Contributing EconomistDiane Whitmore SchanzenbachNorthwestern University

SEPTEMBER 2020Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 2

Prior to COVID-19, even in the midst of a strong economy with a

record streak of job growth and low unemployment rates, in 2018

nearly 8 million (4 percent) American adults reported that members

of their households sometimes or often did not have enough to eat.

According to data from the U.S. Census Bureau’s Household Pulse

Survey (collected April 23, 2020 through July 21, 2020), during

COVID-19, that number has surged to 26–29 million, or 11 percent of

adults. Who are the hungry in America today? This report provides

a description of who didn’t have enough to eat.

The U.S. Department of Agriculture’s (USDA) widely used food

insecurity rate includes a range of those experiencing food

hardships, including those who lack enough money for food and

those who couldn’t afford to eat balanced meals. About 1 in 4 adults

are estimated to be food insecure during COVID-19. This report

goes a step deeper to highlight the subset of this group who are

experiencing severe food insecurity, meaning they reported they

sometimes or often didn’t have enough food (see sidebar).

Not surprising, groups that are most likely to not have enough to

eat are those that typically disproportionately experience poverty:

Black and Latinx families, those with less than a college education,

and children. Approximately 1 in 5 Black and Latinx adults report

they do not have enough to eat. What is more surprising is the

extent of hunger. It’s not just the poorest families who are facing this

struggle; among those who don’t have enough to eat, 1 in 4 have

usual incomes above $50,000 per year. During this crisis, many

have become unemployed, others who have kept their jobs have

seen their earnings decrease due to reduced hours, and others

are expecting to lose earnings in the next month. The economic

shocks they have experienced have pushed many into hunger —

potentially for the first time.

Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis

FRAC gratefully acknowledges sponsorship of this report from the General Mills Foundation, the Kellogg Company Fund, and PepsiCo.

ACKNOWLEDGMENTS

Food Insecurity vs. Not Enough to Eat

The measure investigated

here — sometimes or often

not enough to eat — is a more

severe measure than the

concept of “food insecurity”

that is also tracked by USDA.

Food insecurity is a measure

indicating that a family did not

have consistent, dependable

access to enough food to live

an active, healthy lifestyle.

Food insecurity has not been

measured directly during

COVID-19; using available data

we can estimate that food

insecurity is 25 percent for

adults overall, and 32 percent

for those with children.

Eleven percent of adults and 14

percent of those with children

report that they sometimes

or often don’t have enough

to eat during COVID-19. The

share reporting they don’t

have enough to eat closely

tracks with the share identified

by USDA as having Very Low

Food Security (VLFS), meaning

that the food intake of some

household members was

reduced and normal eating

patterns were disrupted due to

lack of resources.

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 3

Each year, USDA collects a detailed survey on

households’ experiences with food access, including

asking respondents to choose which statement best

describes the food eaten in their homes in the last

year: enough of the kinds of foods we want to eat;

enough, but not always the kinds of food we want;

sometimes not enough to eat; or often not enough to

eat. In 2018, 3.7 percent of respondents overall and

with children reported that they sometimes or often

did not have enough to eat as shown in Figure 1.

During COVID-19, the Census Bureau has been

asking one question, drawn from the larger set

of annual USDA questions, in which respondents

choose among the same options to describe

the food available during the past week. Overall,

between 10.2 and 11.0 percent of respondents

report that they don’t have enough food. Rates are

higher among those with children. In May and June,

13.6 percent of respondents with children said that

they didn’t have enough food. In July, that number

was 14.4 percent.3

Figure 1 also shows that the share reporting

they don’t have enough to eat increases during

economic recessions. In 2008, the first year of

the Great Recession, the share without enough

to eat increased by just over one-third overall

and for respondents with children compared with

the prior year. The COVID-19 recession has been

2005

16.0%

12.0%

8.0%

4.0%

0.0%

2007 2009 2011 2013 2015 2017 2019 May2020

Respondentsw/ Children

RespondentsOverall

Pe

rce

nt

rep

ort

ing

th

ey

som

etim

es

or

oft

en

do

no

t h

ave

en

ou

gh

to

eat

June2020

July2020

Who Doesn’t Have Enough to Eat?

