UNDP MZ Socioeconomic Impact

THE SOCIOECONOMIC IMPACT OF
COVID-19 ON THE URBAN INFORMAL

ECONOMY IN MOZAMBIQUE
Results from a panel survey of informal sector operators in Maputo
Maputo 12th November 2021

Acknowledgments
The present report reflects over a year of work going into the implementation of the COVID-19
Informal Sector Survey (CISS), a collaboration between the United Nations Development Programme’s
(UNDP) country office in Mozambique and the Ministry of Economy and Finance, involving many
partners and people. In particular, the following people were instrumental in its making:
UNDP Mozambique
Alex Warren-Rodríguez, Senior Economist
Gustavo Novaes, Research Manager
Joana Sampainho, Programme Analyst - Economic and Policy Analysis
Nilifer Anaç, SDG Policy Analyst
Raphael Esteves, SDG Technical Specialist
Marta Lindstrom, Data Analytics and Management Specialist
Vania Come, Programme Associate
Ministry of Economy and Finance

Enilde Sarmento, Director, National Directorate of Economic and Development Policy (DNPED)
Maimuna Ibraimo, DNPED
José Cardoso, DNPED
ASSOTSI – AEIMO (Associação da Economia Informal de Moçambique)

Armindo Chembane, National Executive Secretary
IPSOS Mozambique
Jaime Fernandes, National Director
Kathleen Angulo, Senior Research Analyst
Rodolfo Albino, Research Assistant
Maxwell Odhiambo, Director of Operations
Alberto Mesa, Field coordinator
Arnaldo Bombe, Data Specialist
Nércia Magudo, Trainee
UNU Wider – Inclusive Growth Programme

Eva-Maria Egger, Research Fellow
Sam Jones, Research Fellow
Ricardo Santos, Research Fellow
The CISS survey would have not been possible without the active engagement of the 600 informal
workers affiliated with ASSOTSI who participated in the various rounds of the CISS surveys. A big
note of appreciation goes to them. It would also not have been possible without the encouragement
and support from Ms. Narjess Saidane and Mr. Francisco Roquette, UNDP Resident Representative
and UNDP Deputy Resident Representative in Mozambique, respectively.
Table of Contents
1 Introduction 4
2 COVID-19 and the informal economy – an international perspective 5
3 The impact of COVID-19 and the informal economy in Mozambique 7

3.1 The Socioeconomic impact of COVID-19 in Mozambique . . . . . . . . . . . . . . . . . 7
3.2 The informal economy in Mozambique . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4 The COVID-19 Informal Sector Survey (CISS) 11

4.1 Background and objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
4.2 Methodological considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5 Characteristics of the sample – a highly vulnerable segment of the population 13

5.1 Individual characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2 Business and employment characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.3 Characteristics of households in the sample . . . . . . . . . . . . . . . . . . . . . . . . 15
6 The socioeconomic impact of COVID-19 on the informal economy in Maputo 17

6.1 Impacts on informal economy businesses and informal employment . . . . . . . . . . . 17
6.2 Impacts on informal workers’ families . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
6.3 Understanding drivers of impacts and coping strategies in the CISS study . . . . . . . 27
7 Conclusion: Main findings and policy recommendations 30

7.1 Main findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
7.2 Policy recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
The COVID-19 Informal Sector Survey (CISS) 4
1. Introduction vey and three follow-up waves – implemented be-

tween August 2020 and March of 2021, the CISS
The urban informal economy has been identified survey has generated a rich set of data capturing
as one of the most exposed sectors to the effects how business and household conditions in the
of the socioeconomic impacts of the COVID- informal economy in Maputo have been affected
19 pandemic in both developed and developing by the socioeconomic impacts of the COVID-19
countries. Several reasons explain this. crisis, as well as the type of coping strategies
adopted by informal workers in response to it.
First, in many countries, informal sector activi-
ties are heavily concentrated in the services sec- The present report is organized as follows. Sec-
tor, a sector which has been directly affected by tion 2 provides an international perspective of
lock-down and social distancing restrictions im- how the COVID-19 pandemic is affecting the in-
posed by many governments around the world formal economy in countries around the world.
to reduce the propagation of COVID-19. This is followed in Section 3 by a brief review
of the socioeconomic impacts that COVID-19
Second, informal sector workers and informal en- has had so far in Mozambique, as well as an
terprises tend to have limited access to govern- overview of the informal sector in the country.
ment support programmes. In the case of infor- Section 4 presents the COVID-19 Informal Sec-
mal sector workers, because they often do not tor Survey (CISS), discussing key methodolog-
qualify for social protection programmes, which ical considerations regarding survey design and
typically target the most vulnerable and the implementation. Section 5 then goes on to de-
poorest segments of the population: the elderly, scribe the main characteristics of the sample,
people with disabilities, single mother-headed outlining key household and business traits that
households, or the ultra-poor, among others. In define informal sector operators in the sample.
the case of enterprises, because they lack regis- This is followed in Section 6 by the presenta-
tration or do not meet key business parameters tion of the main findings of the survey, in terms
(size or volume of business) required to qualify of how the COVID-19 pandemic has been af-
for business support programmes. fecting the business and lives of informal sec-
tor operators in Maputo. It looks at these is-
Finally, in developing countries, the urban in- sues, examining a variety of indicators captur-
formal sector is closely linked to the more for- ing household welfare and business dynamics. It
mal and modern parts of the economy (services, does this under the under-standing that socioe-
manufacturing, the public sector), which tends conomic crises such as the one caused by the
to be concentrated in urban areas. Informal op- COVID-19 pandemic affect people through a va-
erators are therefore more exposed to the dy- riety of channels. The last part of Section 6 also
namics of the global economy than are people explores in greater depth specific impacts and
whose livelihoods largely rely on subsistence- coping strategies, seeking to identify factors or
based primary activities taking place in rural sample characteristics associated with the effects
areas. of the COVID-19 crisis on unemployment, sav-
ings, borrowing, the sale of durable assets, and
The present report presents the results of a
household food consumption. Finally, Section 7
panel survey – the COVID-19 Informal Sec-
summarizes the main findings coming out of this
tor Survey, or CISS – undertaken among infor-
study and briefly discusses policy implications.
mal sector operators in Maputo, Mozambique.
Implemented by the United Nations Develop-
ment Programme’s (UNDP) Country Office in
Mozambique in partnership with the Mozambi-
can Ministry of Economy and Finance (MEF),
the CISS survey has sought to monitor the so-
cioeconomic effects that the pandemic has had
in the urban informal economy in Maputo. Con-
sisting of four rounds of surveys – a baseline sur-
2. COVID-19 and the informal econ- This share, however, varied significantly, both
omy – an international perspective across income levels and regions, ranging from
15 percent in advance economies in the 2010-
Informality is a prevalent feature in much of 2015 period, to 38.8 percent in low-income coun-
the developing world, with the informal econ- tries, and from 20.2 percent in Europe to 36.2
omy remaining a significant, often the highest, in Sub-Saharan Africa, during this same period
source of employment. The ILO estimates that (see Figure 1 for more detail). However, this
up to two billion people, equivalent to 61.2 per- share has been steadily falling since the early
cent of the world’s employed population, earn 1990s, this being true for all income levels and
their livelihoods in the informal economy, with regions, with the average contribution of the in-
93 percent of these living in emerging and devel- formal economy to global GDP dropping from
oping economies.1,2 In these countries, informal an estimated 35.1 percent in 1991 to 27.8 per-
employment accounts for most jobs, on average cent, in 2015.
69.6 percent, as opposed to only 18.3 percent of
employment in developed economies, with this Even though informal sector workers contribute
share rising to 85.8 percent in the Africa region. significantly to the economy, they usually have
limited access to social protection programmes,
The informal economy is also an important con- as well as to healthcare and medical treat-
tributor to GDP and thus to countries’ wealth. ment. This makes them particularly vulnerable
In their study of the shadow economy, Medina to health and economic shocks, as they have no
and Schneider (2018)3 estimate that informal cushion to fall back on in times of crisis.4,5
economy activities accounted on average for 27.8
percent of GDP in 2015 for the 158 countries The informal economy and COVID-19
considered in their analysis. COVID-19 has exacerbated this situation for
millions of informal sector workers and done so
Figure 1: Contribution of the Shadow Economy disproportionately.6 Without social protection,
to GDP (in %) they have only limited ability to cope with the
economic, social and health impacts of the pan-
demic. In a context where informality is of-
ten associated with limited access to medical re-
sources, as well as sanitation and hygiene facil-
ities, this makes informal workers particularly
vulnerable to the health consequences of the
pandemic. Additionally, because many informal
workers are not registered, they often don’t have
access to government benefits.7
Overall, ILO estimates that almost 1.6 billion in-

