not exceeding 60

percent of area median income.

- HOME Investment Partnerships Program. This

program provides annual formula grants to state and local

governments that can be used to assist homeowners,

first-time homebuyers, or renters. Qualifying rents must

be affordable to households with incomes not exceeding

65 percent of AMI or must be less than the local fair

market rent (FMR), whichever is less.

- Housing Opportunities for Persons With AIDS

(HOPWA). HOPWA provides annual formula and

competitive grants available to state and local

governments and nonprofits for rental assistance targeted

to a special-needs population.

- Older rental subsidy programs. Programs named for

sections of the National Housing Act, primarily the Section

221(d)(3) Below Market Interest Rate Program and the

Section 236 mortgage assistance program, were active

from the early 1960s through the early 1970s. They were

designed to produce affordable housing for families with

incomes higher than the public housing income limits.

For further detail on HUD program requirements, see HUD

(2018).

APPENDIX C. FEDERAL HOUSING ASSISTANCE AND AFFORDABLE HOUSING PROGRAMS

83WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX D. PREVIOUS REPORTS TO CONGRESS ON WORST CASE NEEDS

APPENDIX D

Previous Reports to

Congress on Worst

Case Needs

- Priority Problems and “Worst Case” Needs in 1989 (June 1991, HUD-1314-

PDR).

- The Location of Worst Case Needs in the Late 1980s (December 1992,

HUD-1387-PDR).

- Worst Case Needs for Housing Assistance in the United States in 1990 and

1991 (June 1994, HUD-1481-PDR).

- Rental Housing Assistance at a Crossroads: A Report to Congress on Worst

Case Housing Needs (March 1996).

- Rental Housing Assistance—The Crisis Continues (April 1998).

- Rental Housing Assistance—The Worsening Crisis: A Report to Congress

on Worst Case Housing Needs (March 2000).

- A Report on Worst Case Housing Needs in 1999: New Opportunity Amid

Continuing Challenges, Executive Summary (January 2001).

- Trends in Worst Case Needs for Housing, 1978–1999 (December 2003).

- Affordable Housing Needs: A Report to Congress on the Significant Need

for Housing (December 2005).

- Affordable Housing Needs 2005: Report to Congress (May 2007).

- Housing Needs of Persons With Disabilities: Supplemental Findings to the

Affordable Housing Needs 2005 Report (February 2008).

- Worst Case Housing Needs 2007: A Report to Congress (May 2010).

- Worst Case Housing Needs 2009: Report to Congress (February 2011).

84 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

- Worst Case Housing Needs 2011: Report to Congress

(August 2013).

- Worst Case Housing Needs: 2015 Report to Congress

(April 2015).

- Worst Case Housing Needs: 2017 Report to Congress

(August 2017).

- Worst Case Housing Needs: 2019 Report to Congress

(June 2020).

These publications are available online at http://www.

huduser.gov.

APPENDIX D. PREVIOUS REPORTS TO CONGRESS ON WORST CASE NEEDS

85WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

APPENDIX E

Data and Methodology

A report such as this one requires researchers to use a number of specialized

concepts, definitions, and assumptions when analyzing and presenting the data.

This appendix documents such elements for those who wish to understand the

results more fully or replicate and extend the results in their own research.

Using the American Housing Survey Data

This report uses data from the most recently available American Housing Survey

(AHS), conducted in 2019. The AHS, which is the only detailed periodic national

housing survey in the United States, is sponsored by HUD and conducted by

the Census Bureau. It provides nationally representative data on a wide range

of housing subjects, including apartments, single-family homes, mobile homes,

vacant homes, family composition, income, housing and neighborhood quality,

housing costs, equipment, fuel type, size of housing units, and recent moves.

The AHS collects national data every 2 years, originally from a sample of about

84,400 housing units (Census-HUD, 2013) and currently from a new, redesigned

sample of about 85,400 housing units begun in 2015 (Census-HUD, 2017).

The survey, which started in 1973, sampled the same housing units between

1985 and 2013—with occasional adjustments and supplements—plus samples

of newly constructed units to ensure the data’s continuity and timeliness. To

address many challenges in maintaining the AHS longitudinal sample for nearly

30 years, including attrition of housing units, response burden, changes in

geography, and disclosure avoidance and mitigation, HUD and the Census

Bureau undertook a major redesign for the 2015 AHS. The redesign included a

selection of a new national and metropolitan area longitudinal sample, changes

to weighting methodologies and imputation processes, and a reevaluation of

variables. Information from the worst case needs reports has helped inform

public policy decisions, including decisions on targeting existing resources,

determining the need for additional resources, and choosing the form that

housing assistance should take.

