research
a466247690
ImpactsofCOVID-19onfoodsecurity_PaneldataevidencefromNigeria.pdf
IFPRI Discussion Paper 01956
August 2020
Impacts of COVID-19 on Food Security
Panel Data Evidence from Nigeria
Mulubrhan Amare
Kibrom A. Abay
Luca Tiberti
Jordan Chamberlin
Development Strategy and Governance Division
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
The International Food Policy Research Institute (IFPRI), a CGIAR Research Center established in 1975,
provides research-based policy solutions to sustainably reduce poverty and end hunger and malnutrition.
IFPRI’s strategic research aims to foster a climate-resilient and sustainable food supply; promote healthy
diets and nutrition for all; build inclusive and efficient markets, trade systems, and food industries;
transform agricultural and rural economies; and strengthen institutions and governance. Gender is
integrated in all the Institute’s work. Partnerships, communications, capacity strengthening, and data and
knowledge management are essential components to translate IFPRI’s research from action to impact.
The Institute’s regional and country programs play a critical role in responding to demand for food policy
research and in delivering holistic support for country-led development. IFPRI collaborates with partners
around the world.
AUTHORS
Mulubrhan Amare ([email protected]) is a Research Fellow in the Development Strategy and
Governance Division of the International Food Policy Research Institute (IFPRI), Washington, DC.
Kibrom A. Abay ([email protected]) is a Research Fellow in IFPRI’s Development Strategy and
Governance Division, Cairo office.
Luca Tiberti ([email protected]) is Director of the Partnership for Economic Policy (PEP) and
Assistant Professor in the Department of Economics at
the University of Laval, Quebec.
Jordan Chamberlin ([email protected]) is a Spatial Economist at the Maize and Wheat
Improvement Center (CIMMYT), Nairobi Office.
Notices
1 IFPRI Discussion Papers contain preliminary material and research results and are circulated in order to stimulate discussion and
critical comment. They have not been subject to a formal external review via IFPRI’s Publications Review Committee. Any
opinions
stated herein are those of the author(s) and are notnecessarily representative of or endorsed by IFPRI.
2 The boundaries andnames shown and the designationsused on the map(s) herein do not imply official endorsement or
acceptance by the International Food Policy Research Institute (IFPRI) or its partners and contributors.
3 Copyright remains with the authors. The authors are free to proceed, without further IFPRI permission, to publish this paper, or any
revised version of it, in
outlets such as journals, books, and other publications.
Abstract
This paper combines pre-pandemic face-to-face survey data with follow up phone surveys
collected in
April-May 2020 to quantify the overall and differential impacts of COVID-19 on
household food security, labor market participation and local food prices in
Nigeria. We
exploit
spatial variation in exposure to COVID-19 related infections and lockdown measures along with
temporal differences in
our outcomes of interest using a difference-in-difference approach. We
find that those households exposed to higher COVID-19 cases or mobility lockdowns experience
a significant increase in measures of food insecurity. Examining possible transmission channels
for this effect, we find that COVID-19 significantly reduces labor market participation and
increases food prices. We
find that impacts differ by economic activities and households. For
instance,lockdown measures increased households' experience of food insecurity by 12percentage
points and reduced the probability of participation in non-farm business activities by 13 percentage
points. These lockdown measures have smaller impacts on wage-related activities and farming
activities. In terms of food security, households relying on non-farm businesses, poorer
households, those with school-aged children, and
those living in
remote and conflicted-affected
zones have experienced relatively larger deteriorations in
food insecurity. These findings can
help
inform immediate and medium-term policy responses, including social protection policies aiming
at ameliorating the impacts of the pandemic,
as well as guide targeting strategies of governments
and international donor agencies by identifying the most impacted sub-populations.
Keywords: COVID-19, Pandemic, Food security, labor market participation, food price.
JEL Codes: I12, O13, Q18, Q12, Q18
iii
Acknowledgments
This paper has been prepared as an
output of the CGIAR Research Program on Policies,
Institutions, and Markets (PIM), led
by the International Food Policy Research Institute (IFPRI).
We are grateful
to PIM and the Partnership for Economic Policy (PEP) for providing financial
assistance to conduct this study.
.
iv
1. Introduction
The COVID-19 pandemic is
ravishing local, national, and global economies. In addition to
the
direct health impacts, the pandemic is having widespread effects on employment, poverty, food
security, nutrition, education and health, and the overall functioning of food systems (Barrett,
2020; Devereux et
al., 2020; Swinnen, 2020; GAIN, 2020). COVID-19 is destabilizing supply
chains at
all levels, and creating instability in food supply and food prices (Zurayk, 2020; Torero,
2020; Reardon et al. , 2020a; Reardon et al., 2020b; Ihle et
al., 2020; Akter, 2020; FAO, 2020).
The World Bank's recent forecasts show that, globally, the pandemic is
likely to
push 49 million
people into extreme poverty in
2020 (World Bank, 2020a).1 More than 45 percent (23 million
people) of these people are in
Sub-Saharan Africa, implying that the region will be hit hardest in
terms of increased extreme poverty. The United Nations World Food Programme (WFP)
estimated that the number of people globally facing acute food insecurity would almost double by
the end of 2020 (about 135 million people before the crisis), due to
income and remittance losses,
and disruption of food systems associated with the
pandemic (WFP, 2020a; WFP, 2020b).
This paper quantifies the impacts of COVID-19 and associated governmental lockdown
measures on household food security and labor market participation in Nigeria. Nigeria is an
interesting case study, as
about 83 million people were already living below the national poverty
line (World Bank 2020a). According to
the recent World Bank projections, Nigeria is predicted to
be one of the three countries with the highest increase in
the number of poor people.2 About 5
million Nigerians are projected to be pushed into poverty because of COVID-19 and associated
mobility restrictions and lockdown measures (World Bank, 2020a; IMF, 2020). Food insecurity
has been a major longstanding challenge in
Nigeria, as
reflected by Nigeria's high Global Hunger
Index (GHI), low Food Consumption Score3 (FCS), and high-calorie deficiency (Global Hunger
Index, 2019). The country also experiences significant seasonal and geographical food price
fluctuations due to
weather shocks to
agricultural production, limited access to
markets and
1 The share of the world’s population living on less than $1.90 per day is projected to
increase from 632 million to
665 million people (World Bank, 2020a).
2 The three countries with the largest change in
the number of poor are estimated to be India (12 million), Nigeria (5
million) and the Democratic Republic of Congo (2
million) (World Bank, 2020a).
FCS is a composite score constructed on the basis of dietary diversity, food frequency, and relative nutritional
importance of different food groups.
3
1
infrastructure, and global food price volatility on imported staple foods.4 Disruptions in
economic
activities are likely to
have direct repercussions on food security as
household spending on food
comprises 58% of household expenditures, with poorer households spending more than 75% of
their resources on food (USDA, 2016; FAO, 2020). Disruptions in
domestic economic activities
and international food markets are therefore very likely to affect the food security of Nigerian
households through various channels (Eriksson et
al., 2008; Barrett et
al., 2019; Devereux et al.,
2020; Baldwin and Weder di Mauro, 2020; Haddad et
al., 2020; Béné, 2020). In addition, the
availability of a large nationally representative panel of households observed before and after the
start of the pandemic makes Nigeria an
ideal setting for an
early empirical examination of COVID
19’s impacts.
Besides quantifying the impact of the spread of the pandemic and government lockdowns
on food security outcomes, this research also aims to
shed light on key impact pathways and
differential impacts of the pandemic. COVID-19 could affect food security of households through
different pathways (Baldwin and Weder di Mauro, 2020; Devereux et al., 2020). For instance,
COVID-19 related lockdowns and social distancing measures can adversely affect incomes by
reducing economic and livelihood activities (Devereux et
al., 2020; Barett, 2020; Reardon et
al.,
2020b), which directly affects food security. In Nigeria, recent projections show that the economy
will contract by between 3.5 to 5 percent in
2020 during the period the government-imposed
lockdown and mobility measures (World Bank, 2020c; IMF, 2020; Andam et al., 2020). These
lockdowns and restrictions are also disrupting food supply chains and community services,
including education-linked programs (e.g., school feeding) and social protection programs, which
ultimately positively affect food prices (WFP, 2020a). For countries like Nigeria that heavily rely
on imports of major staple foods such as
rice and wheat, which registered marked rapid climbs in
spot prices, this is
creating an
added financial burden that directly affects food security of
households (World Bank. 2020a).5 National and state-level restrictions and lockdowns are
affecting food transportation within the country, with clear implications on food supply and,
4 Nigeria imported 2.4 million metric ton of rice in 2019/2020. Nigeria spentmorethan USD4.1 billion on food import
(NBS, 2020).
5For example, the cost of rice in retail markets soared by more than 30%in March alone (Bloomberg: Key Food Prices
Are Surging After Virus Upends Supply Chains: https://news.bloomberglaw.com/international-trade/key-food-prices
are-surging-after-virus-upends-supply-chains
2
consequently, on food prices. This is
expected to
generate significant repercussions on food
insecurity, particularly in poorer and vulnerable urban households (Ericksen et
al., 2010; Tendall
et al., 2015; Gilligan, 2020). We thus evaluate direct effects on two key important channels
through which food security outcomes are likely affected: the disruption of economic activities
and increases in local food prices.
The effect of the pandemic are expected to
differ both by geography and by type of
household, with preexisting vulnerabilities to
food security likely to
be magnified (Amjath-Babu
et al., 2020; Béné, 2020; Devereux et al., 2020; Ravallion et al., 2020; Mobarak and Barnett
Howell, 2020). Nigeria has significant longstanding geographical variation in poverty and food
insecurity – more than 75 percent of poor Nigerians live in
the north of the country – and the
pandemic is
likely to disproportionately exacerbate food insecurity in
those already fragile and
conflict-affected zones (World Bank, 2020b, 2020c). Impacts are expected to be most severe for
poorer households in both rural and urban areas (Ericksen et al., 2010; Ravallion et al., 2020;
Mobarak and Barnett-Howell, 2020). As the spread of the pandemic initiates in urban areas,
government responses, including mobility restrictions and lockdowns, will likely be most intense
in urban areas and may affect urban residents more directly than rural households in
the short term.