formal workers have been significantly impacted
by the pandemic, either because of lockdown
measures, or because these workers operated in
4
ILO (2020). “Extending social security to workers in
the informal economy: Lessons from international expe-
rience”. Geneva.
5
Horn, Z. (2009). “No Cushion to Fall Back On: The
1
These countries also account for most world employ- Global Economic Crisis and Informal Workers”. WIEGO
ment, although this share is somewhat smaller 82 per- Inclusive Cities.
6
cent. ILO (2021). “ILO Monitor: COVID-19 and the
2
ILO (2018). “Women and men in the informal econ- world of work”. Seventh edition. Geneva.
7
omy: a statistical picture” (3rd edition). Geneva. Yu, S. (2020). “How does informality aggravate the
3
Medina, L., Schneider, M. F. (2018). “Shadow impact of COVID-19?” Published on Data Blog, World
economies around the world: what did we learn over the Bank.
last 20 years?”. International Monetary Fund.
Table 1: Informal economy workers significantly losses and loss of income and livelihoods brought
impacted by lockdown measures (2020) about by the pandemic. A number of reasons ex-
plain this. Women are overrepresented in many
of the industries hardest hit by COVID-19, such
as food services, retail and entertainment. More-
over, as lockdown measures have kept people at
home, the burden of unpaid care and domes-
tic work has increased for women. Women have
also suffered in non-economic terms. Reports
show higher rates of violence against women dur-
ing the pandemic. In addition, women are also
more exposed to the health risks of COVID-19,
as globally women make up for 70 percent of
health and social care workers.10
Even though vaccination programmes have

taken off in most countries around the world, for
many developing countries these efforts are still
in their very early stages. Thus, as of August
heavily affected sectors.8 Estimates for Africa 14th, 2021, vaccinations per 100 people in the
indicate that the share of informal workers that Africa continent stood at 6.2 percent, as com-
have been impacted by COVID-19 could be as pared to 112.8 percent for countries in the Eu-
high as 83 percent, corresponding to 325 mil- ropean Union, 89.4 percent in North America,
lion people (see Table 1). For many informal 71.31 in South America or 66.45 in Asia.11 As
workers, remote working or not working is not a result, for many countries, physical distancing
an option, since staying at home translates into continues to be the only way to break the chain
lost income. Without alternative sources of in- of infection, especially for low-income countries.
come and no social protection, ILO estimates These countries are likely to continue maintain-
that these lost earnings could result in an in- ing COVID-19 restrictions in place, until vacci-
crease in relative poverty for informal workers nations have picked up and reached levels consis-
by almost 34 percent, globally, and by up to 56 tent with heard immunity, continuing to put ad-
percent in lower income countries. ditional strain on informal workers livelihoods.
The pandemic is also having a major impact on In conclusion, the available evidence indicates
unemployment and underemployment among in- that informal sector workers around the globe
formal workers. Many informal workers in food are an especially vulnerable group of the popula-
supply sectors are at risk of losing their liveli- tion to the socioeconomic impacts of COVID-19,
hoods due to lockdowns and restrictions on the workers who have already been severely affected
movement of people. Their vulnerability can in- by the pandemic and continue to suffer its eco-
crease further, as they may have to resort to nomic, social and health consequences.
negative coping strategies, such as sale of house-
hold assets, child labour or predatory loans from
informal moneylenders9 .
It is important to note that the impact of

COVID-19 is not being gender neutral. Women
have been particularly hit by the massive job 10
Azcona, G., Bhatt, A., Encarnacion, J., Plazaola-
8 Castaño, J., Seck, P., Staab, S., Turquet, L. (2020).
ILO (2020). “ILO Briefing note: Impact of lock-
down measures on the informal economy - A summary”. “From Insights to Action: Gender equality in the wake
Geneva. of COVID-19”. UN Women.
9 11
FAO (2020). Impact of COVID-19 on informal work- Data retrieved from the Our World in Data online
ers. Rome. database on 16 August 2021 (https://ourworldindata.
org/).
3. The impact of COVID-19 and the Figure 2: COVID-19 figures for Mozambique,
informal economy in Mozambique daily new cases and cumulative deaths
3.1. The Socioeconomic impact of

COVID-19 in Mozambique
Mozambique reported its first COVID-19 case
on 22nd March 2020. Since then, the country
has registered a total of 137,413 cases of con-
firmed COVID-19 infections and 1,690 official
COVID-19 deaths (see Figure 2).12 Actual fig-
ures are likely to be significantly higher, given
Mozambique’s weak healthcare system, as well
as its limited coverage beyond urban and peri-
urban parts of the country.
As elsewhere around world, the COVID-19 pan-

demic has had a very significant social and eco-
nomic impact in Mozambique. A number of
concurrent demand and supply shocks hitting
largely driven by industry-specific factors, not
Mozambique since the beginning of the pan-
by the impact of COVID-19.
demic account for this. These range from the
sharp slowdown experienced by the world econ- On the demand side, this sharp slowdown in
omy and its impact in Mozambique, which has economic activity has also reflected in Mozam-
one of the most open economies in Sub-Saharan bique’s weak external sector performance, with
Africa13 , to the adoption by government of strin- the value of exports dropping by 15 percent in
gent social distancing and lockdown measures. 2020, that of imports falling by 0.4 percent,
Altogether, these shocks have severely affected while foreign direct investment in projects other
economic activity and disrupted the lives and than mega-projects experiencing a very sharp
livelihoods of Mozambicans across the country. decline, from US$ 875.3 million in 2019 to only
US$ 212.5 million in 2020. Aggregate consump-
At the macroeconomic level, Mozambique’s
tion, on the other hand, fell by 7.2 percent in
economy experienced in 2020 its first contrac-
2020, with reductions in both private and pub-
tion in almost thirty years, largely due to the
lic consumption, of 2.1 and 19.3 percent, re-
pandemic, with GDP falling by 1.23 percent,
spectively (INE, 2021). While the most recent
according to the National Institute of Statistics
macroeconomic figures released by INE point to-
(see Figure 3).14 The hardest hit sectors were
wards a recovery of economic activity, Mozam-
tourism, services, transport, manufacturing and
bique’s short to medium-term outlook remains
construction, all of which contracted in 2020.
lukewarm, with the IMF projecting GDP to
Extractive industries, which in 2019 accounted
grow by only 2.1 percent in 2021 and by 4.7 per-
for 7 percent of GDP, also experienced a very
cent in 2022.
sharp decline of 15.1 percent, although this was
12
Data retrieved from WHO’s Coronavirus (COVID- An important driver behind the sharp contrac-
19) Dashboard on 16 August 2021 (https://covid19. tion in economic activity in Mozambique, expe-
who.int/). rienced since the beginning of the pandemic, has
13
Mozambique’s average exports-to-GDP ratio for the
been the adoption of a range of confinement and
2015-2019 period stood at 37.8 percent, significantly
higher than the average for Sub-Saharan Africa (24.3 per- social distancing measures by government since
cent) and higher than its regional peers: Tanzania (16.2 the beginning of the pandemic aimed at reducing
percent), Kenya (13.9 percent) or South Africa (30.0 per- the spread of COVID-19 in the country. Since
cent).
14 the introduction of State of Emergency legisla-
INE (2021). «Produto Interno Bruto, Óptica de Pro-
dução». Retrieved online on 18.08.2021.
Figure 3: Mozambique, GDP and GDP growth, first six rounds of the survey, covering the June
1995-2020 to November 2020 period, point towards a sig-
nificant impact on food security, with around
60 percent of households surveyed consistently
reporting having missed a meal as a result of
the crisis, and with up to 37 percent of fami-
lies having gone without eating for a whole day
during the early stages of the pandemic. The
survey also suggests a considerable impact on
(un)employment as a result of the crisis.
A nationally representative survey of 1,333

