Risk perception and resource scarcity in food procurement during the early outbreak of COVID-19

Abstract

Objectives
The retail food industry, a major essential business, is among the very few thriving sectors during the COVID-19 pandemic. However, such prosperity on the store side does not guarantee a sufficient food supply for all populations. This study aims to understand if people’s risk perception and food security status shaped their food procurement behaviors during the early outbreak of the pandemic.

Study design
Extended from the theory of risk perception, food consumers may behave differently during a disastrous event in terms of store patronization. The study evaluates how food procurement behaviors are affected by perceived risk aversion, resource scarcity, and consumers’ food security status.

Methods
The study examines how people with different food security statuses made grocery shopping decisions at the risk of epidemic exposure based on a nationwide survey of 2,590 participants in the U.S. during the early break of the pandemic in April 2020. The study employs a moderated mediation analysis on in-store shopping frequency and food expenditure.

Results
People having a food-secure status before the pandemic spent significantly more as a result of the reduced shopping frequency (i.e., the secure-insecure subgroup β = -.18, p < .01; the secure-secure subgroup β = -.35, p < .01). The increase in food expenditure was insignificant for people who were food-insecure before the pandemic (i.e., the insecure-insecure subgroup, β = -.01, p > .05; the insecure-secure subgroup, β = -.11, p > .05).

Conclusions
The study reports that in general people reduced the frequency of grocery shopping trips to avoid epidemic exposure while increasing the food expenditure per trip. The increase in food expenditure was not statistically significant among the food-insecure populations likely due to their budget constraints.

Full text

Journal Pre-proof
Risk perception and resource scarcity in food procurement during the early outbreak
of COVID-19
Yiru Wang, Xiang Chen, Yang Yang, Yunhe Cui, Ran Xu
PII: S0033-3506(21)00166-9
DOI: https://doi.org/10.1016/j.puhe.2021.04.020
Reference: PUHE 4254
To appear in: Public Health
Received Date: 10 January 2021
Revised Date: 12 April 2021
Accepted Date: 28 April 2021
Please cite this article as: Wang Y, Chen X, Yang Y, Cui Y, Xu R, Risk perception and resource scarcity
in food procurement during the early outbreak of COVID-19, Public Health, https://doi.org/10.1016/
j.puhe.2021.04.020.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
Published by Elsevier Ltd on behalf of The Royal Society for Public Health.

Risk perception and resource scarcity in food procurement during the early
outbreak of COVID-19
Yiru Wang1, Xiang Chen2, 3, *,Yang Yang4, Yunhe Cui2, Ran Xu3, 5
1. Department of Marketing and Management, State University of New York at
Oswego, Oswego, NY 13126, USA
2. Department of Geography, University of Connecticut, Storrs, CT 06269, USA
3. Institute for Collaboration on Health, Intervention, and Policy (InCHIP), USA
4. Department of Tourism and Hospitality Management, Temple University,
Philadelphia, PA 19122, USA
5. Department of Allied Health Sciences, University of Connecticut, Storrs, CT 06269,
USA
*Corresponding author: Xiang Chen
Email: xiang.chen@uconn.edu
ORCiD: 0000-0002-5045-9253
Yiru Wang: yiru.wang@oswego.edu
Xiang Chen: xiang.chen@uconn.edu
Yang Yang: yangy@temple.edu
Yunhe Cui: yunhe.cui@uconn.edu
Ran Xu: ran.2.xu@uconn.edu
Acknowledgment
This work was supported by the University of Connecticut Institute for Collaboration
on Health, Intervention, and Policy (InCHIP) Covid-19 Rapid Grant. This work was
based in part upon the work of the Geospatial Fellows program supported by the
National Science Foundation (NSF) [grant number: 1743184]; any opinions, findings,
and conclusions or recommendations expressed in this material are those of the
authors and do not necessarily reflect the views of NSF.
Disclosure statement
No potential conflict of interest was reported by the authors. The field survey in the
study was approved by the Institutional Review Board at University of Connecticut
under Protocol X20-0096.
Journal Pre-proof