However, the impact of COVID-19 is also expected to
vary across livelihood options, with those
activities that require face-to-face interactions likely to
experience a significant loss in
demand
(e.g., Abay et
al., 2020; Baldwinand Weder di Mauro, 2020). Value chain disruptionsmay extend
deeply into rural areas, affecting both input supply and output demand for farmers and affecting
the income of those employed in
both upstream and downstream agricultural value chains (Barrett
et al., 2019; Amjath-Babu
et al., 2020, Reardon
et al., 2020a). Closure or disruption of informal
food markets, where the poor obtain the majority of their food, may be more severe in
extent and
food security impacts than impacts on formal markets (Devereux et
al., 2020; Barrett, 2020). We
thus explore potential differential impacts along these dimensions, including livelihood strategies
and options.
Combining pre-COVID-19 face-to-face surveys with post-COVID-19 phone surveys and
primary data on states' infections and lockdown measures, we exploit spatial variations in exposure
to COVID-19 along with temporal changes
in various food security indicators using a difference
in-difference approach. By
comparing food security outcomes of households with varying
exposure to the pandemic before and after the outbreak of the pandemic we can
plausibly quantify
3
the overall and differential causal impact of the pandemic. We
also quantify similar impacts
associated with government state-level responses, mainly lockdown and associated mobility
restrictions. With a similar methodology, we also test two main pathways that would, directly and
indirectly, impact households' food security, i.e., the effects of COVID-19 on labor market
participation and food prices.6
We find that those households exposed to higher COVID-19 cases or more strict
government responses experience significant increase in food insecurity indicators. Also, as
plausible drivers of this result, because of COVID-19, labor market activities deteriorate, and food
prices increase in
those areas most affected by the spread of the pandemic and lockdown measures.
For instance, doubling of the number of confirmed cases increase households' experience of food
insecurity by 2-3 percentage points, while it leads
to a reduction in major economic activities by
1-3 percentage points. State-level lockdown measures have much larger impacts on these
outcomes: lockdowns increased households' experience of food insecurity by 12 percentage points,
and reduced non-farm business activities by 13 percentage points. These results remain consistent
across alternative indicators of food insecurity and labor market participation, albeit some
differences in the impacts across alternative activities. We
show that food insecurity is
also
affected by COVID-19-related increases in
food prices. We also document important differential
impacts across various economic activities and households. For instance, state-level lockdown
measures are more impactful in
disrupting non-farm business activities, while farming activities
appear to
be less affected by state-level lockdowns. Similarly, poorer households, those with
school-aged children, and those households living in
remote and conflict-affected zones bear the
highest brunt of the pandemic. In terms of livelihood options, those households engaged in non
farm business activities appear to be hardest hit, while those engaged in
wage-related activities are
relatively less affected.
Although considerable anecdotal evidence has been generated recently on the impacts of
COVID-19 on the labor market and food security outcomes in Africa, rigorous empirical studies
based on household-level survey data have largely not been available before now. Understanding
the magnitude, distributional differences, and pathways of impacts of the COVID-19 pandemic on
households' food security and economic activities is
critical for designing effective policies and
6 https://www.ifpri.org/blog/how-covid-19-may-disrupt-food-supply-chains-developing-countries
4
interventions to
mitigate the adverse effects of the pandemic. Using nationally-representative
household survey, this paper contributes new evidence on the effects of COVID-19 on food
security as
well as
on key impact pathways. These findings have important implications and hence
can inform immediate and medium-term policy responses. For instance, our findings
can inform
social protection policies aiming at
weathering the impacts of the pandemic, which rely heavily on
effective targeting strategies. This is
particularly imperative for governments like Nigeria, which
has limited fiscal space and competing needs for post-COVID-19 recovery investment. Our
findings can
help governments and international donor agencies improve their targeting strategies
to identify the most impacted sub-populations. The evidence that government responses, such
as
lockdowns and other mobility restrictions, have disproportionately large negative impacts on
poorer households is
consistent with arguments made by those who are critical of such policies for
low- and middle-income countries (e.g., Ravallion et al., 2020, Mobarak and Barnett-Howell,
2020; Bargain and Aminjonov, 2020).
The remainder of this paper is
organized as follows. Section 2 describes the context and
data. Our empirical strategy is presented in
Section 3. Section 4 presents estimation results and
associated discussions, while Section 5 provides concluding remarks.
2. Context and Data
2.1. Context
Nigeria is Africa's most populous country, with a high poverty rate, large informal sector economy,
high dependence on imported staples, and high exposures to
shocks. Nigeria is
one of the few
African countries that first recorded COVID-19 cases and hence among those African countries
who experienced significant economic disruptions because of the pandemic. The first COVID-19
case in
Nigeria was recorded on February 27, and by late June, the number of confirmed cases
passed the 30,000 mark (NCDC, 2020).7 As part of the measures to
contain the spread of the
pandemic, federal and state-level governments have introduced social distancing and mobility
restrictions in March 2020 (FMBNP, 2020). The federal government closed all schools in mid
7 The Nigerian Centre for Disease Control (NCDC) is
responsible for overall management of testing, isolation, and
treatment of COVID-19 patients.
5
March, and several states and local authorities introduced bans on public and social gatherings. By
late March, the Nigerian government closed its land and air
borders to
all travelers and suspended
passenger rail services within the country (Ogundele, 2020; NCDC, 2020). Furthermore, the
federal government announced fiscal and stimulus measures, amounting up to
50 billion Naira to
support households, and small and medium-scale enterprises affected by COVID-19 (FMBNP,
2020).
Nigeria's lockdown and mobility restrictions were mostly introduced by federal and state
level governments. On March 29, 2020, the federal government announced lockdown measures
and strict mobility restrictions for Abuja FCT, Lagos, and Ogun states, which lasted for five weeks
from March 30 until May 4.8 The federal government also introduced similar lockdown measures
for Kano state, which started in mid-April and lasted for seven weeks. Lockdowns restrictions in
other states were introduced by state governments independently of the federal government,
including in Akwa Ibom, Borno, Osun, and Rivers. In most cases, the lockdowns remained
in force
for about 5-8 weeks. These measures restricted movement of residents and led to the closure of
business operations, and closure of regional borders linking lockdown areas with the rest of the
country.
These lockdown and mobility restrictions are likely to disrupt major economic activities,
including local businesses. Nigeria is highly susceptible to
income shocks and food insecurity
associated with the spread of the pandemic. As we show in
the next sections, food prices are
already soaring in the country, food supply chains (domestic and international) are being disrupted,
informal sector unemployment rates are likely to be increasing, and poor households are likely to
be facing food shortages. All these effects are likely to
increase food insecurity.
2.2. Data and sampling strategy
In this study, we combine the pre-COVID-19 face-to-face survey with post-COVID-19 phone
survey to
quantify the overall and differential impact of COVID-19 on households' food security
8 In addition to lockdown measures, federal and state government implemented different measures includes: i) travel
bans which includes restricted entry into the country for travelers from high risk countries; closure of two main
international airports; suspension of all railway passenger services in the country; closure of all air and land borders.
ii) closure of schools and religious institutions. iii) Bans on
public and social gatherings across all states in Nigeria.
iv) Curfew hours which restrict movement of people.
6
and labor market participation. These data and surveys are part of the World Bank's Living
Standards Measurement Study – Integrated Surveys on Agriculture (LSMS-ISA) and are collected
in collaboration with the Nigerian Bureau of Statistics (NBS). The LSMS-ISA data for Nigeria,
also known as
General Household Survey-Panel (GHS-P) include four rounds collected in 2010
11, 2012-13, 2015-16 and 2018-19. These data are nationally representative, and they provide
detailed information on employment, income, food, and nutrition security indicators.
Following the COVID-19 pandemic, the LSMS-ISA program has initiated tracking of
national samples of households that had been interviewed during the latest rounds of the LSMS
ISA surveys using phone surveys.9 Among the total sample of households (4,976) interviewed in
the latest round (post-harvest January/February visit) of the GHS-P survey in
2019, 4,934 (99.2%)
provided at
least one phone number. Out of the full sample of households with phone numbers, a
random sample of 3,000 households was selected for the phone survey, to collect a complete
sample of 1800 households that enable statistical monitoring of (monthly) changes in
key
outcomes of interest. Out of these 3,000 households prepared for phone survey, 69 percent of
sampled households were successfully contacted, and among these, 94 percent (1,950) households
were fully interviewed (NBS and World Bank, 2020). The final complete sample for the phone
survey constitutes these 1,950 households, and they are expected to be contacted in subsequent
rounds of the survey. To
create a balanced panel across rounds, we merged these households with
the immediately previous round (2019) and kept those households with complete information in
both rounds.
To adjust for potential (systematic) attrition in
the phone survey and construct nationally
representative statistics, one must construct and
apply appropriate sampling weights. This is
important, although a comparison of observable characteristics from the GHS-P and the phone
survey shows reasonably comparable statistics (NBS and World Bank, 2020). The LSMS-ISA
team constructed the sampling weights using the weights for the GHS-Panel as
the basis, with
further adjustment for attrition in the phone survey. The weights for the final sample of households
from the phone survey were calculated in several stages, and readers are referred to NBS and
World Bank (2020).10
9 These phone surveys have been (are being) conducted in Ethiopia, Malawi, Nigeria, Tanzania and Uganda.
See http://documents1.worldbank.org/curated/en/717901591889288314/pdf/Basic-Information-Document.pdf for
detail information on sampling weights.