adults conducted in February 2021 as part of
the Partnership for Evidence-based Response to
COVID-19 paints a similarly bleak picture.17
Hence, 30 percent of respondents reported hav-
ing difficulty getting medicines in the three
months prior to the survey, while 23 percent in-
dicated they had skipped or delayed health care
tion at the end of March 202015 , the country has visits during the six months prior to the survey
seen measures being introduced to reduce busi- as a result of the COVID-19 crisis. On the other
ness hours, limit and, in some cases, ban certain hand, 61 percent of respondents claimed to have
type of business activities (e.g. gyms, the sale of experienced income loss because of the pandemic
alcohol, barracas, etc.), reduce mobility and lim- and 54 percent reported missing meals. Key bar-
itations on social gatherings, close down inter- riers to foods access reported in this survey in-
national air traffic, limit access to public spaces cluded reduced income, higher food prices, food
(e.g. beaches), tighten controls on international markets being closed, mobility restrictions and
freight and travel, or rotation schemes for both food supply shortages.
public and private sector workers, among others.
In addition, the disruption of international and, Against this backdrop, Barletta et al (2021)18
especially, regional supply chains, also played an estimate that the COVID-19 crisis could result
important role. This is particularly the case for in a drop in household consumption of between
supply chains with neighbouring South Africa, 7.1 and 14.4 percent in 2020, and an increase in
a country which in 2019 accounted for 28.5 per- consumption-based poverty of between 4.3 and
cent of Mozambique’s imports and on which it 9.9 percentage points, with up to 2 million addi-
relies for the supply of all type of goods, from tional people entering poverty. Accordingly, the
spare parts, to equipment, construction mate- impact of poverty is likely to have been more in-
rial, fresh food, or consumer goods. tense in rural areas, as consumption levels there
were already low prior to the pandemic.
Unsurprisingly, the impact of the COVID-19
pandemic has been heavily felt in both busi- Businesses have also been heavily hit by the pan-
nesses and families. At the household level, INE demic. A survey by the National Institute of
and the World Bank’s High Frequency Survey16 , Statistics of over 89,000 firms across the coun-
which has been regularly monitoring the situa- try undertaken between April and July 202019 ,
tion of urban households since early in the pan- 17
PERC (2021). “Finding the Balance: Public Health
demic, helps illustrate the severe distress faced and Social Measures in Mozambique”. Partnership for
by families across the country. Results for the Evidence-based Response to COVID-19.
18
Barletta et al. (2021). The impact of COVID-19
15
Presidential Decree No. 11/2020 of 30 March 2020. on consumption poverty in Mozambique. WIDER WP
16
INE and WB (2021) Relatório do Inquérito sobre 2021/94.
19
o Impacto da COVID-19 nos Agregados Familiares Ur- INE (2020). “Resultados do inquérito do Impacto da
banos em Moçambique: Junho-Novembro de 2020. Ma- COVID-19 nas empresas.” Instituto Nacional de Estatís-
puto. tica. Maputo.
Table 2: Impact of COVID-19 on business and matic shocks hitting the country in recent years,
work low commodity prices for some of Mozambique’s
main commodity exports and the hidden debt
crisis, among other factors. It also follows sev-
eral years of falling income per capita, a result of
Mozambique’s weak economic performance and
high population growth, currently at close to 3
percent per year, one of the highest in the world.
In response to these impacts, the government

of Mozambique has articulated a series of mea-
sures aimed at relieving the situation of both
families and businesses. Social policy responses
included the introduction of VAT and import
duty exemptions for medicines, medical equip-
ment, food (sugar, oil) and other basic consump-
tion goods (e.g., soap). A general 10 percent
reduction of electricity tariffs, increasing to 50
percent for families paying the so-called “social
found that 2.3 percent of these companies had tariff”, as well as the delayed payment of water
closed as a result of COVID-19, resulting in a bills and its exemption of payment for low con-
total of 43 thousand job losses (see Table 2 for sumption users. A key measure on the social
more details). Small enterprises were particu- front has been the expansion of Mozambique’s
larly affected by these closures, with 3.1 percent existing social protection programme aimed at
of them closing, accounting for 95 percent of increasing coverage from 592,179 to 1,695,004
COVID-19 related job losses. Estimates from households to include families affected by the
this same survey put the loss of business rev- COVID-19 pandemic who would not otherwise
enue for companies during the first half of 2020 qualify for support.21
at 41 percent, with this drop being particularly
Measures to support businesses have included
intense for small firms: 48.7 percent.
wavers and the postponement of payments of
The impact of the pandemic on business has corporate taxes for firms with turnovers below
been especially intense in sectors such as MZN 2.5 million, the compensation of VAT tax
tourism, heavily dependent on international vis- credits with other taxes, a 10 percent reduction
its. Thus, a study of June 2020 by the CTA, of electricity tariffs for businesses, the suspen-
Mozambique’s main business association, found sion of mobile phone money fees, a temporary
that close to 75 percent of tourist establishments increase in mobile money transaction limits or
surveyed for this study in the province of In- the forgiveness of social contribution fines, the
hambane, Mozambique’s main tourist destina- introduction of a US$ 22.9 million subsidized
tion, had closed.20 line of credit to support SMEs, as well as more
general measures by the Bank of Mozambique
It is important to note that these events take to reduce interest rates and banking regulatory
place against an already weakened economy, a requirements, contributing to raise liquidity in
frail business sector and social fabric. Hence, the the financial sector.22
pandemic follows several years of lukewarm eco- 21
Prior to COVID-19 these programmes mainly fo-
nomic performance, with the Mozambican econ- cused on supporting the very vulnerable, namely the el-
omy growing only at 3.3 percent between 2015 derly, persons with disabilities, those chronically ill, as
to 2019, down from an average of 7.9 percent well as certain categories of children (e.g. child headed
between 1993 and 2015, the result of adverse cli- households, orphans).
22
World Bank (2021) Mozambique Economic Update:
20 Setting the Stage for Recovery. Maputo.
CTA (2020) Impacto da pandemia da COVID-19 no
sector empresarial e medidas para a sua mitigação. Ma-
puto.
3.2. The informal economy in Table 3: Number of informal workers aged 7+