1
Risk perception and resource scarcity in food procurement during the early outbreak of
1
COVID-19
2
3
Abstract
4
Objectives: The retail food industry, a major essential business, is among the very few thriving
5
sectors during the COVID-19 pandemic. However, such prosperity on the store side does not
6
guarantee a sufficient food supply for all populations. This study aims to understand if people’s
7
risk perception and food security status shaped their food procurement behaviors during the early
8
outbreak of the pandemic.
9
Study design: Extended from the theory of risk perception, food consumers may behave
10
differently during a disastrous event in terms of store patronization. The study evaluates how
11
food procurement behaviors are affected by perceived risk aversion, resource scarcity, and
12
consumers’ food security status.
13
Methods: The study examines how people with different food security statuses made grocery
14
shopping decisions at the risk of epidemic exposure based on a nationwide survey of 2,590
15
participants in the U.S. during the early break of the pandemic in April 2020. The study employs
16
a moderated mediation analysis on in-store shopping frequency and food expenditure.
17
Results: People having a food-secure status before the pandemic spent significantly more as a
18
result of the reduced shopping frequency (i.e., the secure-insecure subgroup β = -.18, p < .01; the
19
secure-secure subgroup β = -.35, p < .01). The increase in food expenditure was insignificant for
20
people who were food-insecure before the pandemic (i.e., the insecure-insecure subgroup, β = -
21
.01, p > .05; the insecure-secure subgroup, β = -.11, p > .05).
22
Journal Pre-proof

2
Conclusions: The study reports that in general people reduced the frequency of grocery shopping
23
trips to avoid epidemic exposure while increasing the food expenditure per trip. The increase in
24
food expenditure was not statistically significant among the food-insecure populations likely due
25
to their budget constraints.
26
Keywords: food procurement; food expenditure; risk perception, food security; COVID-19
27
28
29
Introduction
30
The coronavirus disease 2019 (COVID-19) has shaken the world in every possible way,
31
including how people acquire and prepare food to meet nutritional needs. Food consumers play a
32
critical role in the food supply chain, which was considerably undermined by the pandemic.1-3
33
With an increasing number of away-food outlets (e.g., restaurants, school cafeterias) temporarily
34
closed or permanently running out of business due to the pandemic, people had to rely more on
35
food preparation at home.4,5 Compared with the pre-COVID-19 time, the consumer survey by
36
Hunter6 reported that 54% of the respondents switched to home cooking during the pandemic,
37
for reasons such as saving money, relaxing, and maintaining a healthy diet. More home cooking
38
also changed people’s food procurement behaviors. For example, it was reported that 53% of
39
American consumers stockpiled groceries during the pandemic.7 Many also purchased food
40
storage equipment or developed urban gardens to prepare for lasting and unforeseeable impacts.
41
The changes in food procurement patterns could be perceived as the result of risk aversion
42
against adverse health outcomes.8 Specifically, as people were uncertain about whether their
43
interpersonal interactions and contacts with in-store facilities would expose them to the virus,
44
they may break the regularity of food patronization, such as reducing the frequency of food
45
Journal Pre-proof

3
procurement trips. These behavioral changes have posed formidable challenges for low-income,
46
food-insecure populations, who had difficulty in maintaining healthy diets even before the
47
pandemic.9,10 To alleviate the financial hardships for food-insecure populations, the Food and
48
nutrition services (FNS) of the U.S. Department of Agriculture (USDA) increased the coverage
49
of the nutrition assistance programs, such as the Supplemental Nutrition Assistance Program
50
(SNAP) and the Special Supplemental Nutrition Program for Women, Infants, and Children
51
(WIC), by increasing onsite free meals and providing meal delivery services during business
52
shutdowns and school closures.11 However, it was reported that more than 54 million people in
53
the U.S. still faced the food insecurity issue as the pandemic evolved,12 indicating that the
54
imbalance between food provisioning and food demand has been a paramount societal concern.
55
However, little research has explored the impacts of COVID-19 on food-insecure populations
56
given their relative resource scarcity.
57
This paper examines the food procurement behaviors during the early outbreak of
58
COVID-19 using a nationwide survey, with particular attention paid to food-insecure
59
populations. We employed a moderated mediation analysis to examine changes in food
60
procurement behaviors, which had not been explored via a thorough literature search at the time
61
of the study.
62
63
Methods
64
Conceptual framework
65
The emergence of the COVID-19 pandemic imposed a safety barrier on all populations.
66
Due to the risk of pandemic exposure and the implementation of social distancing orders, food
67
consumers, in general, minimize their essential travel, including grocery shopping trips.
68
Journal Pre-proof