10
7
In this paper, we use the first round of the phone survey (the only available at the time of
writing), which was administered in April-May 2020.11 The LSMS-ISA phone surveys are planned
to be monthly surveys and hence are high-frequency surveys. These high-frequency phone surveys
covered topics including (1) knowledge regarding the spread of COVID-19; (2) prices and access
to food and non-food necessities; (3) employment and income losses; (4) food insecurity; and (5)
subjective wellbeing. We
are more interested in those outcomes, which can be observed in
the
face-to-face (pre-COVID-19) and phone (post-COVID-19) surveys. Since both the pre-and post
COVID-19 LSMS-ISA data contain important information on households' participation in
economic activities, types of employment, income, and food insecurity experience, we can
examine the patterns of food insecurity and labor allocation along multiple periods. As we discuss
below, we are particularly interested in tracking impacts on food insecurity and disruptions in
economic activities, which are both followed and measured in
similar ways in both rounds.
Table 1 presents the weighted summary statistics of selected variables used in our analysis.
For comparison purposes and for those variables observed in both rounds (and that are not expected
to change significantly because of COVID-19), we report these summary statistics separately for
each round. Those observable household characteristics that are observed for both rounds appear
to be statistically comparable across both rounds. This is encouraging as
most of these household
characteristics are not expected to
change in
such a short period significantly. About 19 percent of
our sample are female-headed in
the 2019 round, while the corresponding figure for the 2020 round
amounts 18 percent. We
also show a few other pre-COVID variables which are used to capture
eventual heterogeneity effects across the population.
11
The outcomes and information from 2019 were collected during January and February 2019.
8
Table 1: Descriptive results of key
explanatory variables
Pre-COVID-19
(2019)
Post-COVID-19
(2020)
Male headed households (yes=1) 0.81 0.82
Age of head (years) 49.64 49.42
Education of head (years) 8.21 8.87
Family size (numbers) 5.53 5.52
Value of assets (PPP US) 1677.66 -
Urban households 0.38
Households with school going children 0.74
Households living in North East Nigeria 0.17
Distance to road (km) 5.31 -
Livelihood (income) sources during the last 12 months12Farming /agriculture
0.77 -
Non-farm business 0.64 -
Wage employment 0.34 -
Remittances and assistances 0.38 -
No. observations 1,906 1,906
Source: Authors' calculations based on Nigeria LSMS-ISA 2019 and 2020 rounds. Sample weights have been applied.
2.3. Definition of variables and descriptive results
Outcome variables
Food insecurity indicators: We measure food insecurity using three indicators, capturing
households' experience of food insecurity. In both rounds, households' food insecurity experience
are elicited using the self-reported experience of
hunger and food shortage in the last 30 days
(Hoddinott, 1999; Carletto et al., 2013; Bellemare and Novak, 2017). The first indicator asks if a
household head or any other adult in
the household had to
skip a meal because there was not
enough money or other resources to
get food. The second indicator elicits whether the household
has run out of food and takes a value of 1 if the household ran out of food because there was not
enough money or other resources to get food. Finally, the third indicator takes a value of 1 if the
household or any other adult in
the household went without eating for a whole day because of a
lack of money or other resources.
12
We note that households were asked to
mention multiple sources of livelihood and hence choices are not mutually
exclusive.
9
Labor market participation: The 2019 and 2020 surveys collect information on households'
participation in
income-generating activities over the last seven days. The major income
generating activities include farming, non-farm business, and wage-related activities. We
thus can
measure and quantify changes in labor allocation across both rounds. We
define an
indicator
variable for farming activities, which takes a value of 1 if the household head or any member of
the household worked on a household farm growing crops, raising livestock, or fishing, and 0
otherwise. Similarly, we define an
indicator variable for non-farm business, which takes a value
of 1 if
the household head or any member of the
household operated family business and zero
otherwise. Both farm and non-farm activities are observed at the household level. We also generate
an indicator variable for participation
in wage-related activities (observed
at the individual level),
which assumes a value of 1 if the household head or any other member of the household did work
wage job, either at their place of work or from home, and 0 otherwise. We also generate an
indicator variable for participation in any economic activity which assumes a value of 1 if the
household head or any member of the household participated in any of the above economic
activities, and zero otherwise
Food consumer price index (CPI): The Food ConsumerPrice Index (CPI) weemploy in this study
is collected and constructed by the Nigeria Bureau of Statistics (NBS), which measures the average
change in
prices over time consumers pay for a basket of food items. Food CPI measures changes
in the retail prices of food items and
is the principal indicator of changes
in retail food prices. It is
used to
measure consumer inflation in Nigeria's economy. We use food CPI for May 2019 and
May 2020, corresponding to
the both survey rounds we employ in
this study.
Table 2 reports key outcome variables: households' food security and labor market
participation rates in
both rounds. The results in
Table 2 show significant increases in
all food
insecurity indicators. For example, households' food insecurity experiences, as
measured by
incidence of skipping a meal, running out of food, and going without eating in the last 30 days
have increased by 47, 32, and 20 percentage points, respectively. Our empirical estimations
explore whether these changes and increases in food insecurity can
be attributed to COVID-19 and
associated mobility restrictions. On the other hand, participation in
income-generating activities
significantly reduced in
the post-COVID-19 round, while the food consumer price index increased
substantially.
10
Table 2: Descriptive results of key
outcome variables
Pre-COVID-19
(2019)
DifferencePost-COVID-19
(2020) test
Food security indicators
Skip a meal 0.26 0.73 0.47***
Run out of food 0.25 0.57 0.32***
Went without eating for a whole day 0.05 0.24 0.20***
Labor market participation
Farm activities 0.65 0.45 -0.19***
Non-farm business activities 0.57 0.37 -0.20***
Wage employment 0.27 0.11 -0.15***
Work in any activity 0.95 0.69 -0.26***
Food consumer price index (CPI) 289.98 359.59 69.61***
No. observations 1,906 1,906
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds. Values are weighted using the
sampling weights discussed above.
Notes: Food security indicators are measured as household-level responses to
a question that elicits food insecurity
experienced in the last 30 days. Labor market participation indicators take a value of 1 if any adult member of
the household reported labor allocation for that category of activity within the last 7 days. The food consumer
price index is computed at the state level.
1 1
Households were also asked about the impact of the pandemic on major livelihood sources.
These are self-assessed subjective indicators, but they can provide suggestive evidence on the
differential sectoral (livelihood) impacts of the pandemic, which can complement our forthcoming
difference-in-difference estimations. Households were asked for major sources of livelihood in the
last 12 months and changes in
associated income since the outbreak of COVID-19. As shown in
Figure 1, 72 percent of households reported that
their income from farming and agricultural
activities has reduced, 83 percent of households reported a reduction in
income from non-farm
businesses, and about half of them report reductions in
wage-related incomes. These suggest that
non-farm businesses are the most affected, and wage-related activities are relatively least affected.
This is
not surprising as
some wage-related activities are likely to be under formal contractual
agreements, and some of these activities may
be performed remotely and hence less affected by
mobility restrictions.
Figure 1: Changes in income by sources since the outbreak of the pandemic. These statistics are
adjusted for sampling weights.
1
0.9
0.8
0.53 0.7
0.72 0.6
0.83 0.77 0.78
0.5
0.4
0.3
0.43
0.2 0.19
0.17 0.1
0.09 0
0.11
0.06
0.19
0.04 0.04
Non-farm business Wage employmentincome Remittances&
income assistances income
0.05
Total incomeFarm income
Increased Stayed the same Reduced
Key explanatory variables: state-level COVID-19 cases and government lockdowns
We compiled the COVID-19 cases and lockdown measures from the Nigerian Centre for Disease
Control (NCDC) (NCDC 2020; IFPRI. 2020). As our post-COVID-19 survey was fielded in April
and May 2020, we extract confirmed COVID-19 cases until the end of May 2020.
12
We compile government measures based on policy announcements by Federal and State
Governments of Nigeria (FGN. 2020a 2020b; NCDC. 2020). We
focus on the strictest mobility
restrictions, defining an
indicator variable that takes a value of 1 for those states introducing
lockdown measures to
contain the spread of the virus, while those states which did not introduce
lockdown measures take a value of 0.13 Thus, our main explanatory variables of interest are the
number of COVID-19 cases and an indicator variable for those states who introduced lockdown
measures to
contain the spread of the pandemic.14
The average state-level COVID-19 cases (at the end of May) is
about 222, and about 22
percent of the states have imposed lockdown restrictions. Figure 2 presents the geographic
distribution of confirmed COVID-19 cases and lockdown interventions (measured at
the state
levels) across states in
Nigeria. As expected, federal and state-level governments are likely to
introduce lockdown measures with increasing confirmed COVID-19 cases. However, some states
with a high level of COVID-19 cases have abstained from introducing lockdown measures while
some other states with low COVID-19 cases have announced lockdown measures, variations we
exploit in some of our estimations.
13 Nigeria has 36 states and one federal territory (the Federal Capital Territory). For simplicity, we refer to all these
as 37 states. All states are included in the analysis.
14 We also construct an indicator variable assuming a value of 1 for states above the median confirmed COVID-19
case and 0 for those states below the median COVID-19 case in
our sample.
13
Figure 2: Confirmed COVID-19 cases Lockdown restrictions by states
Source: Federal Government of Nigeria (2020) and Nigeria Center for Disease Control (NCDC, 2020).
"Heterogeneity" variables:
To better understand the differential impacts of COVID-19 cases and associated lockdown
measures on households' food security and labor market participation rates, we employ baseline
characteristics of households to differentiate "vulnerable" households and livelihoods. As the
impacts of the pandemic are likely to vary across households, we aim to uncover heterogeneous
impacts across various groups, especially those deemed to
be vulnerable households and regions.
The availability of baseline surveys allows us to
estimate the impact of the pandemic across various
socioeconomic groups and regions (see the mean value of these variables in
Table 1). For instance,
we explore potential differential impacts across rural and urban households as well as
across poor
and non-poor households. We
also classify households in remote and more accessible areas as
well
as across households living
in conflicted affected and other states. Households with school-going
14
children may experience further deterioration in
food security due to
the nationwide school
closures and associated school-feeding programs. To test this hypothesis, we estimate differential
impacts for those households withand without school-going children. We
also construct indicators
of household sources of livelihood in
the past twelve months, including farming/agriculture, non
farm business, wage employment, remittances, and
assistances.15 We
then estimate heterogeneous
responses and impacts across livelihood options.