To accurately estimate worst case needs for federal rental assistance from AHS

data, it is essential to determine whether household incomes fall below HUD’s

official very low-income limits (50 percent of HUD-Adjusted Median Family

Income [HAMFI], also termed area median income [AMI]), whether a household

already receives housing assistance, and whether an unassisted income-

An online codebook that documents all variables available in all American Housing Survey years

is available at https://www.census.gov/data-tools/demo/codebook/ahs/ahsdict.html.

86 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

eligible household has one or more of the priority problems

that formerly conferred preference in tenant selection for

assistance (rent burdens exceeding 50 percent of income,

substandard housing, or being involuntarily displaced).

HUD and the Census Bureau provide a Table Creator for the

2011 to 2019 AHS surveys. The Table Creator enables users

to create customized tabulations of AHS data without the

difficulties and special skills needed to analyze microdata

Public Use Files. Content includes variables similar to those

provided in this report.

A national data source that is a

reasonable alternative to the AHS for measuring housing

needs is the American Community Survey (ACS). The ACS

has the advantage of a larger sample size that supports

estimates for small geographic areas. Disadvantages of

the ACS include addressing housing assistance status

less comprehensively and providing much less information

about housing unit characteristics. For example, the ACS no

longer ascertains whether units contain complete plumbing

systems. HUD also sponsors special tabulations of ACS data

that have HUD income limits information and can be used by

the public to estimate housing needs similar to those in this

report for various small geographies.

Weighting. Because the AHS is based on a sample of

housing units rather than a census of all housing units,

estimates based on the data must be “weighted up” so that

totals for each year match independent estimates of the total

housing stock and better represent the full housing stock.

The Census Bureau weights up responses to account for

undercoverage of households and household nonresponse

(about 15 percent). The weights for 2001-through-2009 AHS

data used in this report are based on the 2000 Census of

Housing, with adjustments for estimated change since then.

Since 2011, AHS data have been weighted to 2010 census

benchmarks. AHS datasets for recent years are provided

with multiple “replicate” weights for each observation that can

be used to estimate standard errors and evaluate statistical

significance without knowledge of stratification and cluster

sampling parameters of complex sample designs. See

Statistical Significance below.

Exclusions. Households reporting incomes that are zero or

negative are excluded from estimates of worst case needs.

However, they are included in counts of total households.

If such households pay rents greater than the fair market

rent and report zero or negative incomes, then their income

situation is presumably temporary, and so they are included

and higher incomes are imputed to them.

Disclosure Review. The local income limits and HAMFI

values that are required to estimate worst case needs are

The AHS Table Creator tool is found at https://www.census.gov/data/data-tools/ahs-table-creator.html.

The Comprehensive Housing Affordability Strategy (CHAS) datasets can be downloaded from https://www.huduser.gov/portal/datasets/cp.html.

linked to local geographies and therefore pose a risk of

disclosing AHS respondents. Accordingly, the analysis

relies on restricted access Internal Use files maintained at

the Census Bureau’s Research Data Center. Tabulations

are reviewed, and values are suppressed as necessary, to

comply with Census disclosure prevention requirements

under the authority of the Data Review Board. The U.S.

Census Bureau reviewed this data product for unauthorized

disclosure of confidential information and approved the

disclosure avoidance practices applied to this release:

CBDRB-FY21-POP001-0037.

Statistical Significance. Assessments of statistical

significance in this report are made regarding 95-percent

confidence intervals. Standard errors associated with

estimates are estimated using the AHS replicate weights and

Fay “Balanced Repeated Replication” method in the SAS

SurveyMeans statistical procedure. Changes between survey

years are judged to be statistically significant if the difference

between estimated values exceeds a critical value based on

the square root of the sum of squared variances. A limitation

of this method is that it is not robust to non-independent

samples inherent to the AHS panel design.

Household and Family Types

In this report, the terms family and household are not

interchangeable because not all households are families.

Families refers only to a subset of households that have one

or more people in the household related to the householder

(the first household member age 18 years or older who is

listed as an owner or renter of the housing unit) by birth,

marriage, or adoption.