Figure 1. Share of Adults (Overall and With Children) Reporting Their Household Sometimes or Often Does Not Have Enough to Eat: 2005–2018 and May–July 2020

Research has already found that the following have

reduced hunger and other measures of food hardship:

Payments from the new Pandemic Electronic Benefit

Transfer (P-EBT) program, which provides resources

to families who lost access to free or reduced-price

school meals, lifted 2.7–3.9 million children out of

hunger in the subsequent weeks1 since the pandemic

began; and evidence from the prior recession shows

that increasing Supplemental Nutrition Assistance

Program (SNAP, formerly Food Stamps) benefits

helped people and the economy, reducing food

insecurity and also stimulating the local economy.2

In the face of continued high rates of hunger,

Congress should turn to these proven programs —

continuing to provide P-EBT payments as long as

schools aren’t following their normal schedules, and

increasing maximum SNAP benefits by 15 percent.

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 4

characterized by a larger and more rapid increase

in unemployment rates than the Great Recession,

which explains some of the reasons why the rate of

those reporting not enough to eat has spiked so high

in recent months. Of course, widespread closures of

schools and child care centers that usually provide

meals to children is also a contributing factor.

The shares without enough to eat during COVID-19

are disproportionately high among Black and

Latinx respondents — especially among those

with children. As shown in Figure 2, more than 1

in 5 Black and Latinx adults with children reported

in July that they sometimes or often did not have

enough to eat. Note that even prior to COVID-19, in

2018 there were stark disparities across groups with

Black respondents more than three times as likely

as white and Asian respondents who reported not

having enough to eat. Every group has seen their

rates of hunger more than double between 2018

and today, with rates quadrupling among Latinx

respondents (overall and with children). Across every

group, respondents with children in their homes are

more likely to report that they don’t have enough to

eat. A similar pattern holds across education levels

(not shown), with 16 percent lacking enough to eat

among those with a high school diploma or less,

10 percent among those with some college, and

3 percent among those with a college degree or

more. For respondents with children, the rates are

21 percent (high school or less), 13 percent (some

college) and 4 percent (college degree). Rates of

hunger are consistently high across all adult age

levels, and are lower among older adults.

The share reporting not enough to eat varies

predictably by annual income — with rates much

higher among those with the lowest incomes. During

COVID-19, 28 percent of those with usual incomes

below $25,000 per year reported not having enough

to eat (compared with 11 percent in 2018). The share

declines across groups of usual annual income but

is still reasonably common among those with higher

income levels — 1 in 4 of those reporting not having

enough to eat had usual incomes above $50,000

per year. Although across the board every group

experienced job losses during COVID-19, those with

lower levels of usual income were more likely to lose

their jobs.

Pe

rce

nt

rep

ort

ing

th

ey

som

etim

es

or

oft

en

d

o n

ot

have

en

ou

gh

to

eat

Pe

rce

nt

rep

ort

ing

th

ey

som

etim

es

or

oft

en

d

o n

ot

have

en

ou

gh

to

eat

25%

20%

15%

10%

5%

0%

25%

20%

15%

10%

5%

0%

2% 2%

7%

2% 2%

5%

8%

4%

9%10%

23%

20%

17%

7%9%

20%

5%6%

19%

15%

White Black Latinx Asian Other Race

2018

White Black Latinx Asian Other Race

July 2020July 2020 2018

Panel B. Respondents with ChildrenPanel A. Respondents Overall

Figure 2. Share of Adults (Overall and With Children) Reporting Their Household Sometimes or Often Does Not Have Enough to Eat, by Race/Ethnicity: 2018 and July 2020

Black respondents more than three times as likely as white and Asian

respondents who reported not having enough to eat.

More than 1 in 5 Black and Latinx adults with children

reported in July that they sometimes or often did not

have enough to eat.

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 5

A deeper dive into the data shows how vast the

economic distress of COVID-19 has been — and how

closely it is tied to not having enough to eat. Panel

A of Figure 4 shows the share of adults who have

become economically vulnerable during COVID-19.

Twenty-one percent of adults report having become

unemployed themselves during COVID-19 — though

to be sure, some of those are expecting to be

recalled back to their old jobs when the situation

improves. Another 27 percent are workers “on the

edge”: while they themselves are working, they have

either experienced a loss of income due to another

household member’s job loss, fewer hours worked or

other types of reductions, or expect to experience a

loss of job and/or income in the next month. Among

workers on the edge, 90 percent report they have

already experienced a loss of income, and half report

that they both have already lost income and still

expect to lose more. Those with children are more

likely to have lost a job, or to be an employed worker

on the edge.