4
Intuitively, when people’s needs for food increase and their shopping frequency reduces, the
69
food expenditure per trip would boost.
70
However, COVID-19’s spillover effect on societal issues, especially medical bills and
71
unemployment, may complicate food procurement patterns. By June 2020, 7.7 million workers
72
had lost their jobs because of the pandemic.13 The loss of income has forced people to deplete
73
savings and face challenges of sustaining basic nutrition needs. A cascading issue arising from
74
unemployment is food insecurity, referring to the status in which individuals or households lack
75
resources to maintain healthy and affordable diets.14 Studies have shown that food-insecure
76
populations are regarded as budget shoppers who are more sensitive to price changes and have
77
lower expenditure per shopping trip.15-17 During the COVID-19, the loss of income could
78
transform people into food-insecure, budget shoppers, while bringing contingent yet lasting
79
impacts on their nutritional wellbeing. We thus propose a conceptual framework to explore the
80
impact of risk perception and resource scarcity on food procurement as a result of the pandemic
81
(Figure 1).
82
83
Journal Pre-proof

1
84
85
Figure 1. Conceptual framework of the study
86
In-store safety
perception
Food
expenditure
Frequency of
grocery shopping
trips
Food security
status
Risk perception changes
Travel behavioral changes
Food procurement behavioral changes
Resource scarcity
Journal Pre-proof

1
Survey instrument and data
87
To investigate the impact of risk perception and resource scarcity on food procurement
88
behaviors due to the rise of the pandemic, we operationalized the conceptual framework into
89
measurable items in a questionnaire. The questionnaire included four sections. First,
90
participants were asked to recall their grocery shopping trips during the early outbreak in
91
April 2020. Cognitive and behavioral questions relating to their last grocery shopping trip,
92
including in-store safety perception (in a 5-point Likert scale), in-store duration of stay (in
93
minutes), travel duration from home to store (in minutes), and total food expenditure (in U.S.
94
dollars) were assessed for each participant. Second, respondents were asked to answer the
95
same set of questions by recalling their grocery shopping trips undertaken in 2019. The
96
"2019 responses" were established as a control condition to evaluate the participant’s
97
behavioral changes during the pandemic. Third, we employed an existing 2-item screen to
98
identify participants with food-insecure status: "I worried whether my food would run out
99
before I got money to buy more" and "the food I bought just didn't last and I didn't have
100
money to get more."18 This 2-item screen was assessed using a 5-point Likert scale, where a
101
response of 3 = “somewhat agree,” 4 = “agree,” or 5 = “strongly agree” to either question
102
labeled the participant as food-insecure. Fourth, the questionnaire also asked for respondents’
103
demographic information, including gender, educational attainment, and employment status
104
during the pandemic. To minimize the variance in data collected from the same respondent,
105
the protocols by Kamran-Disfani et al.19 were applied.
106
The web-based survey was distributed on Amazon Mechanical Turk and lasted for
107
about one month in May 2020. Responses from 2,590 participants living in all 50 U.S. states
108
were collected. A total of 2,388 respondents (92.2%) passed the attention check questions and
109
were included as validated responses in the analysis, as shown in Table 1.
110
Journal Pre-proof