Families with children. Households with a child younger

than age 18 present are presumed to meet the definition

of family through relation by birth or adoption (including

grandparents as parents).

Older adult households without children. Households in

which the householder or spouse is age 62 or older and no

children are present. Older adult households may be either

family or nonfamily households.

Other family households. Households with people younger

than 62 as a householder and no children, in which either (1)

one or more people are related to the householder by birth,

marriage, or adoption; or (2) one or more subfamilies reside

there who have members related to each other by birth,

marriage, or adoption.

Other nonfamily households. Households with single

people, younger than 62, living alone or with only

87WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

nonrelatives. Most of these households consist of a single

person living alone rather than unrelated people sharing

housing.

Households with people with disabilities. Before 2009,

no questions in the AHS were designed to directly ascertain

whether individuals suffered from disabilities. Worst case

needs reports for 2007 and earlier identified households

containing people with disabilities using various forms of

income-based proxies. Households with disabilities (1)

were not families with children, (2) were not older adult

households, and (3) received some form of income or

government assistance that is very likely to indicate that an

adult with disabilities is present in the household. The 2009

AHS and subsequent surveys ask direct questions about

impairments and difficulties with activities of daily living for

each household member, including children older than 5

years old. This report, therefore, addresses disability based

on people reported with these problems, except that older

adults who have disabilities do not increase the number

of households counted with disabilities because so many

disabilities are associated with aging.

Housing Assistance Status

In 1997, the AHS questions intended to identify households

receiving rental assistance were changed in both content

and order from those used previously. After careful review,

HUD and the Census Bureau adopted the following

procedure to identify assisted households in a way that

produces results that are more comparable with pre-1997

data. Those questions were further refined in 2007 as a result

of additional cognitive research. In this report, therefore,

receipt of rental assistance is based on respondent reports

designed to determine the following:

- Whether the household must recertify to determine the

rent it pays.

- Whether the rent is less because of a federal, state, or

local government housing program.

- Whether the household has a housing voucher and, if so,

whether it can be used to move to another location.

- Whether the housing authority is the household’s

landlord.

An alternative approach of identifying HUD-assisted

households using an administrative data match is not used

to determine housing assistance status for the purposes of

this report because such an approach excludes assistance

received from other federal, state, or local agencies.

Housing Problems

Rent or cost burden. A ratio of housing costs (including

utilities) to household income that exceeds 30 percent, which

is a conventional standard for housing affordability. To the

extent that respondents underreport total income, the AHS

estimates may overcount the number of households with

a cost burden. A severe cost burden exceeds 50 percent

of reported income. A moderate cost burden exceeds 30

percent but is less than or equal to 50 percent of reported

income. Cost burdens only qualify as potential worst case

needs if they are severe. Households reporting zero or

negative income are defined as having no cost burden.

Inadequate housing. Housing with severe or moderate

physical problems, as defined in the AHS since 1984 and

modified from time to time to reflect changes in the survey.

Severe inadequacies constitute potential worst case needs,

but moderate inadequacies do not. The 2007 AHS eliminated

the questions about hallways (common stairways and light

fixtures) in multiunit structures in the section on selected

physical problems, which affects the classification of units

having severe or moderate physical problems. Briefly, a unit

is defined as having severe physical inadequacies if it has

any one of the following four problems.

1. Plumbing. Lacking piped hot water or a flush toilet or

lacking both bathtub and shower, all for the exclusive

use of the unit.

2. Heating. Having been uncomfortably cold during the

past winter for 24 hours or more, or three times for at

least 6 hours each, because of broken-down heating

equipment.

3. Electrical. Having no electricity or having all of the

following three electrical problems: exposed wiring,

a room with no working wall outlet, and three or more

blown fuses or tripped circuit breakers in the past 90

days.

4. Upkeep. Having any five of the following six

maintenance problems: leaks from outdoors, leaks from

indoors, holes in the floor, holes or open cracks in the

walls or ceilings, more than 1 square foot of peeling paint

or plaster, and rats in the past 90 days.

A housing unit has moderate physical inadequacies if it has

any of the following four problems but none of the severe

problems listed previously.

1. Plumbing. Having all toilets break down simultaneously

at least three times in the past 3 months for at least 3

hours each time.

2. Heating. Having unvented gas, oil, or kerosene heaters

as the main source of heat (because those heaters may

88 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

produce unsafe fumes and unhealthy levels of moisture).

3. Upkeep. Having any three of the six upkeep problems

associated with severe inadequacies.