3. Empirical Strategy
To quantify the impact of COVID-19 on households' food insecurity (our main outcome of
interest) as
well as
labor market participation and food prices (our intermediate outcomes of
interest), we exploit spatial variations in
the spread of the pandemic across states in Nigeria, along
with the temporal variations in our outcomes of interest. We
specifically estimate the following
fixed effects specification to
quantify the impact of COVID-19:
����ℎ����= ����ℎ+ ����0��������������������+ ����1������������������������∗ ��������������������+ ����ℎ���� (1)
where ����ℎ����stands for food insecurity and labor market outcomes for each household h and round
t.16 ����ℎcaptures household fixed effects, Cases represent the number of confirmed COVID-19
cases for each
state, which is expressed
in absolute numbers
as well as, in alternative specifications,
per million population in
each state. ��������������������is a dummy variable, assuming a value of 1 for the post
COVID-19 round and 0 for the pre-COVID-19 round. The parameter associated with this round
dummy captures aggregate trends in food security and labor market outcomes. This variable also
captures aggregate potential differences in
our outcomes of interest driven by differences in survey
methods (face-to-face or phone survey). ����ℎ����is an
error term that is assumed to
be uncorrelated
with COVID-19 cases, at least conditional on household fixed effects and state-level policy
responses. The household fixed effects in
equation (1) capture time-invariant heterogeneities
across households. The specification in
equation (1) is
a standard difference-in-difference
approach, except that our treatment intensity variable is continuous.
15
Remittances and assistances are defined as remittances and assistances received from inside the country or foreign
sources.
16
For the food price estimation, the unit of analysis is the state, and we control for state fixed effects.
15
Our identifying variation in
equation (1) comes from a combination of spatial variations in
COVID-19 and temporal variations in
our outcome of interest. The interaction term, between
COVID-19 cases and post-COVID-19 round dummy, captures differential temporal evolution in
our outcome of interest across states with varying exposure to the pandemic. We hypothesize that
those states experiencing a higher intensity of the
pandemic are more likely to
witness a higher
reduction in labor market participation and a higher increase in
food insecurity. Thus, the
estimation in
equation (1) entails comparing the temporal evolution of food security and labor
market outcomes for those states with high and low
exposure to
the pandemic.
Potential temporal variations in food security and labor market participation rates are
likely to
be driven by both government responses to
the pandemic as
well as
household-level
responses associated with precautionary measures to
reduce the contraction of the virus. The
economic repercussions of the pandemic are expected to
vary depending on individuals'
precautionary measures and state-level government responses (Abay et
al., 2020; Koren and Peto,
2020). In line with this, various states in Nigeria have imposed alternative forms of restrictions
and lockdowns, which are likely to
affect individuals' mobility and hence the labor market and
food security outcomes. To quantify the differential and compounding impact of these lockdown
measures, we estimate the following fixed effects specification:
����ℎ����= ����ℎ+ ����0��������������������+ ����1������������������������������������∗ ��������������������+ ����ℎ���� (2)
where ������������������������������������now stands for a dummy variable indicating for the introduction of
lockdown measures.17 To gauge and identify the
relative impact of government measures and
individually-driven precautionary measures driven by the spread of the pandemic, we further
interact the spread of the pandemic with lockdown measures. More specifically, we create an
indicator variable for states recording above-median cases and interact this with lockdown
measures. However, the breadthand implementation of these lockdown measures are likely to
vary
across states. Thus, although such an
exercise can
give us some latitude to identify the relative
impacts of the spread of the pandemic and government-induced restrictions, such results can only
provide suggestive evidence.
The impacts of the pandemic are likely to
vary across households with varying
socioeconomic status, livelihood options, and underlying conditions. We,
thus, aim to uncover the
17
Wenote that as the number of COVID-19 cases are strongly correlated with government responses to
the pandemic,
we cannot control for both COVID-19 cases and government measures in the same specification.
16
potential differential impact of COVID-19 across various groups of households. In particular, the
impacts are expected to be higher among those households and regions deemed to be vulnerable,
including poor households, those households with school-going children and those living in
conflict-affected and remote zones. Using baseline information on households' residence,
socioeconomic status, and livelihood options, we quantify the differential impact of the pandemic
on households' food security and labor market participation using the following empirical
specification:
����ℎ����= ����ℎ+ ����0��������������������+ ����1��������������������∗ ����������������������������������������ℎ∗ ��������������������+ ����ℎ���� (3)
where all
the terms, except the term "Vulnerable", are as
defined above. Our vulnerable group of
households includes poorer, households with school children, urban households and those living
in remote and conflict-affected zones and neighborhoods. ����1
in equation (3) capture differential
trends in food security and labor market outcomes of those deemed "vulnerable" households,
which can
be attributed to
the spread of the pandemic and associated lockdown restrictions.
We also examine potential differential impacts across households with varying exposure
to the pandemic because of their livelihood strategies and sectoral engagement in labor markets.
For example, some sectors are likely to experience a disproportionally higher impact associated
with social distancing andlockdownmeasures. Forexample, several recent economywide analyses
of the impact of the pandemic show that services are the most affected sectors (e.g., Breisinger et
al., 2020). Abay et
al. (2020) show that those sectors and services involving face-to-face
interactions experience much higher loss in
demand for services, while those services meant to
substitute personal interactions (e.g., ICT services) enjoy a significant boost in
demand.
Traditional small non-farm businesses in
Africa are likely to involve personal interactions and
hence may be more affected than those activities that can be performed remotely. Similarly, rural
activities might be less prone to the spread of the pandemic and associated lockdown measures for
several reasons. First, the spread of the pandemic is
likely to be higher among urban areas. Second,
government responses and restrictions are expected to
be more strict and intense among urban
areas. Third, urban food systems and value chains are likely to
be more affected by short-term
shocks than rural livelihoods. We
thus estimate the following empirical specification to
quantify
the differential impact of the pandemic across livelihood options.
����ℎ����= ����ℎ+ ����0��������������������+ ����1��������������������∗ ������������������������ℎ������������ℎ∗ ��������������������+ ����ℎ���� (4)
17
Where all notations except "Livelihood" are as
defined above. As shown in
Table 1, households’
livelihood options and sources of income in
our sample include farming (agriculture), non-farm
business, wage-employment, and remittances and assistances.
To account for systematic non-response in
the post-COVID-19 phone survey, we weighted
all our estimates by the sampling weight associated with the LSMS-ISA phone survey data.18 This
weighting procedure enables recovering unbiased and representative statistics under the
assumption that data are “missing at random” conditional on some observable factors that are
accounted in the construction of weights (e.g., Wooldridge, 2007; Korinek et al., 2007).As we are
following households across rounds, error terms are expected to be correlated across time. We,
thus, cluster standard errors at the household level.
4. Results
4.1. Food security impacts
In this section, we present estimation results on
the impact of the pandemic and associated
lockdowns on food security, corresponding to
equations (1) and (2). Table 3 shows the impact of
COVID-19 on food security outcomes, measured as
binary indicators of food insecurity
experience.19 The number of reported COVID-19 cases for each state are transformed using an
inverse hyperbolic sine transformation, to
accommodate those few states with zero reported
cases.20
The interaction between COVID-19 cases and the post-COVID-19 dummy captures the
temporal variation in
the evolution of our outcomes of interest associated with varying exposure
to the spread of the pandemic. A positive and significant impact shows that states registering higher
numbers of COVID-19 cases are likely to
experience greater increases in the probability of food
insecurity, relative to the pre-COVID-19 period. The coefficients in Table 3 show that doubling
the number of COVID-19 cases is associated with a 2.4-2.7 percentage point increase in
the
18
A discussion on the construction of these sampling weights are given in NBS and World Bank (2020).
19Because of the lownumber of zero-valued cases in
theinfection rates, the inverse hyperbolic sine transformed values
differ very little from log transformed values and can be effectively interpreted in the same manner. The only state
which didn’t report COVID-19 cases in May 2020 was Cross Rivers.
20
As we have large positive values of COVID-19 cases for most states, such a transformation is
expected to be
innocuous (e.g., Bellemare and Wichman, 2019).
18
probability that a household ran out of food or skipped a meal in the last 30 days.21 The size of the
impact is
plausible, although we expect significant heterogeneities across different types of
households and contexts, an
empirical question we
address in the next section.22
Table 3: Impact of COVID-19 cases on household food security outcomes
0.369***
(1) (2) (3)
Skip a meal Ran out offood Went without eating for a whole
day
Post dummy (2020 round) 0.217*** 0.072
(0.052) (0.054) (0.050)
COVID-19 cases*Post 0.025** 0.024** 0.027***
(0.011) (0.011) (0.010)
Constant 0.262*** 0.251*** 0.056***
(0.010) (0.010) (0.009)
Household fixed effects Yes Yes Yes
R-squared 0.39 0.22 0.15
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria's LSMS-ISA 2018-19 and 2020 rounds.
Notes: All estimation results are adjusted by sampling weights accounting for systematic non-response in the phone
survey. The number of confirmed COVID-19 cases are transformed using an inverse hyperbolic sine
transformation to accommodate one state with zero case. Standard errors, clustered at the household level,
are given in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
The impacts shown in Table 3 are likely to
be compounded by national and state-level
government responses to
the pandemic, which included social distancing and mobility restrictions
as well
as partial and complete lockdown measures.
We thus estimate the empirical specification
in equation (2) to quantify the implication of variations
in state-level responses to the pandemic.
We mainly focus on the strictest mobility restrictions and hence generate
an indicator variable for
states introducing lockdown measures. We then compare temporal evolutions in
food security
outcomes across states with and without lockdown measures. Table 4 generally shows that
lockdowns increase food insecurity. For example, we find that state-level lockdowns increase the
probability that a household skips a meal in
the last 30 days by 12 percentage points.
21
We note that, as the spread of the pandemic remains fast globally, doubling of COVID-19 cases takes only a few
weeks (in some cases less than a week) in many countries, including in
Nigeria.
22
When we use the number of infections per 1 million inhabitants, the results are qualitatively the same. Results are
reported in the Appendix (1A-5A).