2
We utilized a series of statistical analyses to examine the people’s in-store safety
111
perception change and the resulting food procurement behavioral changes. First, a t-test was
112
conducted to examine the change in perceived in-store safety, shopping frequency, and food
113
expenditure. Then, we employed a moderated mediation analysis to identify the associations
114
among the in-store safety perception change, shopping frequency change, and food
115
expenditure change, and how an individual’s food security status mediated these changes.
116
Lastly, a regression analysis was conducted to quantify the changes in food procurement
117
behaviors by the consumer’s food security statuses before and during COVID-19.
118
Journal Pre-proof

3
Table 1. Demographics of survey participants
119
Variable
Subgroup
N (Percentage)
Gender
Male
1,359
(56.9%)
Non-male
1,029
(43.1%)
Age
18–24
156
(6.53%)
25–34
1,039
(43.51%)
35–44
595
(24.92%)
45–54
365
(15.28%)
55–64
177
(7.41%)
65 and above
56
(2.35%)
Ethnicity
Caucasian
1,692
(70.85%)
African American
342
(14.32%)
Latino
129
(5.40%)
Asian
160
(6.70%)
Native American
37
(1.55%)
Other
28
(1.17%)
Educational attainment
Finished middle school
8
(0.34%)
Finished high school
201
(8.42%)
Some college
398
(16.67%)
Completed 2-year college
220
(9.21%)
Completed 4-year college
1,278
(53.52%)
Attended graduate school
283
(11.85%)
Employment
Employed for wages
1,998
(83.67%)
Not employed for wages
390
(16.33%)
Food security status
(before-during the pandemic)
Secure-secure
1,535
(64.28%)
Secure-insecure
438
(18.34%)
Insecure-secure
24
(1.01%)
Insecure-insecure
1,535
Journal Pre-proof

4
(64.28%)
120
Results
121
We first performed the t-test to compare the changes in food procurement behaviors
122
before and during the pandemic, as shown in Table 2. The table reveals that, in general, food
123
consumers’ safety perceptions of the in-store shopping environment significantly decreased
124
by 0.9 point on a 5-point Likert scale (t = -40.42, p < .01); consumers significantly reduced
125
their shopping frequency (t = -25.23, p < .01) and increased their food expenditure from
126
$119.91 to $131.42 per trip (t = 8.24, p < .01).
127
128
Table 2. Comparison of food procurement behaviors before and during the pandemic
129
(N = 2,388)
130
Before
COVID-19
During
COVID-19
Difference
t-value (t)
In-store safety
perception1
4.66
(.01)
3.76
(.02)
-.90**
-40.42
Shopping frequency2
(standard error [S.E.])
4.43
(.02)
3.95
(.02)
-0.47**
-25.23
Food expenditure3
(S.E.)
4.35
(.02)
4.47
(.02)
.12**
8.24
1. On a 5-point Likert scale, where 1 denotes “very unsafe” and 5 denotes “very safe.”
131
2. Evaluated by the frequency of in-store visits per month.
132
3. Evaluated by U.S. dollars in natural logarithm.
133
**Difference is significant at .01.
134
135
Then, we conducted a moderated mediation analysis using the PROCESS Model 14 to
136
explore variables associated with the food expenditure change (i.e., dependent variable). The
137
PROCESS model is an advanced macro built in the SPSS software to perform customized
138
mediation and moderation analyses.20 The independent variable was the in-store safety
139
perception change, the mediator variable was the shopping frequency change, and the
140
Journal Pre-proof

5
moderator was the food security status during the pandemic (secure or insecure). The control
141
variables included travel time change (in minutes), shopping duration change (in minutes),
142
online food procurement during the pandemic (Y/N, dummy variable), the difference in free
143
time (in minutes), gender (male/non-male, dummy variable), employment status (Y/N,
144
dummy variable), education attainment (college or above/else, dummy variable), residential
145
neighborhood (urban/rural, dummy variable), local infection rate (infections per total
146
population), and local death rate (deaths per total population). The result is shown in Table 3.
147
Table 3 reveals the relationships among the behavioral changes: the increase in in-store
148
safety perception was associated with both the decrease in shopping frequency (β = .18, p <
149
.01) and the increase in food expenditure (β = -7.00, p < .01). Also, people’s food security
150
status during the pandemic further impacted the relationship between shopping frequency and
151
food expenditure, as shown by the interaction term (β = -22.68, p < .01). This result indicates
152
that the mediation effects on food procurement differ among people in different food security
153
statuses.
154
Journal Pre-proof