4. Kitchen. Lacking a sink, range, or refrigerator for the

exclusive use of the unit.

Overcrowding. The condition of having more than one

person per room in a residence. Overcrowding is counted

as a moderate problem rather than a severe problem that

constitutes a potential worst case need.

“Priority” problems. Problems qualifying for federal

preference in admission to assisted housing programs

between 1988 and 1996, including paying more than one-

half of income for rent (severe rent burden), living in severely

substandard housing (including being homeless or in a

homeless shelter), or being involuntarily displaced. These

problems informed the original definition of worst case

needs. Because the AHS sample tracks housing units and

thus cannot count people experiencing homelessness, AHS

estimates of priority problems are limited to the two severe

problems described previously: (1) rent burdens greater

than 50 percent of income or (2) severe physical problems.

In accordance with the intention to estimate the number of

unassisted very low-income renters with priority problems,

the exhibits in appendix A classify households with a

combination of moderate problems and severe problems as

having severe problems.

Income Measurement

Income sources. Income means gross income reported by

AHS respondents for the 12 months preceding the interview.

For each person in the household, the AHS questionnaire

collects the amounts of several different types of income.

Income includes amounts reported for wage and salary

income, net self-employment income, Social Security or

railroad retirement income, public assistance or welfare

payments, and all other money income before deductions

for taxes or any other purpose. Imputed income from equity

is not included as income in this report. Following HUD rules

for determining income eligibility for HUD programs, the

earnings of teenagers age 17 and younger are not counted

as income for this report.

Household income. Reported income from all sources for

all household members age 18 and older.

Income Categories

HAMFI and official income limits. HUD is required by law

to set income limits each year that determine the eligibility of

For details about how HUD sets income limits, see https://www.huduser.gov/portal/datasets/il.html.

applicants for assisted housing programs. In 1974, Congress

defined low income and very low income for HUD rental

programs as incomes not exceeding 80 and 50 percent,

respectively, of HAMFI. HAMFI is more commonly referred

to as area median income (AMI), although the latter term

may be subject to misinterpretation. Note that income limits

are based on median family income (MFI), not on median

household income. HUD determines base income limits for

a household of four, and adjusts them further by household

size: one person, 70 percent of base; two people, 80 percent;

three people, 90 percent; five people, 108 percent; six

people, 116 percent; and so on. Each household is assigned

to an income category using the HUD-determined income

limit for its geographic area and number of household

members.

The Quality Housing and Work Responsibility

Act of 1998 (Pub. Law 105–276) first applied an extremely

low income standard based on 30 percent of HAMFI for

admissions targeting in public housing and the tenant-based

Section 8 program. (See Extremely Low Income below.)

Income cutoffs in association with AHS geography. The

Census Bureau matches AHS survey addresses with HUD

income limit geography and assigns the appropriate income

limits to each case, making the appropriate adjustments for

household size.

Because developing estimates of official income limits for

the geography identified in the AHS microdata was time

consuming, before the 2003 AHS release, HUD prepared

income limits to use with AHS geography for only 3 years:

1978, 1986, and 1995. Income cutoffs for the 2003 AHS

release and each subsequent dataset have been based on

HUD’s current income limits for those years, weighted by

AHS weights. The Census Bureau included those cutoffs to

the AHS public use file through 2013. To protect respondent

confidentiality, income limit variables were restricted to the

AHS internal use file (IUF) in 2015. Additional detail about

income limits can be found in the housing costs-affordability

section of the AHS Codebook interactive tool (Census-HUD,

2019).

Categorizing households by income. For this report,

when households are categorized using the extremely low-,

very low-, and low-income cutoffs, the cutoffs are adjusted

for household size using the same adjustment factors that

HUD programs use. (See additional considerations under

Extremely Low Income below.)

In addition, households reporting negative income are

attributed incomes of slightly more than AMI if their monthly

housing costs exceed the FMR and they lived in adequate

and uncrowded housing. The justification for imputing higher

incomes is that many households in this situation live in

89WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

housing with amenities such as dining rooms, balconies,

and off-street parking and thus may be reporting temporary

accounting losses.

- Extremely low income (ELI). Income not in excess of

30 percent of HAMFI, as determined by the extremely

low-income cutoff used for Section 8 programs. In 2014,

Congress required HUD to begin setting ELI cutoffs for

each area to the greater of 30 percent of HAMFI or the

federal poverty guidelines, but necessarily capped by

the VLI cutoff.