Panel B of Figure 4 shows the rate of reporting not

enough to eat among those made economically

vulnerable during COVID-19. Overall, 21 percent of

those who lost their jobs during COVID-19 report

not having enough to eat. Workers have not been

Pe

rce

nt

rep

ort

ing

so

me

tim

es

or

oft

en

no

t e

no

ug

h t

o e

at

30%

25%

20%

15%

10%

5%

0%

COVID-19

2018

$100,000+ $75,000–$99,999

$50,000–$74,999

$35,000–$49,999

$25,000–$34,999

$0–$24,999

1 % 1 % 2 %3 %

5 %

8 %

11 %

4 %

8 %

12 %

16 %

28 %

Figure 3. Share of Adults Overall Reporting Their Household Sometimes or Often Does Not Have Enough to Eat, by Usual Annual Income: 2018 and May–July 2020

Figure 4. Economic Vulnerability During COVID-19

Families Don’t Have Enough to Eat because of the COVID-19 Economy

Pe

rce

nt

of p

eo

ple

wh

o lo

st t

he

ir jo

bs

or

we

re “

on

th

e e

dg

e”

du

rin

g C

OV

ID-1

9P

erc

en

t o

f e

ceo

no

mic

ally

vu

lne

rab

le w

ho

re

po

rt s

om

etim

es

or

oft

en

no

t e

no

ug

h t

o e

at

35%

30%

25%

20%

15%

10%

5%

0%

30%

25%

20%

15%

10%

5%

0%

21%

21%

13%

24%

14%

27% 26%

31%

Lostown job

Worker“on the edge”

Worker“on the edge”

Lostown job

Lostown job

Worker“on the edge”

Worker“on the edge”

Lostown job

Panel A. Percent Economically Vulnerableduring COVID-19

Panel B. Share of Economically Vulnerablewithout Enough to Eat

Adults Overall Adults w/ Children

Adults Overall Adults w/ Children

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 6

spared: 13 percent of those who are themselves

still working, but are on the edge due to household

income losses, or who expected losses, report not

having enough to eat.

As has consistently been the case throughout this

analysis, those with children are worse off. Not

only are they more likely to have experienced

job losses or income losses, but, conditional on

experiencing these, they are more likely to not have

enough to eat. The impact on children is further

compounded by children’s loss of meals in schools

and childcare sites.

Figure 4 shows the high rates of economic

vulnerability overall, but job losses have been higher

for Black and Latinx households, as well as for those

with lower levels of education. Job losses have also

been higher in some service sectors, such as food

service and travel.4 Figure 5 separates the measures

of economic vulnerability — among those who either

lost their own job, or lost or expect to lose household

income — by race and ethnicity. A majority of Black

and Latinx respondents have experienced economic

vulnerability, with much higher rates than whites.

Asian respondents and those in the “other race”

category fall in between. Economic vulnerability

results in not having enough to eat at substantially

higher rates for Black and Latinx respondents.

Among the economically vulnerable, 28 percent of

Blacks, 22 percent of Latinx, and 21 percent of the

other race group report not having enough to eat

overall, with 29, 23, and 22 percent, respectively,

among those with children. Similar patterns hold

across income level and education — with those

usually earning less or with lower levels of education

losing jobs and incomes at higher rates, and with

economic vulnerability resulting in not having

enough to eat at higher rates.

Pe

rce

nt

of ad

ults

wh

o li

st jo

bs

or

are

“wo

rke

rs o

n t

he

ed

ge

”

Pe

rce

nt

of e

ceo

no

mic

ally

vu

lne

rab

le w

ho

re

po

rt s

om

etim

es

or

oft

en

no

t e

no

ug

h t

o e

at

70%

60%

50%

40%

30%

20%

10%

0%

35%

30%

25%

20%

15%

10%

5%

0%

Panel A. Percent Economically Vulnerableduring COVID-19, by Race/Ethnicity

Panel B. Percent of the Economically Vulnerable without Enough to Eat

Adults Overall Adults w/ Children Adults Overall Adults w/ Children

55% 55%

61% 62%68%

58%

42%

50%52% 51%28%

29%

22%

Black Hispanic

Asian White

Other Race

23% 22%

10%12% 12%

13%

21%

Black Hispanic

Asian White

Other Race

Figure 5. Economic Vulnerability During COVID-19, by Race/Ethnicity

Despite how stark these numbers are, they would

surely be worse if not for the historic relief efforts

from Congress. Between topping up Unemployment

Insurance benefits, providing Economic Impact

Payments, creating P-EBT, and implementing smart

reforms to SNAP, Congress took helpful steps

at the beginning of the pandemic to buffer the

impact of the economic shock; however, this is

not enough. Too many have not received relief

payments adequate to weather the current

economic storm, and Congress has not yet adopted

or extended policies known to alleviate effectively

food hardship — and do so in a targeted and

temporary manner.5

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 7

Who Doesn’t Have Enough to Eat During COVID-19?