6
Table 3. Moderated mediation analysis of changes in food procurement behaviors
155
Mediator: Shopping frequency change
Type
Variable
Coefficient (β)
S.E.
t
Independent
variable
In-store safety perception change
0.18**
0.02
10.59
Control
variable
Travel time change
0.01
0.01
-1.15
Shopping duration change
0.01**
0.01
5.01
Online food procurement (Y)
0.01
0.04
0.31
Difference in free time
-0.02**
0.01
-3.17
Gender (male)
0.07
0.04
1.85
Employment status (Y)
-0.08
0.05
-1.53
Education (college or above)
-0.02
0.01
-0.42
Residential neighborhood (urban)
-0.04
0.04
-0.87
Local infection rate
-4.73
8.14
-0.58
Local death rate
108.94
101.62
1.07
Intercept
Intercept
-0.18*
0.07
-2.49
Dependent variable: Food expenditure change
Type
Variable
β
S.E.
t
Independent
variable
In-store safety perception change
-7.00**
1.53
-4.59
Mediation
term
Shopping frequency change
-9.98**
1.89
-5.27
Food security status
-9.30
4.91
-1.90
Interaction term
-22.68**
4.84
-4.68
Control
variable
Travel time change
0.01
0.01
0.16
Shopping duration change
1.05
0.09
11.65
Online food procurement (Y)
3.87
3.32
1.17
Difference in free time
-0.41
0.61
-0.67
Gender (male)
2.18
3.16
0.69
Employment status (Y)
6.38
4.34
1.47
Education (college or above)
2.41
3.68
0.65
Residential neighborhood (urban)
0.22
3.82
0.06
Local infection rate
-557.45
697.20
-0.80
Local death rate
13866.50
8704.23
1.59
Intercept
Intercept
-2.98
6.17
-0.48
* Difference is significant at .05.
156
**Difference is significant at .01.
157
Journal Pre-proof

7
We further examined how people’s food security statuses before and during the
158
pandemic (before-during) affected their food procurement behaviors. Based on Figure 2, we
159
categorized the subjects into four subgroups: insecure-insecure (N = 1,535, 64.28%),
160
insecure-secure (N = 24, 1.00%), secure-insecure (N= 438, 18.34%), and secure-secure (N =
161
391, 16.37%). We found that people who experienced the transition in food security status
162
during the pandemic reduced the frequency of food trips (i.e., the green and red lines in
163
Figure 2), compared with other two subgroups.
164
165
166
167
Figure 2. In-store shopping frequency change categorized by food security status
168
before and during the pandemic (before-during)
169
170
We then conducted a regression analysis on each subgroup to quantify how the changes
171
in food procurement behaviors differ by food security status, where the food expenditure
172
change was the dependent variable and the shopping frequency change was the independent
173
variable (Table 4). It is found that the people having a food-secure status before the pandemic
174
spent significantly more as a result of the reduced shopping frequency (i.e., the secure-
175
insecure subgroup β = -.18, p < .01; the secure-secure subgroup β = -.35, p < .01). The
176
0
0.1
0.2
0.3
0.4
0.5
0.6
-5 -4 -3 -2 -1 0 1 2 3 4
Percentage in the relevant subgroup
In-store shopping frequnecy change
Insecure-insecure Insecure-secure Secure-insecure Secure-secure
Journal Pre-proof

8
increase in food expenditure was insignificant for people who were food-insecure before the
177
pandemic (i.e., the insecure-insecure subgroup, β = -.01, p > .05; the insecure-secure
178
subgroup, β = -.11, p > .05).
179
The results signify that those in a food-secure status before the pandemic were more
180
likely to spend on and stockpiling food because of their relative financial advantages. In
181
contrast, those subject to food insecurity before the pandemic were less likely to increase
182
spending on food even if they tentatively reduced the frequency of store patronization. The
183
finding further suggests that although risk perception can significantly influence people’s
184
food procurement behaviors, such as food expenditure, these influences do not manifest on
185
food insecure populations because of the resource scarcity or the lack of financial resilience
186
to disastrous events.
187
Journal Pre-proof