Mozambique years by main sector of activity (in %)
The informal economy in Mozambique is esti-

mated to account for 31 percent of GDP.23 It in-
cludes a very wide and diverse range of activities
that go from the millions of Mozambicans earn-
ing a livelihood from subsistence farming and
fishing, to the many informal enterprises and
self-employed people operating in the country’s
urban economy as petty traders, or involved in
small-scale manufacturing, construction and ar-
tisanal activities.
INE’s 2005 Survey of the Informal Sector, primary sectors, constituting a very different
INFOR-200524 , provides the most indepth ac- profile to that of informal economy workers in
count of this sector. Based on a nationally rep- urban areas. Thus, data from this survey indi-
resentative sample of 6,900 households, the sur- cates that 97 percent of rural informal workers
vey provides detailed information of both formal were employed in agriculture and other primary
and non-formal employment activities for people activities. In urban settings, however, this share
over the age of seven, generating robust indica- was significantly lower, although still consider-
tors down to the provincial level. Any activity able: 68.9 percent, and was followed by those
in which participants of the survey are engaged employed in trade and services, 24.3 percent,
as employees, self-employed or entrepreneurs is and those engaging in construction activities:
considered part of the informal economy if none 6.9 percent (see Table 3).
of the following two conditions are met:
Interestingly, the INFOR-2005 survey found
1. The activity is registered either with the that a significant share of formal workers, 17
provincial authorities or in the correspond- percent, also engaged in secondary informal ac-
ing Repartição de finanças/local tax office. tivities. This share was particularly high in rural
2. The enterprise is in possession of an official areas, where up to 36.4 percent of formal work-
document, either a license (alvará) or a reg- ers engaged in informal activities, but was also
istration record (Ficha de Registro), or, for significant in urban settings, in both cases most
the case of employees, these hold an official of these formal workers engaging also in informal
work contract. activities in agriculture.
The results of the survey indicate that 75.2 per- Informal sector workers were not particularly
cent of the workforce was employed in the in- young in the INFOR-2005 survey: 42.9 percent
formal economy in Mozambique in 2005, an- of them were 35 years of age or older. Those un-
other 7.9 percent was employed in formal activ- der the age of 25 only represented 31 percent of
ities, while the remaining 17 percent was unem- the total. The survey also found important gen-
ployed. Informality was more prevalent in rural der dimensions to informality, with 58.5 percent
areas, where of 86.9 percent of the workforce was of informal workers in the country being women.
involved in informal activities, with this same These differences were particularly noticeable in
share only reaching 50.4 percent in urban areas. the south of the country, were up to 65.8 per-
cent of informal sector workers were women. In
It is important to note that a very large share terms of employment status, the survey found
of people in the informal rural economy identi- that a large majority of informal economy oper-
fied in the INFOR-2005 survey engaged mainly ators outside agriculture, 73.2 percent, were self-
in subsistence activities in agriculture and other employed or entrepreneurs, the rest (26.8 per-
23
cent) being employed by others.
See footnote 3.
24
INE (2006). “Resultados do Primeiro Inquérito Na-
cional ao Sector Informal: INFOR-2005”. Maputo.
A study on informality in Mozambique by Byiers between 17 and 19 times less profitable than for-
(2009) using data from INE’s INFOR-2005 sur- mal companies. A significant part of these dif-
vey and focusing only on non-agricultural self- ferences can be explained by differences in firm
employed entrepreneurs, finds that 57.7 percent and ownership traits that exist between formal
of these operators could be classified as involun- and informal enterprises. But even after control-
tarily informal, with 44.1 percent of them citing ling for these differences, informal firms are still
a lack of salaried employment opportunities as between 2 to 3 times less productive than formal
the main reason for operating an informal busi- micro enterprises. Despite these differences, Aga
ness, and an additional 13.6 percent explicitly et al (2019) find that few informal firms tran-
citing survival-related reasons. According to this sition into the formal sector, with very few of
same study, the estimated effect of being for- them having even considered this move. Mostly,
mally registered on labour productivity for this because these firms see no benefit of registering
particular segment of the informal sector was of their businesses, with 33.5 percent of surveyed
between 31.6 to 34.5 percent, controlling for all firms citing this as the main reason for not do-
other possible firm characteristics, with the im- ing so, although other factors such as the time,
plied average annual value-added premium for fees and paperwork involved in formalizing (26.1
a microenterprise registering reaching between percent) or the lack of information (18.6 per-
US$513 and US$560 per year. cent) were also important.
The World Bank’s recent 2018 Mozambique En- Despite these findings regarding the structural
terprise Survey, which includes a specific mod- differences that exist between formal and in-
ule on informal enterprises, also provides useful formal enterprises, Aga et al (2019) find that
insight into the country’s informal economy.25 there is significant heterogeneity among infor-
Based on a sample of 554 informal firms operat- mal firms. In fact, some of these present firm
ing in Beira, Maputo and Nampula26 , the sur- characteristics and levels of performance that
vey finds that informal firms tend to be smaller very much resemble those of formal enterprises.
and younger than formal enterprises, employing These are firms, they argue, that could poten-
on average 1.6 employees, as opposed to 14.1 tially become formal enterprises with the ade-
employees for formal enterprises, and having an quate support (training and skills development,
average age of 4.7 years, against 8 years for mi- formalization assistance, market and technology
cro formal enterprises and 14.9 years for larger access support, etc.). The rest, representing
formal companies. Informal enterprises in these close to 60 percent of surveyed firms are, how-
locations are also more likely to be owned by ever, unlikely to be able to make this shift to-
women and mostly operate in the retail sector, wards formalization.
with an estimated 72 percent of them involved
in retail activities, against 52 percent of formal
4. The COVID-19 Informal Sector
enterprises included in the survey.
Survey (CISS)
Beyond these descriptive traits of the informal
sector, the results coming out of this 2018 enter- 4.1. Background and objectives
prise survey indicate that informal firms sell less,
From the outset, the urban informal sector in
are less productive and are less profitable than
Mozambique has been identified as one of the
micro formal firms. Hence, Aga et al (2019) find
potentially most vulnerable and at greater risk
that informality for firms in this survey is associ-
groups of the population to the socioeconomic
ated with selling 25 times less than formal firms
impacts of the COVID-19 pandemic.27 Three
and selling 15 times less than formal firms with
main reasons underlie these concerns.
fewer than 5 employees. Informal firms are also
25
See Aga et al (2019) for an analysis of main results
coming out of this survey. 27
26
Firms were considered as informal if they lacked ei- See for instance UN (2020) “Unpacking the potential
ther an operating license, a business registration certifi- socioeconomic impact of the Coronavirus pandemic in
cate, or a taxpayer’s identification number (NUIT). Mozambique” a United Nations situation analysis and
policy recommendations. Maputo, March 30th.
First, as an intrinsic part of the urban land- 4.2. Methodological considerations