We have seen that Black and Latinx populations,

and those with lower levels of education, are more

likely to not have enough to eat and have been

disproportionately impacted by the COVID-19

economic shock.

Figure 6 shows population characteristics of those

without enough to eat during COVID-19. As shown

in Panel A of Figure 6, 42 percent of those without

enough to eat are whites, followed by 27 percent

Latinx, 22 percent Black, 3 percent Asian, and the

remainder comprising other races. As shown in

Panel B, 61 percent have an education level no

higher than a high school diploma, with 30 percent

having some college, and 9 percent with a college

degree or more. As shown in Panel C, 44 percent

have usual incomes less than $25,000 per year;

one-third have usual incomes between $25,000 and

$49,999, and 23 percent have usual incomes above

$50,000 — many of whom have seen large drops in

their incomes at this time.

Women disproportionately do not have enough to

eat, and have been more likely to lose their jobs

during COVID-19. Of those without enough to eat,

54 percent overall are women, and among those

with children who lack adequate food, 60 percent

of them are women. Those age 60 or older without

enough to eat are also more likely to be women

(58 percent).

In 21 states and the District of Columbia, more than 1

in 10 adults overall report not having enough to eat,

as shown in Figure 7, Panel A. As always, rates are

worse for those with children. In 38 states and the

District of Columbia, more than 1 in 10 adults with

children say they don’t have enough to eat.

White Black

Asian Other

Latinx

$0 to$24.9K

$25K to$49.9K

$50K to$99.9K

<=HS Some College BA+

42% 61% 44%30%

33%22%

27%

23%

3% 6% 9%

Panel A. Race/Ethnicity ofThose Without Enough to Eat

Panel B. Education Level ofThose Without Enough to Eat

Panel C. Usual Annual Income of Those Without Enough to Eat

Figure 6. Characteristics of Those Without Enough to Eat

In 21 states and the District of Columbia, more than 1 in 10 adults overall report not

having enough to eat.

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 8

11CA

7CO

8UT

7ME

8KS

7ID

9MT

9SD

9VA

9PA

9OR

8ND

8WA

8WY

8NE

10AZ

11NM

10IL

10MO

10MI

10IN

10OH

11KY

11WV

11NY

11NC11

TN11OK

14TX

7IA

6MN 7

WI

12AR

12AL

12NV

15LA

17MS

12FL

12GA

12SC

9 DE

11 MD

8 HI

12 DC

10 NJ

10 CT

9 RI

7 MA

6 NH

6 VT

8AK

17%

16%

15%

14%

13%

12%

11%

10%

9%

8%

7%

6%

15CA

9CO

9UT

9ME

11KS

10ID

9MT

13SD

11VA

13PA

13OR

11ND

10WA

11WY

11NE

14AZ

15NM

14IL

12MO

13MI

14IN

14OH

14KY

15WV

14NY

15NC15

TN14OK

17TX

10IA

9MN 10

WI

15AR

16AL

14NV

20LA

20MS

17FL

13GA

15SC

12 DE

16 MD

11 HI

22 DC

13 NJ

13 CT

15 RI

9 MA

9 NH

7 VT

11AK

22%

20%

17%

16%

15%

14%

13%

12%

11%

10%

9%

8%

7%

6%

Figure 7. Percent Reporting Not Enough to Eat, by State

Respondents Experiencing Hunger, Overall

Respondents With Children Who Are Experiencing Hunger

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 9

Food is Key, but It’s About More than Food

Conclusion

Those who report not having enough to eat are also

experiencing other aspects of economic turmoil. As

shown in Figure 8 below, a majority of those without

enough to eat say they expect to lose employment

income in the next month. One-third say that they

are not at all confident that they can make their

next housing payment, and nearly half are not at all

confident that they can afford the food they need

over the next month. More than one-third report that

they did not make their last housing payment. More

than 80 percent of those without enough to eat

reported experiencing bad mental health symptoms

at least half of the days in the prior week, including

nervousness/anxiety, inability to stop worrying, little

interest in doing things, and feeling down, depressed

or hopeless. Rates of bad mental health symptoms

among those without enough to eat are 20 to 30

percentage points higher than those among the

population overall.