1
188
Table 4. Regression analysis of changes in food procurement behaviors by food security status subgroup
189
Dependent variable: food expenditure
Insecure-insecure
(N = 1,535)
Insecure-secure†
(N = 24)
Secure-insecure
(N = 438)
Secure-secure
(N = 391)
β
S.E.
t
β
S.E.
t
β
S.E.
t
β
S.E.
t
Independent
variable
Shopping frequency change
-0.01
0.02
-0.14
-0.11
0.17
-0.66
-0.18**
0.03
-5.41
-0.35**
0.04
-9.76
In-store safety perception change
-0.06**
0.02
-2.48
0.22
0.10
2.30
-0.11**
0.03
-4.07
-0.08**
0.02
-3.53
Control
variable
Travel time change
0.01
0.01
6.80
-0.01
0.03
-0.35
-0.01
0.01
-1.08
0.01
0.01
0.68
Shopping duration change
0.01**
0.01
8.97
0.02
0.01
2.88
0.01**
0.01
6.67
0.01**
0.01
6.42
Online food procurement (Y)
0.11*
0.03
3.03
0.33
0.24
1.35
-0.03
0.04
-0.62
0.01
0.05
0.24
Difference in free time
-0.01
0.01
-0.49
-0.03
0.02
-1.33
-0.01
0.01
-0.15
0.01
0.01
1.20
Gender (male)
-0.04
0.04
-1.03
-0.32
0.20
-1.64
-0.04
0.05
-0.93
0.06
0.05
1.21
Employment status (Y)
0.04
0.05
0.78
0.21
0.29
0.73
0.15*
0.07
2.14
0.03
0.07
0.46
Education (college or above)
0.04
0.05
0.89
0.39
0.19
2.11
-0.04
0.05
-0.84
0.05
0.05
1.02
Residential neighborhood (urban)
-0.03
0.05
-0.68
-0.80
0.63
-1.28
0.05
0.06
0.98
0.01
0.05
0.04
Local infection rate
-4.47
7.90
-0.57
-34.83
69.94
-0.50
-11.95
9.06
-1.32
-6.47
9.87
-0.66
Local death rate
91.30
97.50
0.94
939.25
2370.16
0.40
229.95
125.45
1.83
6.54
171.76
0.04
Intercept
Intercept
0.07
0.08
0.97
0.81
0.73
1.10
-0.13
0.09
-1.46
-0.19*
0.10
-1.86
190
† Less than 1% of respondents declared insecure-secure. The small sample size did not warrant the significance of the result in the subgroup.
191
* Difference is significant at .05.
192
**Difference is significant at .01.
193
194
Journal Pre-proof

1
Discussion
195
The study is situated within the unique context of the COVID-19 pandemic, during
196
which food consumers, in general, perceived that shopping groceries at a brick-and-mortar
197
store may pose considerable health risks because of the potential epidemic exposure.
198
Therefore, examining how people behaved differently for food procurement during the
199
pandemic in terms of their trip frequencies and expenditures could help evolve the theory of
200
risk perception and has important health policy implications.
201
The results show that, on average, food consumers chose to reduce the frequency of
202
visits to grocery stores during the pandemic. The reduction in food procurement trips did not
203
lower the demand for food among food-secure consumers; instead, food-secure consumers
204
tended to stockpile food by spending more during each trip, likely due to the fear of epidemic
205
exposure and the preparation for food shortage events.21 Specifically, the study reveals the
206
mediating effects of two important factors on food procurement, referring to risk perception
207
and resource scarcity. First, the study identifies that all people intentionally undertook fewer
208
in-store shopping trips to avoid epidemic exposure. This avoidance behavior, as a result of
209
the change in risk perception, manifested the most among those characterized as food-secure
210
before and during the pandemic (i.e., the secure-secure subgroup). Second, the study reveals
211
that food-insecure populations had a significantly smaller margin in food expenditure. Thus,
212
the food-insecure populations may not have changed their food procurement behaviors due to
213
their budget constraints and the lack of resources to prepare for emergencies.
214
The study provides evidence for stakeholders to develop strategic initiatives and
215
support populations who were victimized by the pandemic. Specifically, food-insecure
216
populations had to reduce the frequency of regular grocery shopping because of the elevated
217
risk perception and safety concerns. However, they could not afford a large expenditure on
218
food purchases due to their deteriorating financial situations in the pandemic and could thus
219
Journal Pre-proof