Because of this requirement, 78 percent

of geographic areas had 4-person ELI cutoffs set above

30 percent of HAMFI in 2019. The average increase in

the cutoff among such areas was 32 percent, and for

3 percent of areas that were capped by the VLI cutoff,

the increase was 67 percent. Because federal poverty

guidelines use larger household size adjustments than

do HUD income limits, increases in the ELI cutoffs

were both more likely and more substantial for large

households than for small households.

- Very low income (VLI). Income not in excess of 50

percent of HAMFI, as determined by the very low-

income cutoff. Very low income thus includes extremely

low income, although the term sometimes is used loosely

in specific contexts, such as mismatch analysis, to mean

incomes of between 30 and 50 percent of HAMFI.

- Low income. Reported income not in excess of 80

percent of HAMFI, as determined by the low-income

cutoff.

- Middle income. For this report, income exceeding 80

percent and less than 120 percent of HAMFI.

- Upper income. For this report, income exceeding 120

percent of HAMFI.

HUD allows some jurisdictions exceptions in the

definition of the ELI and VLI cutoffs. Those exceptions

are intended to prevent loss of benefits to assisted

households caused by improvement in local economic

conditions. Thus, the official income limits for ELI and

VLI are in some cases set above 30 or 50 percent of

HAMFI, respectively. The AHS (and thus this report) uses

those official income limits in all its measures.

- Poverty. Household income of less than the U.S.

national poverty guidelines for that household size. As

discussed in appendix A of the Census Bureau’s AHS

publications, AHS poverty estimates differ from official

poverty estimates made from the Current Population

See Frequently Asked Question 4, https://www.huduser.gov/portal/datasets/il.html#2015_faq.

For more detailed information on 2015 AHS metropolitan areas, see https://obamawhitehouse.archives.gov/sites/default/files/omb/bulletins/2013/b13-01.pdf.

Survey. AHS poverty estimates are based on the income

of households rather than the income of families or

individuals, and AHS income questions are much less

detailed and refer to income during the past 12 months

rather than during a fixed period. The poverty guidelines

for a family of four approximates 33 percent of HAMFI.

Comparisons of income limits with poverty thresholds

are presented in exhibits A-6a, A-6b, A-7, and A-8.

- Earnings at minimum wage. Households with incomes

from salary or wages totaling at least as much as one

could earn working full-time (40 hours per week for 50

weeks per year) at the federal minimum wage of $7.25

per hour are defined as having at least full-time earnings

at minimum wage. Thus, the sum of salary and wage

income earned by all persons in the household totals at

least $14,500 annually. Households with incomes from

salary or wages totaling at least one-half that amount

($7,250 annually) are defined as having at least half-time

earnings at minimum wage. Comparisons of household

earnings characteristics are presented in exhibits A-6a,

A-6b, A-7, and A-8.

Location

Metropolitan Statistical Area (MSA). From 1973 to

1983, the definitions of metropolitan location in AHS data

corresponded to the 243 Standard MSAs used in the 1970

census. From 1984 to 2013, metropolitan location in the AHS

has referred to the MSAs defined in 1983, based on the 1980

census. The 2015 AHS redesign that selected a new national

and metropolitan area longitudinal sample for the first time

since 1985 brought metropolitan area definitions up-to-date

with the most current Office of Management and Budget

(OMB) delineations based on the 2010 census, which, at the

time the 2015 AHS sample design took place, was February

2013.

These areas are now termed Core-Based Statistical

Areas (CBSAs).

Region. The four census regions are the Northeast, Midwest,

South, and West.

Mismatch of Supply and Demand for

Affordable Rental Housing

Mismatch. HUD assesses the state of the housing market

by examining the extent of mismatch between the supply of

the rental housing stock and the number of renters whose

household incomes fall below specified thresholds. Three

summary measures are used to characterize the extent of

mismatch at selected income levels:

90 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

APPENDIX E. DATA AND METHODOLOGY

- affordable units per 100 renters;

- affordable and available units per 100 renters; and

- affordable, available, and adequate units per 100

renters.

These mismatch measures can be understood as measuring

the sufficiency of the quantity of housing supplied relative to

the quantity of housing demanded. The italicized terms are

defined and discussed below.