19
Table 4: Government responses and food security indicators
0.046***
(1) (2) (3)
Skip a meal Ran out of food Went without eating
for a whole day
Post dummy(2020 round) 0.467*** 0.315*** 0.194***
(0.016) (0.016) (0.013)
Lockdown*Post 0.121*** 0.067* 0.024
(0.039) (0.040) (0.020)
Constant 0.256*** 0.250***(0.007) (0.007)
(0.005)
Household fixed effects Yes Yes Yes
R-squared 0.41 0.27 0.19
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds.
Note: Dependent variables are as defined in Table 2. All estimations are adjusted by sampling weights accounting for
non-response in the phone survey. Standard errors, clustered at the household level, are given in
parentheses.
Lockdown is an indicator variable taking a value of 1 for those states which introduced lockdown measures to
contain the spread of the virus. Standard errors, clustered at the household level, are given in parentheses. * p
< 0.10, ** p < 0.05, *** p < 0.01.
To jointly examine the effects of infection rates and lockdowns, we also interact the
indicator variables for the spread of the pandemic with lockdown measures. To facilitate this, we
construct an
indicator variable assuming a value of 1 for states above the median confirmed
COVID-19 case and 0 for those states belowthe median COVID-19 case in our sample. Interacting
these indicators gives us four groups: high COVID-19 cases with lockdown, high COVID-19 cases
without a lockdown, low COVID-19 cases with lockdown, and low COVID-19 cases without
lockdown. The estimation results are shown in
Table 5. As expected, households in
states
recording high COVID-19 cases and withlockdown measures are hit hardest and hence experience
the greatest increase in
food insecurity. Coefficient estimates suggest that both the spread of the
pandemic as
well as
government-induced lockdown measures are increasing food insecurity.
However, the former seems to dominate, given the
larger magnitudes of the estimates.
20
Table 5: Disentangling the impact of COVID-19 cases and government measures
(1) (2) (3)
Skip a meal Ran out of food Went without eating for a
whole day
Post dummy (2020 round) 0.423*** 0.267*** 0.137***
(0.030) (0.029) (0.026)
High COVID-19 cases*Lockdown*Post 0.154*** 0.156*** 0.057
(0.052) (0.053) (0.049)
High COVID-19 cases*No-lockdown*Post 0.094** 0.112*** 0.122***
(0.044) (0.043) (0.040)
Low COVID-19 cases*Lockdown*Post -0.047 -0.110 0.061
(0.086) (0.086) (0.053)
Constant 0.262*** 0.251*** 0.056***
(0.009) (0.010) (0.008)
Household fixed effects Yes Yes Yes
R-squared 0.41 0.26 0.16
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: "High" and "Low" COVID-19 cases are defined as above and below the median confirmed values in our sample,
respectively. All estimations are adjusted by sampling weights for accounting non-response in the phone
survey. The number of confirmed COVID-19 cases is
transformed using inverse hyperbolic sine transformation
to keep zero cases for one state. Lockdown stands for indicator variables for those states who introduced
lockdown measures to
contain the spread of the virus. Standard errors, clustered at the household level, are
given in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
4.2. Mechanisms and Intermediate outcomes
The impact of COVID-19 on labor market participation
Reduction in
income is
one of the most important mechanisms through which the COVID-19
pandemic can
affect food insecurity. Results in
Table 6 show the implication of the spread of the
pandemic on labor market participation rates. As expected, the spread of the pandemic is
associated
with a significant reduction in
economic activity. The interaction term in column 1 of Table 6
shows that doubling the number of COVID-19 cases is associated with a 2 percentage points
reduction in
the probability of participation in any
economic activity (in the last seven days). The
second column presents impacts on-farm activities, while the third and fourth columns report
impacts on non-farm business and wage-related activities. Overall, these results imply that
households in
areas with a higher degree of exposure to
the pandemic have experienced significant
reductions in
economic engagement. Wage-related activities are the least affected, probably
because some of these activities can
be performed remotely (e.g., Dingel and Neiman,2020) and/or
are under formal contracts. As shown in
our data (not reported here for lack of space), the majority
of wage workers are employed in
the public sector and non-governmental organizations, which
are less likely to
fire employees and allow employees operate some of their functions in
some
21
form. These findings are consistent with global evidence on the differential sectoral impact of the
pandemic (e.g., Dingel and Neiman, 2020; Abay et
al., 2020).
Table 6: Impact of COVID-19 cases on participation in
economic activities
(1) (2) (3) (4)
Working any Farm Non-farm
activity activities
Wage
business employment
Post dummy (2020 round) -0.459*** -0.056 -0.214*** -0.073*
(0.052) (0.056) (0.063) (0.043)
COVID-19 cases*Post -0.021** -0.033*** -0.016* -0.013
(0.010) (0.012) (0.009) (0.009)
Constant 0.958*** 0.676*** 0.662*** 0.235***
(0.009) (0.010) (0.008) (0.007)
Household fixed effects Yes Yes Yes Yes
R-squared 0.54 0.18 0.22 0.26
No. observations 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for few states. Standard errors, clustered at the household level, are given in parentheses. * p < 0.10, **
p <
0.05, *** p < 0.01.
Similar to the food security results, the results shown above are likely to
be compounded
by government lockdown measures. Table 7 shows that lockdowns limit economic activities and
hence households' participation in labor market activities. In particular, lockdowns are associated
with a 6 percentage point reduction in
households' participation in
any economic activity.
Interestingly, state-level lockdown measures are more impactful in
reducing non-farm business
activities. As expected, farm activities appear to
be less affected by state-level lockdownmeasures,
perhaps due to
the following important reasons. First, farm activities require limited mobility and
face-to-face interactions. Second, lockdown measures are likely to
be strict and effectively
implemented in
urban than in
rural areas. It is noteworthy to stress that results
can be reinforcing
across sectors, not only because sectors (at the demand and supply level) are interconnected, but
also because households may
draw their revenue from different sources. For example, due to the
pandemic, rural households may be impeded to diversify their income by moving off the farm to
wholesalers.
2 2
Table 7: Government responses and participation in economic activities
-0.515***
(1) (2) (3) (4)
Working any Farm
activity activities
Non-farm Wage
business employment
Post dummy (2020 round) -0.219*** -0.230*** -0.121***
(0.014) (0.016) (0.017) (0.013)
Lockdown*Post -0.069* -0.004 -0.122*** -0.062*
(0.040) (0.044) (0.046) (0.037)
Cons 0.955*** 0.645*** 0.640*** 0.266***
(0.006) (0.007) (0.007) (0.005)
Household fixed effects Yes Yes Yes Yes
R-squared 0.54 0.14 0.19 0.25
No. observations 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds.
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. Lockdown
stands for indicator variables for those states who introduced lockdown measures to contain the spread of the
virus. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, ** p < 0.05, *** p <
0.01.
The impact of COVID-19 on food prices
As discussed in
Béné (2020), COVID has various adverse effects on local food systems' agents
and is likely to have negative impacts on food security. These effects include, but are not limited
to, the disruption in
inputs’ supply chain, the drop in
the demand of certain food commodities, the
reduction in
workers availability, and the disruption in
transportation (of inputs and final products),
as well
as the effect on food retailers and vendors' activities. All these inevitably have
an upward
effect on food consumer prices. The food price increase is another likely supply-side channel
through which COVID-19 can affect food insecurity (e.g., Mogues, 2020). We used the pre
COVID-19 food consumer price index (May-2019) and post-COVID-19 food consumer index
(May-2020) to
quantify the effects of the pandemic on the change in
the consumer price index, as
well of the COVID-19 related change in food prices on household food security. As reported in
Table 8, the country experiences an increase in
food price due to
the pandemic, both when related
to infection rates and
to lockdown measures. In Table 9, we also show that pandemic-related food
price increase harms food security, irrespective of
the food security indicator that is
used.
2 3
0.212***
Table 8: Impact of COVID-19 cases and lockdown on the food consumer price index
(1) (2) (3)
CPI CPI CPI
Post dummy (2020 round) 0.201*** 0.203***
(0.002) (0.007) (0.007)
Lockdown*Post 0.011*** 0.008**
(0.003) (0.003)
COVID-19 Cases* Post 0.003** 0.002*
(0.001) (0.001)
Constant 6.362*** 6.362*** 6.362***
(0.001) (0.001) (0.001)
R-squared 0.76 0.81 0.88
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for one state. CPI values are log-transformed. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, ** p < 0.05, p < 0.01. ***
Table 9: COVID-19, Food Consumer Price Index and food security
Panel A:
Skip a meal Ran out of food Went without eating for a
whole day
(1) (2) (3) (4) (5) (6)
Post dummy (2020 round) 6.895 4.983 4.627 2.707 0.609 -1.630
(5.837) (5.950) (5.855) (5.949) (4.328) (4.463)
CPI -0.522 -0.673** -0.134 -0.294 -0.337 -0.498*
(0.323) (0.331) (0.333) (0.343) (0.268) (0.275)
CPI*Post -0.975 -0.701 -0.653 -0.378 -0.062 0.258
(0.888) (0.904) (0.891) (0.904) (0.659) (0.677)
COVID-19 Cases*CPI*Post 0.004** 0.004** 0.004***
(0.002) (0.002) (0.002)
Constant 0.262*** 0.262*** 0.251*** 0.251*** 0.056*** 0.056***
(0.010) (0.010) (0.010) (0.010) (0.009) (0.008)
Household fixed effects Yes Yes Yes Yes Yes Yes
No. observations 3812 3812 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for few states. CPI values are log-transformed. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
4.3. Heterogeneous impacts
Differential household impacts of COVID-19 on food security and labor market participation
The impact of COVID-19 is
likely to vary across households due to
differences in
underlying
conditions and exposure to
the pandemic as
well as in
associated government measures and
responses. For instance, urban households are likely to
experience higher exposure to
the
pandemic, and hence they are likely to
experience reductions in
economic activities. Similarly,
2 4
poorer households' and those in remote areas and conflict zones could see further deterioration in
food insecurity because of disruptions in local and national transportation systems and markets.