scape, which in developing countries tends to
The CISS survey was designed as a quantitative
concentrate the more advanced and open parts
longitudinal study, consisting of an initial base-
of the economy (trade, industry, logistics), the
line survey and three follow-up waves, and was
informal economy is typically more exposed to
implemented between August 2020 and March
global economic dynamics, in this case, the
2021. The baseline survey was undertaken in
sharp downturn experienced by the global econ-
September 2020 and sought to generate general
omy as a result of the COVID-19 pandemic.
background information on informal sector op-
Second, informal activities in developing coun- erators and their families, in addition to collect-
tries, which often take place in streets and infor- ing a first round of data on the socioeconomic
mal market areas, have been heavily hit by con- impact of the pandemic for this group of the
finement measures and other COVID-19 related population. The remaining three follow-up sur-
restrictions introduced by governments across veys aimed at monitoring over time the socioeco-
the developing world to restrict mobility and nomic effects of COVID-19 in the informal sector
business activity, especially in urban settings. by regularly collecting data on a set of indica-
tors defined for this purpose. They were imple-
Finally, by their very nature, informal busi- mented between October 2020 and March 2021,
nesses and workers are likely to have limited ac- as per the calendar presented in Table 4.
cess to social protection and business support
programmes rolled out by governments in re- The initial baseline survey was conducted on
sponse to COVID-19. As businesses, because a randomly generated sample of 600 informal
they typically lack registration and other ba- workers, 18 years of age or older, affiliated with
sic business capacities (e.g., bank accounts) to ASSOTSI, for whom ASSOTSI had basic iden-
benefit from government loans and other sup- tity information (age, gender, etc.), as well as
port measures. As informal workers and busi- contact details. This sample was stratified by
nesspeople, because they are unlikely to meet age, in order to better represent workers in the
eligibility criteria established under most social 2017 National Census, group which was taken
protection programmes, which typically target as a reference of the relevant sample population
the most vulnerable segments of the population. of informal workers in Maputo. A subsample of
500 of these informal operators was randomly
It is against this background, that the UNDP assigned every two months for each of the three
Country Office in Mozambique decided in mid- follow-up waves conducted for this study.
2020 to undertake, in collaboration with the
Ministry of Economy and Finance (MEF), a se- The survey questionnaires comprised both stan-
ries of surveys to monitor and assess how the dard household and enterprise survey questions,
COVID-19 pandemic was impacting, from a so- with the idea of capturing both these two dimen-
cioeconomic perspective, informal sector opera- sions in our analysis of the socioeconomic impact
tors in the city of Maputo: The COVID-19 In- of the COVID-19 pandemic on the informal sec-
formal Sector Survey, CISS. tor in Maputo. An effort was made to include or
adapt questions to the COVID-19 context, based
The CISS study has been done in partnership
with ASSOTSI-AEIMO, Mozambique’s main as- Table 4: CISS Survey timelines
sociation of informal economy operators, and
has involved a variety of independent and UN
experts in its design, implementation and anal-
ysis. Within the UN, UN-WIDER’s Inclusive
Growth Programme in Mozambique provided
specific support in the design of the survey and
the analysis of results. The surveys were admin-
istered by the local office of IPSOS, the French
market research and opinion polling firm.
on similar pandemic-related work done in other obviously does not include in its membership all
countries, as well as on work in emergencies set- informal sector operators in the city of Maputo,
tings, adapting these to the context of informal least so of Mozambique. The history of AS-
sector workers in Maputo. SOTSI and, in particular, of its creation, closely
tied to the privatization process Mozambique
The CISS questionnaires were designed around underwent in the 1990s, which saw the layoff of
five broadly defined areas deemed relevant for thousands of SOE employees, most likely builds
the purpose of the study: certain biases into the survey, with an overrep-
resentation of older and female informal work-
(1) Individual, business and household charac- ers and businessowners operating in particular
teristics – this module was only adminis- activities of the informal economy in Maputo.
tered in the initial baseline survey. Therefore, the conclusions of this study need to
(2) Evolving general business conditions in the be interpreted with caution and not necessarily
informal economy. extrapolated to the population of informal ur-
(3) Impacts on employment, revenue, earnings, ban workers in Maputo and Mozambique as a
savings and intra-household nutrition. whole.
(4) Business and household coping strategies.
(5) Impact of support measures.
5. Characteristics of the sample – a
Interviews for the CISS surveys were conducted highly vulnerable segment of the
in Portuguese and Changana, depending on the population
preference of participants, and took an average
of 40 minutes, in the case of the baseline survey,
5.1. Individual characteristics
and of 20 minutes, for the three follow-up waves.
As noted in the previous section, a total of 600
In order to comply with COVID-19 social dis- informal sector workers took part in the CISS
tancing measures, interviews were conducted by surveys. This is a largely female, middle-aged
means of telephone calls, which limited the scope group of people, with basic levels of education.
of topics that could be inquired and the preci- Hence, 64 percent of them are women, a share
sion of some of the more complex information significantly higher than for self-employed work-
gathered. To address these concerns and min- ers residing in Maputo reported in the 2017
imize any form of bias or inaccurate collection National Census – taken here as the reference
of data, quality checks were conducted on every sample population – 48.8 percent of which were
round of data collection. These involved having women.
16 percent of interviews monitored by a supervi-
sor, reviewing 24 percent of interview recordings The average age of informal sector workers par-
for the more complex questions included in the ticipating in the survey is of 39.8 years, with the
survey, and reviewing an additional 5 percent of median and modal ages both at 38. Women are
interview recordings in their entirety. On the somewhat older than men: 42.3 years of age, on
other hand, to deal with issues of attrition typ- average, as opposed to 35.3 years for men. The
ical of longitudinal studies, interview substitu- youngest person was 19 years old at the time
tions of the original 500-person sample gener- of the survey. The oldest, was 70. Compara-
ated for each round were only made after 3 days tively, the CISS sample is slightly older than the
attempting to contact the selected interviewee, relevant group of the population in the 2017 Na-
with at least two calls made per day. tional Census, for which the average age is 39.5.
It is important to note that the design of the sur- Most informal workers in the CISS survey live
vey carries a series of limitations that need to be in the greater Maputo metropolitan area, either
taken into account in interpreting results. Chief in Maputo city (42.7 percent of the sample), in
among these is the fact that the survey draws the neighbouring city of Matola (43.9 percent)
exclusively from a sample of Maputo-based reg- or in Marracuene district (2.4 percent), north
istered members of ASSOTSI, association that of Maputo. There is, however, a sizeable 9.2
Table 5: Profile of participants 5.2. Business and employment

characteristics
From a business and employment perspective,
the largest group of informal workers participat-
ing in the CISS survey is that of “Fixed mer-
chants”, which constitute half of the sample (see
Table 6). These are people operating in the
various markets that exist in Maputo, and who
typically own a fixed stall from which they sell
their produce. These people trade in all type of
goods and services, although there is a prepon-
derance of those selling fresh and dry food (to-
gether, both these groups account for 48.4 per-
cent of the sample), clothes (34 percent) and ac-
cessories (13.2 percent). This group is followed
by that of street vendors (22.3 percent), and by
percent of informal workers who live in Gaza
informal workers providing plumbing, carpentry
Province, around 2-3 hours’ drive north of Ma-
or shoe repair services (4.8 percent). In this
puto and who travel daily or weekly to Maputo
regard, most survey participants are traders or
to sell their produce in the capital. Interest-
service providers, with hardly any engaging in
ingly, this group only includes women, as there
manufacturing or artisanal activities.
are no men in the CISS sample living and trav-
eling from Gaza. Compared to men, women tend to be largely
fixed or street merchants, with 60 percent of
In terms of educational attainment, up to 58.8
street vendors and 69 percent of fixed merchants
percent of informal workers in the sample have
being female. Men, on the other hand, are com-
completed primary school. An additional 34.3
paratively dominant in services, with more than
percent have attended, without completing, pri-
three quarters (77.4 percent) of all workers ac-
mary school. Out of 19.1 percent that had at-
tive in services being male. There are also dif-
tended middle-school, 12.2 percent have com-
ferences in activity between people with different
pleted middle-school and 1.3 percent that have
educational background. Service providers show
higher education (see Table 5 for details).
a higher number of people with secondary edu-
These figures are broadly in line with those
cation compared to fixed and street merchants,
found in the 2017 National Census for self-
most of whom only have primary schooling.
employed workers residing in Maputo, which can
be considered as the reference sample population Most of informal economy workers participat-
of the CISS study. ing in the CISS survey have been engaged in
these activities for many years, with 40 percent
Older workers show lower educational attain-
of them doing so for more than 10 years, 22.7
ment. Thus, while the share of those with no
percent for between five and ten years and only
formal education stand at 20.3 percent for work-
37.3 percent being engaged in informal economy
ers aged 55 and above, it is only 11.7 percent for
activities for five years or less. In addition, most
workers aged 46-54 years of age. Moreover, all
of these informal operators work on their own,
workers aged 18 to 31, have at least attended
although a sizeable 12.6 percent employ other
primary school. There are also substantial gen-
informal workers in their businesses. This per-
der differences in education: while the share of
centage is close to the share of adults residing in
women that attended primary school without
Maputo in the 2017 National Census reporting
completing it is 42.8 percent, for men this share
to be own-account workers with employees: 14.9
drops to 19.1 percent. Additionally, 10.2 percent
percent.
of women never had school education, compared
to only 0.9 percent of men.
Table 6: Type of activity of participants There are also considerable gender differences in
business and occupational characteristics in the
sample, especially with regards to the type of
activities that male and female informal workers
engage in. Women are more likely to operate fix
market stalls, selling fresh and dry food prod-
ucts, and also to be unemployed. Men, on the
other hand, are more active as street hawkers,
selling clothes and accessories, and in the pro-
vision of services, such as plumbing, carpentry,
shoe repair, etc. Many of them also reported to
be unemployed, but less so than women.
5.3. Characteristics of households in the