In addition to being more likely to not have enough

to eat, others have found that death rates from

COVID-19 are higher for Black and Latinx people

of all ages — with differences in death rates much

higher for those at younger ages, including death

rates 10 times higher among Black individuals

between the ages of 35 to 44 years old, and seven

times higher for those who are 45 to 54 years old.

Rates for Latinx people are slightly lower than for

Black people, but still well above white death rates.6

During COVID-19, far too many Americans report

that they do not have enough to eat. This is not a

problem limited to certain populations or regions.

Approximately 1 in 5 Black and Latinx adults say they

do not have enough to eat, as do 1 in 14 white and

Asian adults. In the vast majority of states more than

1 in 10 adults with children don’t have enough to eat.

Most have been pushed into this state by job loss.

Many others are still employed, but have lost hours

or earnings, as well as other supports, which have

pushed them into hunger. Across the board, those with

children are more likely to not have enough to eat.

Congress has a number of proven policies that it

can adopt to help struggling families get enough to

eat. In particular, increasing maximum SNAP benefits

and extending the P-EBT program, which provides

benefits to those who lost access to free or reduced-

price school meals, are two of the most effective

food assistance policies.7 Congress should, without

delay, act to increase maximum SNAP benefits by 15

percent and extend P-EBT payments, so that fewer

people in America will go without enough to eat.

w/ ChildrenOverall

Expect an HH job loss in the next 4 weeks due to COVID

“Not at all” confident inability to pay mortgage

or rent

“Not at all” confident that HH will be able to a�ord the kinds

of foods needed for the next 4 weeks

Didn’t pay last mortgage/rent

80%

60%

40%

20%

0%

65% 68%

33% 35%

48%45%

36% 38%

Figure 8. Widespread Economic Distress among Those Without Enough to Eat

FRAC n Not Enough to Eat: COVID-19 Deepens America’s Hunger Crisis n www.FRAC.org 10

Endnotes1 Bauer, Lauren, Abigail Pitts, Krista Ruffini, and Diane Schanzenbach. “The Effect of Pandemic-EBT on

Measures of Food Hardship,” Brookings Institution, July 30, 2020. https://www.brookings.edu/research/

the-effect-of-pandemic-ebt-on-measures-of-food-hardship/

2 Schanzenbach, Diane, Lauren Bauer, and Greg Nantz. “Twelve Facts about Food Insecurity and SNAP,” The

Hamilton Project, Brookings Institution, April 2016. https://www.hamiltonproject.org/assets/files/twelve_

facts_about_food_insecurity_and_snap.pdf

3 Differences between May and June are not statistically significant, but July rates are statistically higher

than those in May and June. The Census Household Pulse is collected online, while the annual survey

typically is collected through a telephone survey. Research suggests that reported rates of food hardship

are lower in surveys collected in person or via telephone compared to those collected online, possibly due

to respondents’ social desirability bias when responding to a live interviewer (Karpman, Michael, Stephen

Zuckerman, and Dulce Gonzalez. 2018. “The Well-Being and Basic Needs Survey: A New Data Source for

Monitoring the Health and Well-Being of Individuals and Families.” Washington: Urban Institute).

4 Montenovo, Laura, Xuan Jiang, Felipe Lozano Rojas, Ian M. Schmutte, Kosali Simon, Bruce A. Weinberg, and

Coady Wing. “Unequal Employment Impacts of COVID-19.” Econofact, June 1, 2020. https://econofact.org/

unequal-employment-impacts-of-covid-19

5 Bitler, Marianne, Hilary Hoynes, and Diane Schanzenbach. “The social safety net in the wake of COVID-19.”

Forthcoming, Brookings Papers on Economic Activity.

6 “Race gaps in COVID-19 deaths are even bigger than they appear” by Tiffany Ford, Sarah Reber, and

Richard V. Reeves, Brookings Institution. https://www.brookings.edu/blog/up-front/2020/06/16/race-gaps-

in-covid-19-deaths-are-even-bigger-than-they-appear/

7 Hoynes and Schanzenbach argue for an automatic 15 percent increase in maximum SNAP benefits during

recessions as an automatic stabilizer. See Hoynes, Hilary and Diane Schanzenbach, “Strengthening SNAP

as an Automatic Stabilizer,” in Boushey, Heather, Ryan Nunn, and Jay Shambaugh, eds., Recession Ready:

Fiscal Policies to Stabilize the American Economy. The Hamilton Project and Washington Center on

Equitable Growth.