Such heterogeneous impacts may also vary by type of outcome. For instance, while urban
households are likely to
experience reductions in
economic activities, poorer and remotely located
households may be more likely to
face food security challenges. For this purpose, we define
indicator variables for urban and poor households, those in
remote (measured by the distance to
the main road) and conflict-affected areas, and those with school-going children. These variables
are interacted with state-level COVID-19 cases and lockdown indicators to
quantify the
differential impact of the spread of the pandemic and associated lockdown measures.
Results, presented in Table 10, show that
those households living in remote areas and
conflict-affected North-East Nigeria (Yobe, Borno, Bauchi,Gombe,Taraba, and Adamawa States)
are more likely to
experience deterioration in
food security. On the other hand, although urban
households reduce economic activities (as we show in
Table 11), they do not suffer from
significant reductions in food security. This finding is probably because of better underlying food
security and improved access to markets. On the other hand, poorer households experience
significant increases in
all indicators of food insecurity, although some of their activities are not
meaningfully affected by state-level lockdowns (Table 11). This is
likely driven by differences in
responses to
lockdown measures between poor and non-poor households. Consistent with this
argument, Bargain and Aminjonov (2020) use Google mobility data to document that mobility
reductions are relatively smaller in
poor neighborhoods in developing countries, possibly because
poorer households in
such settings are less able to
afford the costs of reduced mobility in
compliance with government restrictions.
Our results also show that households with school attending children, who are likely to
miss government school meals because of school closure, are likely to
face food security
challenges. This is anticipated in the context of Nigeria, where about 7.5 million pupils in
46,000
schools are enrolled in the national school feeding program (World Bank, 2020c). The closure of
schools is
likely to
reduce the food intake of these children, potentially exacerbating any negative
effects on their human capital development.
25
Table 10: Differential impact of COVID-19 on household food security
Panel A: Differential impact of COVID-19 cases on household food security
(1) (2) (3)
Skip a meal Ran out of food Went without eating for a
whole day
Post dummy (2020 round) 0.481*** 0.320*** 0.186***
(0.021) (0.022) (0.019)
COVID-19 cases*Urban*Post 0.006 0.017 0.021
(0.014) (0.015) (0.014)
COVID-19 cases*log(Distance to road)*Post 0.007 0.010** 0.009**
(0.004) (0.004) (0.004)
COVID-19 cases*Asset poor tercile*Post 0.048*** 0.029* 0.031*
(0.015) (0.017) (0.017)
COVID-19 cases*North East Zone*Post 0.011 -0.000 0.042**
(0.017) (0.016) (0.017)
COVID-19 cases*School children*Post 0.021** 0.018* 0.023***
(0.010) (0.011) (0.009)
Constant 0.262*** 0.251*** 0.056***
(0.010) (0.010) (0.009)
Household fixed effects Yes Yes Yes
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for few states. Lockdown stands for indicator variables for those states who introduced lockdown measures to
contain the spread of the virus. Standard errors, clustered at the household level, are given in parentheses. * p <
0.10, **
p < 0.05, p < 0.01. ***
The results in Table 11 provide differential impacts of COVID-19 cases and associated
lockdown measures on households' labor market participation in
economic activities. These results
indicate that urban households and those households located in remote areas and conflict-affected
areas of North East Nigeria are likely to experience significant reductions in
economic activities.
For example, urban households are more likely to
experience a reduction in economic activities,
despite significant variations across various types of economic activities. Households in
urban
areas reduce non-farm business and wage-related activities while increasing farm activities. This
implies that the pandemic can also lead to the reallocation of labor resources across alternative
economic activities and sectors of economies. Similarly, households in
remote areas and those in
conflicted-affected areas are disproportionally affected by the COVID-19 crisis and hence reduce
all forms of economic activities. These patterns are consistently observed when using both
COVID-19 cases and lockdown measures.
26
Table 11: Differential impacts of COVID-19 cases on economic activities
Panel A: Differential impacts of COVID-19 cases on economic activities
(1) (2) (3) (4)
Working any Farm
activity activities
Non-farm Wage
business employment
Post dummy (2020 round) -0.530*** -0.285*** -0.230*** -0.203***
(0.021) (0.023) (0.025) (0.017)
COVID-19 cases*Urban*Post -0.018** 0.054*** -0.023** -0.025***
(0.008) (0.009) (0.009) (0.007)
COVID-19 cases*Distance to road*Post -0.007** -0.016*** -0.008** -0.001
(0.003) (0.004) (0.004) (0.003)
COVID-19 cases*Asset Poor*Post -0.024*** -0.025*** 0.010 -0.020***
(0.008) (0.009) (0.009) (0.006)
COVID-19 cases*North East Zone*Post -0.022*** -0.048*** -0.032*** -0.005
(0.009) (0.010) (0.011) (0.009)
Constant 0.958*** 0.676*** 0.602*** 0.238***
(0.009) (0.010) (0.010) (0.007)
Panel B: Differential impacts of government measures on economic activities
Post dummy (2020 round) -0.541*** -0.233*** -0.246*** -0.219***
(0.019) (0.021) (0.022) (0.015)
Lockdown*Urban* Post -0.129*** 0.215*** -0.143** -0.186***
(0.049) (0.052) (0.061) (0.048)
Lockdown*log(Distance to road)* Post -0.036* -0.024 -0.078*** -0.027*
(0.023) (0.026) (0.025) (0.014)
Lockdown*Asset poor tercile*Post -0.126* -0.074 -0.114 0.030
(0.078) (0.081) (0.080) (0.076)
Constant 0.958*** 0.676*** 0.602*** 0.238***
(0.009) (0.010) (0.010) (0.007)
Household fixed effects Yes Yes Yes Yes
No. observations 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds.
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for few states. Lockdown stands for indicator variables for those states who introduced lockdown measures to
contain the spread of the virus. Standard errors, clustered at the household level, are given in parentheses. * p <
0.10, **
p < 0.05, *** p < 0.01.
Differential impacts of COVID-19 on various livelihoods
In this section, we explore potentially heterogeneous impacts of the pandemic across households
with varying livelihoods and sources of incomes. Several studies from developed countries, where
administrative and transaction-level data are available, show that the pandemic has had
heterogeneous impacts on different livelihood options and sectors. For instance, livelihoods and
sectors that can operate on a remote basis with limited personal interactions or those functionally
27
dependent on the internet are likely to
be less affected, relative to
those involving personal
interactions (Dingel and Neiman, 2020; Abay et
al., 2020; Gilligan, 2020). Similarly, some
livelihood options and sectors are likely to
experience a relatively higher disruption in
economic
activities. For instance, government-imposed mobility restrictions and shutdowns often disrupt
supply chains, which may
prove the most challenging for small businesses with smaller stock.
Thus, those households relying on non-farm business activities are likely to experience
disproportionally higher impacts associated with disruptions in
value chains caused by the
pandemic and related mobility restrictions. Although not many rural activities in
Nigeria are
functionally dependent on the internet, some activities can
be operated without many personal
interactions with others and hence may be relatively less prone to these restrictions and lockdown
measures.
We hypothesize that households relying on alternative livelihood options and economic
sectors may
be relatively more resilient to
the shocks associated with pandemic. For instance, as
shown in Figure 1 and Table 6, those households relying on non-farm businesses, witness the
highest reduction in income and economic activities. On the other hand, wage-related activities
and income sources are least affected by lockdown measures (Figure 1 and Table 6). The
estimations in Table 12 probe these relationships further to explore heterogeneous impacts across
households' livelihood and income sources.
The results in
Table 12 consistently show that those households relying on on-farm
activities and non-farm businesses experience a significant increase in
food insecurity associated
with COVID-19 cases and associated lockdowns. Those households relying on wage, remittance,
and assistance income are not significantly affected by the pandemic and associated lockdowns.
This is
consistent with the self-assessed evidence from Figure 1. This is
not surprising as
some
wage-related activities may still be operated remotely, or individuals engaged in
wage-related
activities have longer-term contracts or savings that they can draw on during crises of this type.
For instance, more than half of the wage employees in
our data are employed in
government and
non-governmental organizations, which are less likely to fire employees.
28
Table 12: Differential impact of COVID-19 and
government measures on household food
security
Panel A: Differential impacts of COVID-19 cases on household food security
(1) (2) (3)
Skip a meal Ran out of food Went without eating for a
whole day
Post dummy (2020 round) 0.409*** 0.253*** 0.183***
(0.034) (0.037) (0.032)
COVID-19 cases*Farming* Post 0.020*** 0.021** 0.004
(0.008) (0.008) (0.007)
COVID-19 cases*Non-farm business* Post 0.016** 0.015** 0.015**
(0.007) (0.008) (0.007)
COVID-19 cases*Wage employment* Post -0.018** -0.003 -0.017**
(0.008) (0.008) (0.008)
COVID-19 cases*Remittances and assistances*Post -0.001 0.004 -0.002
(0.008) (0.008) (0.007)
Constant 0.262*** 0.251*** 0.056***
(0.010) (0.010) (0.009)
Panel B: Differential impacts of government measures on household food security
Post dummy (2020 round) 0.466*** 0.308*** 0.198***
(0.021) (0.021) (0.019)
Lockdown*Farming* Post 0.096* 0.115** 0.006
(0.054) (0.050) (0.046)
Lockdown*Non-farm business* Post 0.090** 0.060 0.059*
(0.043) (0.046) (0.035)
Lockdown*Wage employment* Post -0.005 0.040 -0.151**
(0.067) (0.074) (0.067)
Lockdown*Remittances & assistances*Post 0.029 0.004 -0.094**
(0.062) (0.079) (0.050)
Constant 0.262*** 0.251*** 0.056***
(0.010) (0.010) (0.009)
Household fixed effects Yes Yes Yes
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. Lockdown
stands for indicator variables for those states who introduced lockdown measures to contain the spread of the
virus. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, **
p < 0.05,0.01. ***
p <
29
5. Concluding remarks
This study employed recent nationally representative data from Nigeria to
document the ways in
which the COVID-19 pandemic has been affecting both urban and rural households' food security
outcomes. Our analysis suggests that the spread of
the pandemic, as
well as
governmental mobility
restrictions (i.e. "lockdowns"), have both had significant impacts on food security outcomes
reported by households in
our sample. Furthermore, our analysis shows clear effects of infection
rates and lockdowns on key intermediate channels, including restricted household economic
activity and increased local food prices. We
do not have enough information to
relate lockdown
restrictions to
the avoided (or delayed) number of
new cases, and so cannot directly speak to the
tradeoffs that government lockdown policies imply. However, our analysis indicates that there are
measurable food insecurity costs associated with infection rates, as
well as
the restrictions designed
to contain the spread of the pandemic. State-level lockdown measures reduced the probability of
participation in non-farm business activities by 13 percentage points and increased households'
experience of food insecurity by 12 percentage points. Our finding that government-imposed
lockdowns are increasing food insecurity is
consistent with a recent review of grey literature
indicating that the main food security impacts of the pandemic have been through lockdown and
mobility restrictions, with direct effects operating through income losses and reduced purchasing
power, particularly for the poorest households (Béné, 2020). This finding is directly relevant to
the debate on the aggregate social welfare and economic impacts of lockdown restrictions in low
and-middle-income countries, which has come under some criticism (e.g., Ravallion et al., 2020,
Mobarak and Barnett-Howell, 2020, Bargain and Aminjonov 2020).