sample
In examining the socioeconomic impacts of the
COVID-19 pandemic on the informal sector in
Maputo, it is important to look not only at
In terms of location, up to 55 percent of infor- the employment or business effects of the pan-
mal workers in the CISS survey operate in the demic on this segment of the population, but
Zimpeto-KaMubukwana neighbourhood, in the also how these might affect their families. Data
northern outskirts of Maputo, where the Zim- collected for this study suggests that this was an
peto wholesale market is found, and which also already highly vulnerable segment of the pop-
has a high concentration of market retail activ- ulation, therefore, susceptible to being heavily
ities. This is followed by informal workers op- impacted by the COVID-19 crisis.
erating in downtown Maputo (KaMpfumo), also
known as A Baixa (9.2 percent), and by those Overall, 65 percent of survey participants are
based in Xipamanine market – Nlhamankulu heads of household. This percentage is consid-
(8.6 percent). erably lower for female informal workers, 54.5
percent, but still significant. On average, CISS
When asked where they operated, many infor- informal workers’ households have six (5.9) peo-
mal workers, totaling 16 percent of the sample ple living in them, of which three (2.8) are chil-
at the time of the baseline survey, reported hav- dren under the age of 18, many of which be-
ing been recently relocated out of the areas in ing significantly younger.28 These are large and
which they worked by local authorities. As dis- young households, both by Maputo and Mozam-
cussed in greater depth in the next section, this bican standards. Hence, figures from the 2017
phenomenon is closely linked to the high levels National Census put average household size for
of unemployment reported by participants in the Maputo at 4.75 people, and at 4.5 people for
CISS study. It is important to note, in this re- Mozambique as a whole, with 1.9 and 2.5 chil-
gard, that there is a significant share of informal dren under the age of 18 living on average in
workers in the sample whom at the time of the these households, respectively.
baseline survey reported to be unemployed: 20.5
percent, if both those actively looking for work A large majority of informal operators partic-
and those who are not are taken into consid- ipating in the CISS survey, 77.5 percent, own
eration. This was a recurrent finding through- the dwellings in which they live, mostly in and
out the various waves of the CISS survey, with around Maputo city, in the greater Maputo
the share of unemployed in the sample actually metropolitan area. Of those renting, the ma-
increasing over time. This issue is explored in jority are men (64.8 percent), and are younger
greater depth in the next section, given its direct than homeowners in the sample.
links to the socioeconomic impact of COVID-19 28
Hence, 57 percent of households have at least one
on the Maputo informal sector. child in the 0-5 years age group.
Table 7: CISS household characteristics Figure 4: Asset ownership in CISS households
In general, living conditions for families in the

CISS study are worse than those of the 2017
National Census reference sample population of
own account workers residing in Maputo. In
terms of sanitation conditions, only 21 percent
percent a gas or electric stove, and 52 percent
of informal workers in the sample had a toi-
hold a bank account. For women in the sample
let in their homes, as opposed to 52.5 percent
these percentages a significantly lower: 41, 46
for own account workers residing in Maputo in
and 29 percent respectively.
the 2017 National census (see Table 7). The
same goes for access to drinking water, with only Overall, these figures on household wellbeing
48.8 percent of informal workers participating point at high levels of vulnerability of infor-
in the CISS study having piped water in their mal workers taking part in the CISS study.
homes, as opposed to 83.6 percent for the refer- The multidimensional poverty figures reported
ence group of the 2017 National Census. in Figure 5, are further evidence of this. These
were obtained broadly following the methodol-
Informal workers participating in this survey
ogy used by the Ministry of Economy and Fi-
also own on average fewer assets than the 2017
nance (MEF) for its 4th National Poverty As-
Census reference sample population. Hence,
sessment of 201629 , which follows the Alkire-
fewer of them own phones, computers, electric
Foster deprivation approach to measuring mul-
and gas stoves, or cars, and also a fewer number
tidimensional wellbeing and poverty. Typically,
of them hold bank accounts (see Figure 4).
this approach yields multidimensional poverty
Of particular concern in the context of the sharp estimates strongly correlated with income or
economic downturn caused by the pandemic, consumption-based poverty measures.
which has reduced economic activity and earn-
To this end, deprivation estimates were obtained
ings throughout the Mozambican economy, are
for five wellbeing dimensions: (1) education at-
the limited sources of income that exist in house-
tainment; (2) access to drinking water; (3) ac-
holds of informal workers participating in the
cess to improved sanitation; (4) food intake; and
CISS study. Up to 5.3 percent of these house-
(5) ownership of durable goods. A household
holds had no one earning an income in the CISS
was then considered poor if it suffered from de-
baseline survey, and an additional 57 percent
privation in at least four of these five areas.30
had only one income earner in the household.
Based on these parameters, our estimates in-
These shares are similarly high regardless of
29
age, gender or household size and have remained MEF (2016) Pobreza e bem-estar em Moçambique:
at similar levels throughout all three follow-up Quarta avaliação nacional (IOF 2014/15). Maputo.
30
This is a more stringent poverty measure than
waves of the CISS survey. There are also impor- MEF’s, which considers someone to be poor if they are
tant gender differences in asset ownership in the deprived in 60 percent of the dimensions they take into
CISS sample: 67 percent of men own a radio, 56 consideration.
Figure 5: Deprivations and multidimensional 6. The socioeconomic impact of

poverty in CISS household COVID-19 on the informal econ-
omy in Maputo
Having drawn the characteristics of informal

workers in the urban informal economy in Ma-
puto and that of their families and households,
this section turns to the question of how they
have been impacted by the pandemic – and how
they have coped and responded to the crisis. It
investigates these issues looking at different di-
mensions and channels of how the COVID-19
crisis has affected this segment of the popula-
tion and their families, recognising that these
impacts are often multifaceted in nature. Tak-
ing this multifaceted approach, by which sev-
eral measures of impact and coping strategies
are examined, also contributes to the robustness
of the findings coming out of this analysis, in
that the overall findings presented in this sec-
dicate that 13.1 percent of the sample suffered tion are backed by more than one data point.
from multidimensional poverty at the time of the
CISS baseline survey, in September 2020, and up Overall, the analysis presented in this section
to 37.3 percent suffered from at least three depri- shows that the COVID-19 pandemic has had a
vations. As a comparison, according to MEF, in very severe impact on informal workers’ busi-
2014/15 only 1 percent of the population of the nesses, employment conditions and on their
city of Maputo suffered from multidimensional lives, as well as that of their families, especially
poverty. during the initial stages of the crisis. While the
situation faced by these people has stabilized
Multidimensional poverty for informal workers over time, informal workers continue to face a
in the CISS survey is predominantly female: 15.8 very challenging outlook. This is particularly
percent of women are multidimensionally poor, worrying in a context in which this group of the
as opposed to only 8.4 percent of men. This is population was already experiencing important
mostly driven by gender differences in both edu- vulnerabilities, as outlined in the previous sec-
cation and ownership of durable goods. Most of tion, and in which they have hardly received any
the poor are between 39-45 years old (27.8 perform of support throughout the pandemic, at
cent), followed by the younger age group of 32 to least up to the time of the last wave of the CISS
38 years (24.1 percent). Yet, when we only com- study, in March 2021. Despite the very chal-
pare the shares of poverty among different age lenging environment and outlook faced by in-
groups, the oldest ones aged 55 and above show formal workers in Maputo, they overall remain
the highest share of multidimensionally poverty optimistic and have done so throughout the var-
in their group (17.3 percent). Most people con- ious waves of the CISS study. However, this op-
sidered as multidimensionally poor have only at- timism is slowly waning.
tended primary school (96.2 percent), including
31.7 percent who have finished the primary ed- 6.1. Impacts on informal economy
ucation. businesses and informal employment
Informal workers faced a sharp deteriora-
tion of business and market conditions
The COVID-19 pandemic has coincided with a
very sharp deterioration of the business envi-
ronment faced by informal workers in Maputo.
This sharp decline has been felt from the out- Table 8: Changes in market conditions
set and has continued to affect informal econ-
omy operators in Maputo throughout the crisis,
as captured in successive waves of the CISS sur-
vey. Mostly, this deterioration can be attributed
to the sharp decline in economic activity regis-
tered in Mozambique since the beginning of the
pandemic, as well as to the restrictions imposed
by the government as part of the COVID-19 re-
sponse. Yet, other factors not related to the
pandemic have also played a role.
The September 2020 baseline survey for the