2
suffer from a potential food shortage crisis. To ameliorate these situations, public health
220
policymakers should customize and implement nutrition assistance programs that prioritize
221
food-insecure populations in a more flexible and timely manner. Some of the ongoing efforts
222
include lifting the qualification requirements for numerous nutrition assistance programs,
223
providing temporary benefits for schools (i.e., P-EBT), and allowing online food purchasing
224
for SNAP benefits.11
225
Nonetheless, the study is subject to limitations. The foremost issue is the short
226
duration of the survey period, which was limited to only the early outbreak in May 2020.
227
With the escalating cases of infection, federal and local governments have stressed the
228
importance of social distancing on possible occasions.22,23 As these non-pharmaceutical
229
interventions became widespread, food consumers developed a herd mentality of avoiding
230
virus exposure, increased awareness of health consequences, and complied with authoritative
231
suggestions to reduce out-of-home activities, including food procurement. Follow-up surveys
232
on the same group of participants and the employment of a latent growth model could reveal
233
these evolving food procurement patterns in a changing pandemic situation. Secondly, using
234
the survey method alone is unable to reveal the context where food is procured.24 For the
235
socio-economically disadvantaged populations, considerable difficulties (e.g., more
236
responsibility for childcare, lack of health insurance) arising from the resource scarcity may
237
exacerbate food insecurity. Therefore, a mixed-method study to incorporate individual
238
interviews and focus-group discussions with food-insecure populations will help recognize
239
these acute challenges in food procurement and will better justify health inequity. Lastly, as
240
the study was situated in the U.S., the conclusion is not applicable to other world regions.
241
Future research could replicate data collection and methods in low- and middle-income
242
countries. This extended effort can facilitate the understanding of food procurement
243
Journal Pre-proof

3
behavioral changes among populations in low- and middle-income countries, eventually
244
providing evidence to ameliorate global food insecurity during and post COVID-19.
245
Journal Pre-proof

4
Acknowledgment
246
This work was supported by the XXX Grant. This work was based in part upon the work of
247
the XXX program supported by the National Science Foundation (NSF) [grant number:
248
XXX]; any opinions, findings, and conclusions or recommendations expressed in this
249
material are those of the authors and do not necessarily reflect the views of NSF.
250
251
Disclosure statement
252
No potential conflict of interest was reported by the authors. The field survey in the study
253
was approved by the Institutional Review Board at XXX under Protocol XXX.
254
255
Journal Pre-proof