Furthermore, we find that impacts vary considerably by household and geographic context.
Most of these results tally with expectations: for example, the food security outcomes of poorer
households, those households with school attending children, and those living in conflict-affected
areas are most sensitive to COVID-19 effects. For instance, lockdown measures are more
impactful in disrupting non-farm business activities, and those households relying on these
activities experience the highest reduction income and increase in
food insecurity. These lockdown
measures have smaller impacts on wage-related activities. It is
important to note that the current
analysis is measuring the short-term impacts of the pandemic, i.e., those occurring during the first
3 months of significant disruption. It is
not yet clear how these impacts may change over time, as
the epidemic continues to
play out over time and space. It is
possible that the estimated elasticities
30
of food insecurity, economic engagement, and food prices to
the spread of the pandemic and
lockdowns may
evolve over time. Nevertheless, income deterioration experienced by households
soon after the outbreak of the pandemic may have longer-term effects because of potential impacts
on agricultural inputs, health care, schooling and
other investments in the coming months. Our
results align with other evaluations of the urgency of effective social safety net expansion to
address such exacerbated vulnerabilities and mitigate against their longer term economic and
welfare implications (Amjath-Babu et
al., 2020; Béné, 2020).
This article contributes important new empirical analysis of the impacts of the pandemic at a
point where there is an
abundance of conceptual papers and opinion pieces but still scant evidence
on the actual economic and welfare impacts of the
pandemic, particularly in developing countries.
One of the policy implications of our study is the need to address social safety nets in rural areas
as well
as urban areas, which have been the focus of much of the discussion
in the region to date
(Gentilini et
al., 2020; Gilligan, 2020; Devereux, et al., 2020). These findings can
inform
immediate and medium-term social protection policies as
well as
help governments and
international donor agencies improve their targeting strategies to
identify the most impacted sub
populations.
31
References
Abay, K.A., K. Tafere, and A. Woldemichael. 2020. Winners and Losers from COVID-19: Global
Evidence from Google Search (English). Policy Research working paper; no. WPS 9268;
COVID-19 (Coronavirus). Washington, D.C.: World Bank Group.
Akter, S. 2020. The impact of COVID-19 related 'stay-at-home' restrictions on food prices in
Europe: findings from a preliminary analysis. Food Security (2020): 1-7.
Amjath-Babu,T. S., Timothy J. Krupnik, Shakuntala H. Thilsted, and Andrew
J. McDonald. 2020.
Key indicators for monitoring food system disruptions caused by the COVID-19 pandemic:
Insights from Bangladesh towards effective response. Food Security: 1-8.
Andam, K., Edeh, H., Oboh, V., Pauw, K., and J.
Thurlow. 2020. Impact of COVID-19 on
Production, Poverty and Food Systems. Washington, D.C.: International Food Policy
Research Institute (IFPRI).
Baldwin, R., and , B. Weder di Mauro. 2020. Mitigating the COVID Economic Crisis. London:
Centre for Economic and Policy Research.
Bargain, O., and U. Aminjonov. 2020. Between a Rock and a Hard Place: Poverty and COVID-19
in Developing Countries. IZA DP No. 13297.
Barnett-Howell, Z., & Mobarak, A. M. 2020. The Benefits and Costs of Social Distancing in
Rich
and Poor Countries. https://arxiv.org/abs/2004.04867.
Barrett, C. B. 2020. Actions now can
curb food systems fallout from COVID-19. Nature Food, 1–
2.
Barrett, C., Reardon, T., Swinnen, J., & Zilbermam, D. 2019. Structural Transformation and
Economic Development: Insights from the Agri-food Value Chain Revolution. Working
paper, Dyson School of Applied Economics and Management, Cornell University
Bellemare, M. F., & Novak, L. 2017. Contract farming and food security. American Journal of
Agricultural Economics. 99(2): 357–378.
Bellemare, M.F., and C. J.
Wichman. 2020. Elasticities and the inverse hyperbolic sine
transformation. Oxford Bulletin of Economics and Statistics 82 (1): 50–61.
Béné, C.
2020. Resilience of local food systems and links to
food security–A review of some
important concepts in
the context of COVID-19 and other shocks. Forthcoming in Food
Security.
Breisinger, C., M. Raouf, M. Wiebelt, A. Kamaly, and M. Karara. 2020. Impact of COVID-19 on
the Egyptian economy: Economic sectors, jobs, and households. International Food Policy
Research Institute (IFPRI): Regional Program Policy Note 06.
Carletto, C., Zezza, A., & Banerjee, R. 2013. Towards better measurement of household food
security: Harmonizing indicators and the role of household surveys. Global Food Security,
2(1), 30–40
Devereux, S., C. Béné, and J.
Hoddinott. 2020. Conceptualizing COVID-19's impacts on
household food security. Food Security,1-4.
32
Dingel, J.
I., and B. Neiman. 2020. How many jobs can be done at home? Forthcoming, Journal
of Public Economics
Ericksen, P. J.
2008. Conceptualizing food systems for global environmental change research.
Global Environmental Change, 18, 234–245.
Ericksen, P.J., Bohle, H.-G., Stewart, B.
2010. Vulnerability and resilience of food systems. Food
Security and Global Environmental Change. Earth scan, London & Washington DC: 67–
77.
FAO. 2020. National Policy Responses to Limit the Impact of COVID-19 on Food Markets. Food
Policy and Market Developments. Rome, Italy: Food and Agriculture Organisation of the
United Nations (FAO). 2020. http://www.fao.org/giews/food-prices/food
policies/detail/en/c/1270543/.
FGN. 2020a. Address By His Excellency Muhammadu Buhari, President of the Federal Republic
of Nigeria, on the COVID-19 Pandemic, 29th March 2020. Abuja, Nigeria: Federal
Government of Nigeria (FGN).
FGN. 2020b. Address by His Excellency Muhammadu Buhari, President of the Federal Republic
of Nigeria, on the Extension of the COVID-19 Pandemic Lockdown, State House, Abuja,
Monday, 13th April 2020. Abuja, Nigeria: Federal Government of Nigeria (FGN).
FMBNP (Federal Ministry of Budget and National Planning). 2020. Ministerial Press Statement
on Fiscal Stimulus Measures in
Response to
the COVID-19 Pandemic & Oil Price Fiscal
Shock. Abuja, Nigeria: FMBNP.
https://statehouse.gov.ng/wpcontent/uploads/2020/04/HMFBNP-Final-Press-Statement-on
Responding-to-the-COVID-19- 06.04.2020-v.7.docx-1.pdf.
GAIN. 2020. Impact of COVID-19 on Food Systems: A Situation Report, Edition 3. May 13, 2020.
Global Alliance for Improved Nutrition, Geneva, Switzerland.
Gentilini, U., M. Almenfi, P. Dale, A.V. Lopez, U. Zafar. 2020. Social Protection and Jobs
Responses to
COVID-19: A Real-Time Review of Country Measures. Working paper
https://openknowledge.worldbank.org/handle/10986/33635
Gilligan, D. 2020. Social safety nets are crucial to
the COVID-19 response. Some lessons to boost
their effectiveness. International Food Policy Research Institute Blog
https://www.ifpri.org/blog/social-safety-nets-are-crucial-covid-19-response-some
lessons-boost-their-effectiveness (18 June 2020).
Global Hunger Index, 2019. The Challenge of Hunger and Climate Change: The Challenge of
Hunger and Climate Change. International Food Policy Research Institute, Washington,
DC, WHH (Welthungerhilfe), Bonn, and Concern Worldwide, Dublin.
Haddad, L., J.
Fanzo, S. Godfrey, C. Hawkes, S. Morris, and L. Neufeld. 2020. The COVID-19
crisis and food systems: Addressing threats, creating opportunities. Blogpost at
https://www.gainhealth.org/media/news/covid-19-crisis-and-food-systems-addressing
threats-creating-opportunities.
Hoddinott, J.
1999. Choosing outcome indicators of household food security. Washington, DC:
International Food Policy Research Institute.
33
IFPRI. 2020. COVID-19 Policy Response (CPR) Portal. https://www.ifpri.org/project/covid-19
policy-response-cpr-portal
Ihle, R., O. D. Rubin, Z. Bar-Nahum, and Jongeneel, R. 2020. Imperfect food markets in
times of
crisis: economic consequences of supply chain disruptions and fragmentation for local
market power and urban vulnerability. Forthcoming in Food Security.
IMF. 2020. Nigeria: Request for Purchase Under the Rapid Financing Instrument. International
Monetary Fund Country Report No. 20/142, Washington, D.C. 20090.