CISS study already captures this rapidly deteri-
orating business environment faced by informal
workers in Maputo. Back then, up to 83.6 per-
cent of participants in the CISS baseline survey
reported having seen their sales drop since the
beginning of the pandemic. A similar 83.4 per-
cent reported seeing the number of clients fall,
while 60 percent claimed experiencing problems
in purchasing goods or finding these more expen-
sive. This latter trend probably owes to the dis- Beyond these deteriorating market conditions,
ruption faced by global supply chains during the informal workers in the CISS study have also
initial stages of the pandemic, disruption which had to contend with restrictions on informal sec-
also affected distribution channels in Mozam- tor activities put into place by local authorities
bique. It has also been a recurrent finding, with and the police in Maputo. Thus, only in the
informal workers taking part in the CISS study initial baseline survey of September 2020, up to
systematically putting the cost of goods as their 46 percent of informal workers reported having
top concern for all four waves of the CISS sur- had to relocate their businesses to another part
vey. Overall, no discernible differences can be of town. Subsequent waves of the CISS study
appreciated between male and female informal have continued to capture this phenomenon.
operators taking part in the CISS study in any
of these areas. These relocations partly owe to the COVID-19
restrictions on businesses introduced by the gov-
Data collected in subsequent waves of the CISS ernment as part of its social distancing and con-
study indicate a certain stabilisation of market finement measures, which has limited when and
conditions faced by informal operators in Ma- where informal workers can operate. However, it
puto (see Table 8). Hence, around a third of also coincides with the Municipality of Maputo’s
informal workers taking part in this follow-up decision to clear of street vendors the central dis-
surveys – 32 percent in the first wave, 33 per- trict and commercial hub of the city, “A Baixa”,
cent in the second and 38 percent in the third as part of its urban regeneration efforts for this
– reported experiencing no significant changes part of town, an initiative that predates the pan-
during the previous month in the market condi- demic. In addition to these cases of relocation,
tions they faced. In fact, a few saw market con- many informal workers participating in the CISS
ditions improve, reporting having more clients or study, up to 28.1 percent of them, have had their
finding goods and equipment cheaper or easier merchandise seized by the police, leaving them
to find. Yet, mostly, business conditions faced with no goods to sell and, therefore, resulting in
by informal workers in the CISS study have not significant losses.
improved, and for many they have continued to
deteriorate, with falling sales and goods becom-
ing more expensive and harder to find.
Informal operators saw the hours they Figure 6: Impact on hours worked
worked significantly drop because of the
pandemic
This deteriorating business environment faced
by informal workers in Maputo since the begin-
ning of the pandemic has had a direct impact
on their activity, as captured in various perfor-
mance indicators monitored for the CISS study.
This impact has been felt from the outset, with
no significant improvement experienced since.
This is the case, for instance, of the number of

hours worked by informal workers in the CISS
study (see Figure 6). At the time of the base-
line survey in September 2020, a vast major-
ity of these (72.3 percent) reported to be work-
ing fewer hours since the COVID-19 crisis be-
gan, while only 16.6 percent indicated they were
working the same number of hours than before
to the pandemic, suggesting business and work baseline survey, not in relation to pre-pandemic
conditions for this sub-group might have not times. In any case, these trends seemed to
changed significantly since the pandemic. In- have reversed by the time of the third follow-
formal operators with a fixed market stall were up wave, in March 2021, coinciding with the
those most affected in terms of reduced work- tightening of social distancing conditions in Jan-
ing hours: 62.3 percent of those reporting fewer uary 2021, following the sharp rise in COVID-
working hours were fixed merchants, followed by 19 infections and deaths experienced in early
30.1 percent street vendors. This finding is not 2021, with the share of informal workers report-
surprising, given that as part of its COVID-19 ing working fewer hours increasing up to 52.6
response, the government of Mozambique has percent in March 2021.
significantly reduced market opening hours.
. . . They are also earning less during the
As with other indicators, the impact of the pan- pandemic
demic on working hours in the sample stabilized A similar pattern to the one observed for hours
during the last few months of 2020. Hence, the worked by informal sector workers emerges for
share of informal workers reporting to work the reported earnings. Reported earnings need to be
same number of hours as the previous month taken with caution in survey work, such as the
increased to 40.6 percent in October 2020, and CISS study. Hence, survey participants are often
then again to 49.4 percent in December 2020. reluctant to share this type of information with
On the other hand, the proportion of those re- strangers. They also have different understand-
porting working fewer hours, fell to in 33.8 per- ings of what constitutes “earnings”, while recall
cent in October and to 25.8 percent in Decem- data can be problematic and lack accuracy. The
ber 2020, while the share of those indicating latter is especially the case for long recall peri-
longer working hours increased to 23.4 percent ods and for survey participants, such as infor-
in October and then experience a small decrease mal workers and informal businessowners, who
down to 19.9 percent in the CISS wave of De- are unlikely to keep accounts or business records
cember 2020. It is important to note, though, and who are participating in a survey over a tele-
that these variations take the previous month of phone call. Yet, assuming that respondents are
each of these two waves – October and Decem- consistent over time in their attitudes, in their
ber – as a reference, not the pre-COVID-19 sit- understanding of concepts such as earnings and
uation. Therefore, if they are to be interpreted in their responses to survey questions, this data
as an improvement, it is only with regards to can provide useful trend information.
the situation in September, at the time of the
Figure 7: Earnings in the CISS sample (MZM) participate. Xitiques are a form of rotating sav-
ings and credit associations (ROSCA) very pop-
ular among informal workers in Maputo. Partic-
ipants in such schemes, typically informal work-
ers operating in the same market or street, or in
the same line of business, regularly contribute
funds to a given Xitique, for instance, on a on a
daily, weekly or monthly basis. They are then al-
lowed to draw the full amount pooled by all par-
ticipants into the Xitique every so often. Typ-
ically, the contributions are in a fixed amount.
This makes these contributions a particularly ro-
bust indicator, in this case of savings or business
surplus, as survey participants are likely to re-
call in greater accuracy the money they have put
into Xitiques.
As with hours worked, earnings for workers in
the CISS sample experienced a very sharp drop Prior to the pandemic, close to 70 percent (69.6)
between March 2020, prior to the pandemic, and of informal workers in the CISS study regularly
the baseline survey, in September 2020. For the participated in Xitiques. This share was simi-
sample as a whole, earnings fell by 63.1 percent, larly high across all groups, including those who
from an average of 15,108 Meticais per month in were unemployed at the time of our baseline sur-
March 2020, to only 5,568 Meticais in September vey. The only outlier were informal providers of
2020. This drop was particularly felt among par- services, for whom this share was even higher:
ticipants in the survey who are multidimension- 79.3 percent. Many workers participated in
ally poor, who saw their earnings drop by 71.7 more than one Xitique before the pandemic: 20
percent during this same period, as well as by percent did so in two of Xitiques, 6.8 percent in
women participating in the CISS study, whose three and 1 percent in up to four Xitiques.
earnings fell by 66.2 percent, a drop higher than
the sample average. Before the pandemic, informal workers in the
CISS study contributed on average 5,625 Met-
As depicted in Figure 7, from the baseline sur- icais per month to Xitiques. This amount was
vey to the second wave, in October 2020, aver- significantly higher for men (7,606 MZM) than
age earnings continued to drop for the sample for women (4,472 MZM) and than for informal
as a whole, although more moderately and with workers who are multidimensionally poor (3,140
earnings of the multidimensionally poor expe- MZM). Workers would use the money that they
riencing a slight increase. Average sample earn- regularly draw from Xitiques for a variety of pur-
ings then stabilized in the following round of the poses. Some used these funds to buy equipment
CISS survey – although they fell for the group of or merchandise for their business. Others used
multidimensional poor – and then subsequently them to purchase household durable goods, such
fell again for the final wave of the CISS study, as fridges or stoves. Xitiques were also used for
in March 2021. Again, this coincides with the savings, or for family and religious ceremonies.
tightening of COVID-19 restrictions taking place
in early 2021, following the spike in COVID-19 The pandemic has had a dramatic impact on
infections and deaths experience at the time. the use of Xitiques by informal workers, severely
hampering their ability to operate and invest in
Measures of business surplus and savings their businesses and acquire basic household as-
also show a sharp decline sets. To start, a very large number of infor-
In order to get a better sense of how the pan- mal workers have stopped participating in Xi-
demic was affecting business surplus and sav- tiques altogether. Hence, by September 2020
ings, data was collected for the CISS survey on only 30.5 percent of survey participants were
Xitiques in which informal workers in Maputo contributing to Xitiques, down from 69.6 per-
Figure 8: Contributions to Xitiques (MZM) Informal businesses and workers have