Affordability. Several federal rental programs define

affordable rents as those requiring not more than 30 percent

of an income cutoff defined concerning HAMFI. Under

the Low-Income Housing Tax Credit (LIHTC) program, for

example, housing units with rents up to 30 percent of 60

percent of HAMFI qualify as affordable and eligible for the

credit.

This report generalizes the approach developed to define

LIHTC maximum rents for units of different sizes to define

three categories of affordability (ELI, VLI, and low income)

based on incomes that are sufficient for the rents: at or less

than 30 percent of HAMFI, more than 30 percent and not

more than 50 percent of HAMFI, and more than 50 percent

of HAMFI. Units are assigned to affordability categories by

comparing their gross rent, including payments for utilities,

with affordability thresholds calculated as 30 percent of the

income cutoffs for the corresponding income group. Units

with gross rents above those thresholds are not affordable

because they would cause moderate or severe cost burdens

even for the highest income renters of the income group.

Thus, unit affordability depends on the percent of HAMFI

needed to afford the highest rent in each income category

(ELI, VLI, and so on). For example, to be affordable to ELI

renters, a unit’s gross rent would have to be 30 percent

or less (affordability threshold) of 30 percent of HAMFI

(ELI threshold). The method of assigning units to cost

categories was modified in 2017 to account for limited HUD

administrative exceptions to program income limits.

The income limits used to define rent affordability are

adjusted for the number of bedrooms using the formula

codified at 26 U.S.C. 42(g)(2)(C): no bedrooms, 70 percent of

base; one bedroom, 75 percent; two bedrooms, 90 percent;

three bedrooms, 104 percent; four bedrooms, 116 percent;

and plus 12 percent of base for every additional bedroom.

This formula assumes that an efficiency unit houses one

person, a one-bedroom unit houses 1.5 people, and each

additional bedroom houses another 1.5 people.

Availability. For mismatch analysis, housing units that are

affordable at a specified income level are further assessed

for whether they are currently available to households at that

income level. A unit is available if it is either already occupied

by a household of that income level or currently vacant and

available for rent. Units that are occupied by households of

higher income groups are not “available.”

Adequacy. For mismatch analysis, housing units that are

found to be both affordable and available at a specified

income level are further assessed for whether they are free

of severe physical inadequacies, as discussed under the

Housing Problems heading.

Categorization of rental units and households for

mismatch analysis. To analyze the mismatch between

affordability and income, HUD compares household

incomes and housing unit rents with the current income

limits (for income and rent categories up to and including 80

percent of HAMFI) and to a ratio of HAMFI (for categories

exceeding 80 percent of HAMFI). As in the analysis of

household income, households reporting negative income

are redefined as having incomes slightly greater than MFI

if their monthly housing costs were more than the FMR

and they lived in adequate and uncrowded housing. Units

reported as having “no cash rent” are categorized solely on

the basis of utility costs. Utility costs are allocated to vacant

units through hot-deck imputation based on units that are

comparable based on cost, number of units in the structure,

region, and tenure.

Race and Ethnicity

In 2003, the AHS began using revised Census Bureau

categories of race and ethnicity that are not directly

comparable with the categories used in the AHS from 2001

and earlier. Survey respondents may now select more than

one racial group, causing slight but meaningful decreases in

the size of previously monolithic categories.

The 2017 AHS supports producing estimates of worst

case housing needs for more detailed race and ethnicity

categories than were included in previous reports. In

addition to non-Hispanic White, non-Hispanic Black, and

Hispanic renters, households experiencing worst case

housing needs previously enumerated in an “other” race

category are now reported for Asian, American Indian or

Alaska Native, and Native Hawaiian or other Pacific Islander

renters in exhibit 1-7.

91WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

LITERATURE CITED

Literature

Cited

Akana, Tom. 2021. “CFI COVID-19 Survey of Consumers—Relief Programs,

Vaccines, and the Effects of the Crisis on Renters and Mortgage

Holders.” Special Report. Philadelphia, PA: Federal Reserve Bank,

Consumer Finance Institute. https://www.philadelphiafed.org/consumer-

finance/consumer-credit/cfi-covid-19-survey-of-consumers-wave-7-

updates.

Benfer, Emily A., Robert Koehler, Anne Kat Alexander, et al. 2020. “COVID-19

Eviction Moratoria & Housing Policy: Federal, State, Commonwealth,

and Territory.” https://docs.google.com/spreadsheets/d/e/2PACX-1vTH

8dUIbfnt3X52TrY3dEHQCAm60e5nqo0Rn1rNCf15dPGeXxM9QN9Udx

UfEjxwvfTKzbCbZxJMdR7X/pubhtml.

