Food Hardship in the US During the Pandemic: What Can We Learn From Real‐Time Data?

• Published date: 01 June 2022
• Author: Sara Ayllón,Samuel Lado
• Date: 01 June 2022
• DOI: http://doi.org/10.1111/roiw.12564

© 2022 The Authors. Review of Income and Wealth published by John Wiley & Sons Ltd on behalf of

International Association for Research in Income and Wealth

518

FOOD HARDSHIP IN THE US DURING THE PANDEMIC: WHAT CAN

WE LEARN FROM REAL- TIME DATA?

by Sara ayllón*

University of Girona, EQUALITAS & IZA

AND

Samuel lado

University of Girona

We study the potential effect of the declaration of the state of emergency, the beginning and end of

the stay- at- home orders, and the one- off Economic Impact Payments on food hardship in the US

during the first wave of the coronavirus pandemic. We use daily data from Google Trends for the

search term “foodbank” and document the development of a hunger crisis, as indicated by the num-

ber of individuals who need to locate a food pantry through the internet. The demand for charitable

food handouts begins to decrease once families start receiving the stimulus payments, but the biggest

fall comes when economic activity resumes after the lifting of the lockdown orders. Our estimates

indicate that the increased need for emergency help among vulnerable families lasted for at least 10

weeks during the first wave of the pandemic, and we argue that real- time data can be useful in pre-

dicting such urgency.

JEL Codes: I32, I38, H53

Keywords: food hardship, pandemic, Google Trends, CARES Act, stay- at- home orders

1. IntroductIon

The impact of the COVID- 19 pandemic on poverty will be devastating around

the world (Sumner et al., 2020; Palomino et al., 2020), and the US will be no excep-

tion. Yet it may take a long time before researchers can actually measure the full

consequences of the pandemic for vulnerable households, largely because of the

unavailability of data. Surveys that were operating in the field during the COVID- 19

Note: Support from the projects PID2019- 104619RB- C43 and 2017- SGR- 1571 is acknowledged.

Participants at the A&E seminar series at Universidad Carlos III (November 2020), the XXVIII

Meeting on Public Economics (May 2021), the XXIII Applied Economics Meeting (June 2021), the

34th Meeting of the European Society for Population Economics (ESPE) (June 2021), the International

Association for Applied Econometrics (IAAE) Annual Conference (June 2021), the XIV Labour

Economics Meeting (July 2021) and the 9th Meeting of the Society for the Study of Economic Inequality

(ECINEQ) (July 2021) are thanked for their useful comments. Sara Ayllón also thanks the Department

of Social Sciences at the University of Eastern Finland for its warm hospitality while writing the revised

version of this paper. Any errors or misinterpretations are our own.

*Correspondence to: Sara Ayllón, Department of Economics, C/Universtat de Girona 10, 17003,

Girona, Spain (sara.ayllon@udg.edu).

Review of Income and Wealth

Series 68, Number 2, June 2022

DOI: 10.1111/roiw.12564

This is an open access article under the terms of the Creat ive Commo ns Attri butio n- NonCo mmerc

ial- NoDerivs License, which permits use and distribution in any medium, provided the original work

is properly cited, the use is non- commercial and no modifications or adaptations are made.

bs_bs_banner

Review of Income and Wealth, Series 68, Number 2, June 2022

519

© 2022 The Authors. Review of Income and Wealth published by John Wiley & Sons Ltd on behalf of

International Association for Research in Income and Wealth

outbreak had to abruptly terminate all face- to- face interviews.1 While efforts are

being made to keep collecting face- to- face content remotely, the questions that

could not be adapted have been dropped (Sastry et al., 2020). In addition, the qual-

ity of data collected from surveys during the pandemic is of some concern, on

account of the negative effects on response rates, the unrepresentative random sub-

samples (as the virus has affected different socioeconomic groups differently),

comparability of data at different points during the pandemic, respondents’ self-

selection, etc. (Schaurer and Weiß, 2020). Data from multiple surveys are also

expected to be made available to researchers with a substantial delay. However, if

we are to design effective policies to help those most in need, then an accurate diag-

nosis of the economic situation of families is a matter of urgency.

The main objective of this study is to analyze the potential for real- time data

to provide a prompt diagnosis of the effects of the pandemic— and the sudden halt

of economic activity— on one of the most extreme consequences of poverty: food

hardship. With that objective in mind, we use daily data from Google Trends for

the search terms “foodbank” and “food pantry.” The assumption is that families in

need may potentially use the internet to locate the closest foodbank available, once

they run out of goods in their pantries. They might also want to know about open-

ing days and hours and generally how the system works. We want to analyze the

extent to which web search intensities for those terms can be a leading indicator of

subsequent demand for charitable food and ascertain whether real- time data can

serve as an alternative to survey data for poverty analyses. The ultimate goal is to

document the effects of the pandemic— and of the measures imposed to stop its

spread— on the ability of families to provide food for their members. We focus on

data relative to the US, one of the richest countries in the world, with high levels of

food insecurity.2

According to data from the US Department of Agriculture, in 2019, 10.5

percent of households in the US (35.2 million individuals) suffered food insecu-

rity.3 That is, at times during the year, these households were uncertain of having

(or were unable to acquire) sufficient food to meet the needs of all their members,

because they did not have sufficient money or other resources. Some 6.4 percent

of US households qualify as low food secure and 4.1 percent as very low food

secure. In the former case, households obtain sufficient food only by using a num-

ber of strategies, such as eating less varied diets, participating in food assistance

programs or resorting to food pantries. In very low food secure households, the

eating patterns of one or more members have been disrupted and food intake has

1For example, in the US, two major supplements to the Panel Study of Income Dynamics (PSID)—

the Child Development Supplement and the Transition into Adulthood Supplement— were in the field

during the pandemic outbreak and had to cancel all interviews. The same was true of Europe. The most

important data source for the study of poverty in Europe, the European Union– Statistics on Income

and Living Conditions, suspended data collection during the first wave of the pandemic, given the im-

possibility for its interviewers to go into families’ homes. Eurostat has suggested that countries either

postpone/prolong the fieldwork or move from face- to- face to telephone interviewing (Eurostat, 2020).

2We use the term “food hardship” throughout the paper when discussing our foodbank search in-

tensity measure from Google Trends. We refer to “food insecurity” only in those instances in the paper

where the standard definition of food insecurity in the US is used. We would like to thank a referee for

suggesting such use of both terms.

3See www.ers.usda.gov for further details.