Koren, M., and R. Pető. 2020. Business disruptions from social distancing. The Centre for
Economic Policy Research (CEPR), COVID Economics vetted and real-time papers.
ISSUE 2 APRIL 8 2020.
Korinek, A., Mistiaen, J.
A., Ravallion, M., 2007. An Econometric Method of Correcting for Unit
Nonresponse Bias in
Surveys. Journal of Econometrics, 136(1), pages 213-235.
Mobarak, A. and Barnett-Howell, Z. 2020. Poor countries need to think twice about social
distancing, Foreign Policy. Available at: https://foreignpolicy.com/2020/04/10/poor
countries-social-distancing-coronavirus/
Mogues, T.
2020. Food Markets During COVID-19. Special Series on COVID-19. Fiscal Affairs.
NBS and World Bank. 2020. Nigeria COVID -19 National Longitudinal Phone Survey (COVID -
19 NLPS). http://documents1.worldbank.org/curated/en/717901591889288314/pdf/Basic
Information-Document.pdf
NCDC. 2020. First Case of Corona Virus Disease Confirmed in Nigeria. Abuja, Nigeria: Nigeria
Centre for Disease Control (NCDC). Downloaded June 27th 2020 from
https://ncdc.gov.ng/news/227/first-case-of-corona-virus-disease-confirmed-in-nigeria.
Ogundele, K. 2020. Updated: FG Places Travel Ban
on China, Italy, US, UK, Nine Others. The
Punch, March 18, 2020. https://punchng.com/breaking-fg-places-travel-ban-on-china
italy-us-uk-others/.
Ravallion, M. 2020. Pandemic policies in poor places. Center for Global Development,
Washington DC. https://www.cgdev.org/publication/pandemic-policies-poor-places
Reardon, T., M. F. Bellemare, and D. Zilberman, 2020a. How COVID-19 May Disrupt Food
Supply Chains in
Developing Countries. International Food Policy Research Institute Blog
https://www.ifpri.org/blog/how-covid-19-may-disrupt-food-supply-chains-developing
countries (2 April 2020).
Reardon, T., A. Mishra, C.
S. R.
Nuthalapati, M. F. Bellamare, and D. Zilberman. 2020b. COVID
19's disruption of India's transformed food supply chains. Economic and Political Weekly,
55(18).
Swinnen, J.
2020. "Will COVID-19 Cause Another Food Crisis? An Early Review," Issue Post,
April 10, 2020. International Food Policy Research Institute, Washington, DC.
Tendall, D.M., et
al. 2015. Food System Resilience: Defining the Concept. Global Food Security.
6, 17-23.
Torero, M. 2020. Without food, there can
be no exit from the pandemic: Countries must join forces
to avert a global food crisis from COVID-19. Nature, 580, 588–589.
on
34
United State Department of Agriculture (2016). Economics and Agricultural Statistics. 2015
https://www.ers.usda.gov/data-products/food-expenditure-series/
WFP. 2020a. COVID-19 Will Double Number of People Facing Food Crises Unless Swift Action
is Taken. Press Release, April 21, 2020 (Rome: World Food Program).
WFP. 2020b. WFP Global Response to COVID-19: June 2020. World Food Program, Rome.
https://docs.wfp.org/api/documents/WFP-0000117304/download/
Wright B.
D. 2011. The economics of grain price volatility. Applied Economic Perspective
Policy 33: 32–58.
Wooldridge, J. M. 2007. Inverse Probability Weighted Estimation for General Missing Data
Problems. Journal of Econometrics, 141(2): 1281-1301.
World Bank. 2020a.The impact of COVID-19 (Coronavirus)on global poverty:WhySub-Saharan
Africa might be the region hardest hit. Data Blog.
https://blogs.worldbank.org/opendata/impact-COVID-19-coronavirus-global-poverty
why-sub-saharan-africa-might-be-region-hardest.
World Bank. 2020b. Gender Dimensions of the COVID-19 Pandemic. Washington, D.C.: The
World Bank. https://openknowledge.worldbank.org/handle/10986/33622
World Bank. 2020c. World Bank. 2020. Nigeria Development Update: Nigeria in
Times of
COVID-19: Laying Foundations for a Strong Recovery. Washington DC: World Bank.
https://documents.worldbank.org/en/publication/documents
reports/documentdetail/695491593024516552/nigeria-in-times-of-covid-19-laying-foundations-
for-a-strong-recovery
Zurayk, R.
2020. Pandemic and Food security. Journal of Agriculture, Food Systems, and
Community Development, 9(3), 1–5.
35
Appendix
Table 1A: Impact of COVID-19cases (per 1 million inhabitants) on household food security outcomes
(1) (2) (3)
Skip a meal Ran out offood Went without eating for a whole
day
Post dummy (2020 round) 0.391*** 0.229*** 0.106***
(0.042) (0.042) (0.037)
COVID-19 cases*Post 0.031** 0.033*** 0.030***
(0.013) (0.013) (0.012)
Constant 0.262*** 0.251*** 0.056***
(0.010) (0.010) (0.008)
Household fixed effects Yes Yes Yes
R-squared 0.39 0.22 0.15
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria's LSMS-ISA 2018-19 and 2020 rounds.
Notes: All estimation results are adjusted by sampling weights accounting for systematic non-response in the phone
survey. The number of confirmed COVID-19 cases per million inhabitants is transformed using an inverse
hyperbolic sine transformation to
accommodate one state with zero cases. Standard errors, clustered at the
household level, are given in parentheses. * p < 0.10, ** p < 0.05, p < 0.01. ***
Table 2A: Disentangling the impact of COVID-19 cases (per 1 million inhabitants) and government
measures
(1) (2) (3)
Skip a meal Ran out of food Went without eating for a
whole day
Post dummy (2020 round) 0.423*** 0.267*** 0.137***
(0.030) (0.029) (0.026)
High COVID-19 cases*Lockdown*Post 0.164*** 0.166*** 0.061
(0.052) (0.053) (0.049)
High COVID-19 cases*No-lockdown*Post 0.097** 0.117*** 0.126***
(0.044) (0.043) (0.040)
Low COVID-19 cases*Lockdown*Post -0.046 -0.111 0.064
(0.086) (0.086) (0.053)
Constant 0.262*** 0.251*** 0.056***
(0.009) (0.010) (0.008)
Household fixed effects Yes Yes Yes
R-squared 0.41 0.26 0.16
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: "High" and "Low" COVID-19 cases are defined as above and below the median confirmed values in our sample,
respectively. All estimations are adjusted by sampling weights for accounting non-response in
the phone
survey. Lockdown stands for indicator variables for those states who introduced lockdown measures to contain
the spread of the virus. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10,p < 0.05, *** p < 0.01. **
36
Table 3A: Impact of COVID-19 cases on
participation in economic activities
(1) (2) (3) (4)
Working any Farm
activity activities
Non-farm Wage
business employment
Post dummy (2020 round) -0.512*** -0.084* -0.208*** -0.084**
(0.040) (0.043) (0.046) (0.036)
COVID-19 cases*Post -0.014** -0.042*** -0.017* -0.016
(0.007) (0.013) (0.009) (0.011)
Constant 0.958*** 0.676*** 0.602*** 0.238***
(0.009) (0.010) (0.010) (0.008)
Household fixed effects Yes Yes Yes Yes
R-squared 0.54 0.18 0.22 0.26
No. observations 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases per million inhabitants is
transformed using inverse hyperbolic sine
transformation to keep zero cases for few states. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
Table 4A: Impact of COVID-19 cases and lockdown on the food consumer price index
(1) (2) (3)
CPI CPI CPI
Post dummy (2020 round) 0.212*** 0.203*** 0.203***
(0.002) (0.005) (0.005)
Lockdown*Post 0.011*** 0.009***
(0.003) (0.003)
COVID-19 Cases* Post 0.004*** 0.003**
(0.001) (0.001)
Constant 6.362*** 6.362*** 6.362***
(0.001) (0.001) (0.001)
R-squared 0.76 0.81 0.88
No. observations 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases is transformed using inverse hyperbolic sine transformation to keep zero cases
for a few states. CPI values are log-transformed. Standard errors, clustered at the household level, are given in
parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
37
Table 5A: COVID-19, Food Consumer Price Index and food security
Panel A:
Skip a meal Ran out of food Went without eating for a
whole day
(1) (2) (3) (4) (5) (6)
Post dummy (2020 round) 4.240 1.058 3.944 0.659 -1.136 -4.356
(6.221) (6.374) (6.357) (6.548) (4.805) (5.078)
CPI -0.511 -0.665** -0.132 -0.290 -0.336 -0.491*
(0.325) (0.332) (0.333) (0.343) (0.268) (0.275)
CPI*Post -0.554 -0.081 -0.545 -0.056 0.214 0.693
(0.950) (0.972) (0.972) (1.000) (0.735) (0.775)
COVID-19 Cases*CPI*Post 0.005** 0.005*** 0.005***
(0.002) (0.002) (0.002)
Constant 3.515* 4.491** 1.089 2.096 2.195 3.183*
(2.068) (2.114) (2.120) (2.182) (1.707) (1.751)
Household fixed effects Yes Yes Yes Yes Yes Yes
No. observations 3812 3812 3812 3812 3812 3812
Source: Authors' calculations based on Nigeria LSMS-ISA 2018-19 and 2020 rounds
Note: All estimations are adjusted by sampling weights for accounting non-response in
the phone survey. The number
of confirmed COVID-19 cases per million inhabitants is
transformed using inverse hyperbolic sine
transformation to keep zero cases for one states. CPI values are log-transformed. Standard errors, clustered at
the household level, are given in parentheses. * p < 0.10, ** p < 0.05, *** p < 0.01.
38
ALL IFPRI DISCUSSION PAPERS
All discussion papers are available here
They can be downloaded free of charge
INTERNATIONAL FOOD POLICY RESEARCH INSTITUTE
www.ifpri.org
IFPRI HEADQUARTERS
1201 Eye Street, NW
Washington, DC 20005 USA
Tel.: +1-202-862-5600
Fax: +1-202-862-5606
Email: [email protected]