Bureau of Labor Statistics (BLS). 2021a. “Table A-3. Employment status of

the civilian noninstitutional population by sex and age, seasonally

adjusted.” Labor Force Statistics from the Current Population Survey.

https://www.bls.gov/web/empsit/cpseea03.htm.

———. 2021b. “Table 3. Persons unable to work at some point in the last 4

weeks because their employer closed or lost business due to the

coronavirus pandemic by receipt of pay from their employer for hours

not worked and selected characteristics.” https://www.bls.gov/cps/

effects-of-the-coronavirus-covid-19-pandemic.htm.

Bureau of Economic Analysis (BEA). 2021a. “Table 2.1. Personal Income and Its

Disposition.” National Income and Product Accounts. https://apps.bea.

gov/iTable/index_nipa.cfm.

———. 2021b. “Effects of Selected Federal Pandemic Response Programs

on Personal Income, January 2021.” National Income and Product

Accounts. https://www.bea.gov/data/income-saving/personal-income.

Centers for Disease Control and Prevention (CDC). 2021. “Temporary Halt in

Residential Evictions to Prevent the Further Spread of COVID-19.”

https://www.cdc.gov/coronavirus/2019-ncov/covid-eviction-declaration.

html.

Dalton, Michael. 2020. “Geographic impact of COVID-19 in BLS surveys by

industry.” Monthly Labor Review. Washington, DC: U.S. Bureau of

Labor Statistics. https://doi.org/10.21916/mlr.2020.17.

92 WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

Divringi, Eileen, Eliza Wallace, Keith Wardrip, and Elizabeth

Nash. 2019. Measuring and Understanding

Home Repair Costs: A National Typology of

Households. Philadelphia, PA: Federal Reserve

Bank of Philadelphia. https://www.philadelphiafed.

org/community-development/housing-and-

neighborhoods/measuring-and-understanding-

home-repair-costs.

Econometrica. 2016. American Housing Survey Components

of Inventory Change: 2011–2013. Washington,

DC: U.S. Department of Housing and Urban

Development, Office of Policy Development and

Research. https://www.huduser.gov/portal/datasets/

cinch/cinch13/cinch11-13.pdf.

Joint Center for Housing Studies (JCHS) of Harvard

University. 2016. The State of the Nation’s Housing:

2016. Cambridge, MA: JCHS of Harvard University.

https://www.jchs.harvard.edu/sites/jchs.harvard.

edu/files/jchs_2016_state_of_the_nations_housing_

lowres.pdf.

———. 2019. The State of the Nation’s Housing: 2019.

Cambridge, MA: JCHS of Harvard University. https://

www.jchs.harvard.edu/sites/default/files/Harvard_

JCHS_State_of_the_Nations_Housing_2019.pdf.

National Center for Education Statistics. 2019. Digest of

Education Statistics 2018. Table 105.30, “Enrollment

in elementary, secondary, and degree-granting

postsecondary institutions, by level and control

of institution: Selected years, 1869-70 through fall

2028.” https://nces.ed.gov/programs/digest/d18/

tables/dt18_105.30.asp.

Nisar, Hiren, Mallory Vachon, Charles Horseman, and Jim

Murdoch. 2019. Market Predictors of Homelessness:

How Housing and Community Factors Shape

Homelessness Rates Within Continuums of Care.

Washington, DC: U.S. Department of Housing and

Urban Development, Office of Policy Development

and Research. https://www.huduser.gov/portal/

sites/default/files/pdf/Market-Predictors-of-

Homelessness.pdf.

Social Security Administration (SSA). 2021.“Understanding

Supplemental Security Income – 2021 Edition.”

Woodlawn, MD: Social Security Administration.

https://www.ssa.gov/pubs/EN-17-008.pdf.

Urban Institute. “The CARES Act Eviction Moratorium Covers

All Federally Financed Rentals—That’s One in Four

US Rental Units.” April 2, 2020. https://www.urban.

org/urban-wire/cares-act-eviction-moratorium-

covers-all-federally-financed-rentals-thats-one-four-

us-rental-units.

U.S. Census Bureau. n.d. “Table 10. AHS 2019 National:

Housing Costs, All Occupied Units, Renter; Units by

Structure Type.” American Housing Survey. https://

www.census.gov/programs-surveys/ahs/data/

interactive/ahstablecreator.html.