5
256
References
257
1. Aday S, Aday MS. Impact of COVID-19 on the food supply chain. Food Quality and
258
Safety. 2020;4(4):167-180.
259
2. de Sousa Jabbour ABL, Jabbour CJC, Hingley M, Vilalta-Perdomo EL, Ramsden G,
260
Twigg D. Sustainability of supply chains in the wake of the coronavirus (COVID-
261
19/SARS-CoV-2) pandemic: lessons and trends. Modern Supply Chain Research
262
and Applications. 2020.
263
3. Espitia A, Rocha N, Ruta M. Covid-19 and food protectionism: the impact of the
264
pandemic and export restrictions on world food markets. The World Bank; 2020.
265
4. Askew K. Life in lockdown: Coronavirus prompts half of French consumers to
266
reappraise "value" of food. foodnavigator.com. May 29, 2020, 2020.
267
5. Sheth J. Impact of Covid-19 on consumer behavior: Will the old habits return or die?
268
Journal of Business Research. 2020;117:280-283.
269
6. Hunter food study special report 2020;
270
https://www.hunterpr.com/foodstudy_coronavirus/. Accessed April 12, 2021.
271
7. Redman R. U.S. consumers ready to stockpile groceries again 2020;
272
https://www.supermarketnews.com/consumer-trends/us-consumers-ready-stockpile-
273
groceries-again Accessed April 12, 2021.
274
8. Lim N. Consumers’ perceived risk: sources versus consequences. Electronic
275
Commerce Research and Applications. 2003;2(3):216-228.
276
9. Devereux S, Béné C, Hoddinott J. Conceptualising COVID-19’s impacts on
277
household food security. Food Security. 2020;12(4):769-772.
278
10. Fan s. Preventing global food security crisis under COVID-19 emergency.
279
International Food Policy Research Institute. March 6, 2020, 2020.
280
11. FNS. FNS responds to COVID-19. 2020; https://www.fns.usda.gov/coronavirus.
281
Accessed April 12, 2021.
282
12. Blach B. 54 million people in America face food insecurity during the pandemic. It
283
could have dire consequences for their health. 2020; https://www.aamc.org/news-
284
insights/54-million-people-america-face-food-insecurity-during-pandemic-it-could-
285
have-dire-consequences-their. Accessed April 12, 2021.
286
13. Fronstin PW, SA. How many Americans have lost jobs with employer health
287
coverage during the pandemic? . 2020;
288
https://www.commonwealthfund.org/publications/issue-briefs/2020/oct/how-many-
289
lost-jobs-employer-coverage-pandemic. Accessed April 12, 2021.
290
14. Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Affairs.
291
2015;34(11):1830-1839.
292
15. Heath C, Soll JB. Mental budgeting and consumer decisions. Journal of Consumer
293
Research. 1996;23(1):40-52.
294
16. Sheehan D, Van Ittersum K. In-store spending dynamics: how budgets invert relative-
295
spending patterns. Journal of Consumer Research. 2018;45(1):49-67.
296
17. Thaler R. Mental accounting and consumer choice. Marketing Science.
297
1985;4(3):199-214.
298
18. Hager ER, Quigg AM, Black MM, et al. Development and validity of a 2-item screen
299
to identify families at risk for food insecurity. Pediatrics. 2010;126(1):e26-e32.
300
19. Kamran-Disfani O, Mantrala MK, Izquierdo-Yusta A, Martínez-Ruiz MP. The impact
301
of retail store format on the satisfaction-loyalty link: An empirical investigation.
302
Journal of Business Research. 2017;77:14-22.
303
20. Hayes AF. Introduction to mediation, moderation, and conditional process analysis: A
304
regression-based approach. Guilford publications; 2017.
305
21. Eastman J. What did the pandemic teach about stocking up for an emergency?
306
Plenty. The Oregonian. June 1, 2020, 2020.
307
Journal Pre-proof

6
22. Adams A, Li W, Zhang C, Chen X. The disguised pandemic: the importance of data
308
normalization in COVID-19 web mapping. Public Health. 2020;183:36.
309
23. Chen X, Zhang A, Wang H, Gallaher A, Zhu X. Compliance and containment in social
310
distancing: mathematical modeling of COVID-19 across townships. International
311
Journal of Geographical Information Science. 2021;35(3):446-465.
312
24. Chen X, Kwan M-P. Contextual uncertainties, human mobility, and perceived food
313
environment: The uncertain geographic context problem in food access research.
314
American Journal of Public Health. 2015;105(9):1734-1737.
315
316
Journal Pre-proof

We employed a national survey in the US to examine food procurement during COVID-19.
In general, food consumers reduced shopping frequency and increased spending per trip.
We employed a moderated mediation analysis on the food expenditure change.
The increase in food expenditure was insignificant for food-insecure populations.
Journal Pre-proof