———. 2021. “Stimulus Table 1. Stimulus Payment Status

and Use, by Select Characteristics: United States.”

Week 25 Household Pulse Survey: February

17–March 1, 2021. https://www.census.gov/data/

tables/2021/demo/hhp/hhp25.html#tables.

U.S. Census Bureau and U.S. Department of Housing

and Urban Development (Census-HUD). 2019.

“Appendix A. Subject Definitions and Table

Index.” American Housing Survey for the United

States: 2019. Washington, DC: U.S. Department

of Commerce, U.S. Census Bureau. https://www2.

census.gov/programs-surveys/ahs/2019/2019%20

AHS%20Definitions.pdf.

U.S. Census Bureau and U.S. Department of Housing

and Urban Development (Census-HUD). 2013.

“Appendix B. 2013 AHS National Sample Design

and Weighting.” In American Housing Survey for

the United States: 2013. Washington, DC: U.S.

Department of Commerce, U.S. Census Bureau.

https://www.census.gov/programs-surveys/ahs/

tech-documentation/def-errors-changes.2013.html.

———. 2017. 2015 AHS Integrated National Sample:

Sample Design, Weighting, and Error Estimation.

Washington, DC: Department of Commerce, U.S.

Census Bureau and HUD. https://www.census.gov/

programs-surveys/ahs/tech-documentation/def-

errors-changes.html.

———. 2019. AHS Codebook. Interactive Tool. Washington,

DC: U.S. Census Bureau. https://www.census.gov/

data-tools/demo/codebook/ahs/ahsdict.html.

U.S. Department of Housing and Urban Development

(HUD). 2018. Programs of HUD: Major Mortgage,

Grant, Assistance, and Regulatory Programs: 2018.

Washington, DC: HUD. https://www.huduser.gov/

portal/publications/programs-of-hud.html.

U.S. Department of Housing and Urban Development,

Office of Community Planning and Development

(HUD-CPD). 2017. The 2017 Annual Homeless

Assessment Report (AHAR) to Congress: Part

1: Point-in-Time Estimates of Homelessness.

Washington, DC: HUD-CPD. https://files.

hudexchange.info/resources/documents/2017-

AHAR-Part-1.pdf.

LITERATURE CITED

93WORST CASE HOUSING NEEDS: 2021 REPORT TO CONGRESS

———. 2018. The 2018 Annual Homeless Assessment

Report (AHAR) to Congress: Part 1: Point-in-Time

Estimates of Homelessness. Washington, DC: HUD-

CPD. https://www.huduser.gov/portal/sites/default/

files/pdf/2018-AHAR-Part-1.pdf

———. 2019. The 2019 Annual Homeless Assessment

Report (AHAR) to Congress: Part 1: Point-in-Time

Estimates of Homelessness. Washington, DC: HUD-

CPD. https://www.huduser.gov/portal/sites/default/

files/pdf/2019-AHAR-Part-1.pdf

U.S. Department of Housing and Urban Development, Office

of Policy Development and Research (HUD-PD&R).

2008. Housing Needs of Persons With Disabilities:

Supplemental Findings to the Affordable Housing

Needs 2005 Report. Washington, DC: HUD-PD&R.

https://www.huduser.gov/portal/Publications/pdf/

Affhsgneedsdis.pdf.

———. 2010. “Report to Congress on the Root Causes of the

Foreclosure Crisis.” Washington, DC: HUD-PD&R.

https://www.huduser.gov/portal/publications/hsgfin/

foreclosure_09.html.

———. 2019. “3rd Quarter 2019 National Housing Market

Summary.” https://www.huduser.gov/portal/ushmc/

quarterly_commentary.html.

U.S. Department of Labor, Employment and Training

Administration (DOL, ETA). 2020. “UI Replacement

Rates: U. S. Replacement Ratios, Average WBA,

and Average Weekly Wage.” https://oui.doleta.gov/

unemploy/ui_replacement_rates.asp.

Wallace, Eliza, Eileen Divringi, and Keith Wardrip. 2019. “A

New Cost-Based Index of Housing Quality and

Repair Needs.” Cityscape. Washington, DC: HUD-

PD&R. 21(3): 299–309.

LITERATURE CITED

U.S. Department of Housing and Urban Development

Ofce of Policy Development and Research

Washington, DC 20410-6000

September 2021