adapted to the crisis in a variety of ways
Informal businesses and workers are responding
and adapting to this challenging and deteriorat-
ing business in a variety of ways (see Table 9).
Within their businesses, many have decided to
reduce purchases of merchandise to respond to
the drop in the number of clients and falling
demand. This strategy has consistently been
adopted throughout the pandemic by around 40
percent of informal workers in the CISS study,
as captured in the various waves of the survey.
Other workers have opted to sell different type of
products, and in some cases they even increased
purchases of merchandise – although the ratio-
nale behind this latter strategy was not made
cent before the crisis, with this share remaining explicit by these informal operators.
at these levels throughout subsequent waves of
the CISS survey. This observation is very similar Many workers in the Maputo informal sector
for both men and women, but not so for the mul- have gone a step further and have switched busi-
tidimensionally poor who have seen the number ness, altogether, for instance going from having
of participants in Xitiques plumet from a pre- a fixed stall in the local market to selling goods
crisis share of 62 percent, to only 17.5 percent on the streets of Maputo. This has been a re-
in September 2020, with this proportion further current feature throughout the crisis, with the
decreasing in subsequent waves of the survey. share of workers in the CISS sample reporting
to have switched activity during the previous
Moreover, the value of Xitique contributions has month hovering around 10-20 percent: 19 per-
also fallen sharply, by an average of 71 percent in October 2020, 12.8 percent in December
cent between March and September 2020, with 2020 and 15.8 percent in March 2021. One of
this drop being particularly intense for the mul- the groups for which this phenomenon has been
tidimensionally poor (89.3 percent drop) and more prominent is that of informal workers with
for women in the sample (76.8 percent). Since fixed market stalls, who have seen their numbers
then, the situation has stabilised, although with drop from 244 in the September 2020 baseline
these values remaining well below pre-COVID-
19 levels. Thus, between September and Octo- Table 9: Changes made to business
ber 2020, average Xitique contributions contin-
ued to fall, albeit at a slower pace and with aver-
age Xitique contributions actually increasing for
both women and those informal workers who are
multidimensionally poor. There was also a slight
recovery in Xitique contributions at the end of
the year, in December 2020, yet these contri-
butions dropped again between December 2020
and March 2021, with this being the case for all
groups under consideration: men, women and
the multidimensionally poor, following a similar
pattern to that observed for the case of earnings
and hours worked in the sample (see Figure 8).
survey to only 203 by the time the third wave of Figure 9: Unemployment in the CISS sample
the CISS study of March 2021, equivalent to a
16.8 percent drop.
When asked about the reasons for switching ac-

tivity, most informal workers reported business
or livelihood reasons, such as need to earn more
money to make ends meet (25.3 percent in the
March 2021 CISS wave), the drop in sales (19
percent), having fewer clients (5.1 percent), or
the higher costs of the merchandise they sell
(11.4 percent), among other reasons. There is,
however, a small, yet not insignificant number
of informal workers changing business that had
other motives. This includes the fact they were
forced to relocate from their usual workspace
(3.8 percent), that they had had their merchan-
dise seized by the police or the local authorities Detailed data on unemployment collected for
(3.8 percent) or lack of safety (2.5 percent). the baseline survey indicates that 70 percent of
these people were informal workers operating a
Many informal sector workers have fixed market stall before becoming unemployed,
stopped working altogether while 24 percent were street vendors. Unemploy-
A very significant share of those “switching” ac- ment particularly affects informal workers sell-
tivities, have actually become unemployed or ing second-hand clothes and shoes (46 percent
have stopped working and looking for work, alto- of the total), as well as informal workers who
gether. This is perhaps one of the more striking used to operate in the central neighborhood of
findings coming out of the CISS surveys, con- “A Baixa”. Many of these are workers who were
sidering that these workers have no access to relocated out of this area by the local authori-
unemployment insurance and that many live in ties, as part of the municipality’s urban regen-
already highly vulnerable family settings. Un- eration efforts for this part of Maputo. These
employment for participants in the CISS study workers were meant to move to the new mar-
has hovered around the 20-30 percent range ket area of Laulane, in the northern outskirts of
throughout the various CISS surveys. During the city. However, at the time of these reloca-
this time, unemployment in the CISS sample tions, the Laulane market did not have the min-
has gone from 20.5 percent in September 2020 to imum basic infrastructure conditions (transport
26.7 percent in the October 2020 wave, increas- access, washrooms, warehousing space) required
ing again to 28.8 percent in December 2020, and by the municipality for this market to be able to
then dropping slightly to 25.8 percent in March open to the public, leaving these informal work-
2021 (see Figure 9). ers with no place to operate.
Unemployment in the informal economy in Ma- A “deep dive” undertaken for the second follow-
puto is a recent phenomenon. Hence, the bulk up wave of the CISS survey, in October 2020 to
of the unemployed in the CISS study, up to understand better the conditions of those unem-
80 percent, reports having lost their job in the ployed revealed that, while not able to work in
months immediately following the outset of the their usual line of business, some of these work-
pandemic. Very few of them were unemployed ers were actually engaged in other activities that
prior to that. It is also largely a female phe- could be qualified as being working. These work-
nomenon, with women accounting for 80 percent ers represented around 35 percent of those claim-
of unemployed informal workers in third CISS ing to be unemployed. While some of them were
wave of March 2021, a percentage significantly doing odd jobs, or biscates, to earn some money,
higher than the proportion of women in the CISS the majority of these informal workers had de-
sample: 64 percent. cided to stay at home, spending their time fixing
their houses, tending family plots (machambas), wave. This is a sign of the magnitude of the
or taking care of family members, typically chil- impact that the COVID-19 crisis had on the in-
dren not able to go to class due to the closure formal economy during its initial months of the
of schools decreed as part of the government’s pandemic. Yet, these numbers have increased
COVID-19 restrictions. over time: 99 informal workers reported growing
earnings in the October wave, 94 in the Decem-
When asked why they were not able to go back ber wave and 106 in the final follow-up survey
to do their usual work in the Maputo informal conducted in March 2021 (see Table 10). It is
economy, most of these informal workers claimed important to note that not all workers have done
not having enough money to buy products to well throughout the pandemic. In fact, only 18
sell, with 71.6 percent of unemployed workers reported growing earnings over all four waves of
citing this as a major factor. Similarly, an ad- the CISS study.
ditional 10 percent mentioned the high cost of
buying products to sell. Many, however, around Men in the sample seem to have done better by
30 percent, cited the fact that they had been this metric, although the share of women and
forced to relocate, with no alternative location poor informal workers reporting positive growth
offered in which to operate, or that they had had in earnings throughout the surveys was still sig-
their goods seized by the local authorities leav- nificant. Both average and median growth in
ing them with nothing to sell, reason cited by earnings for this group of workers was also sig-
11.9 percent of these workers. nificant; i.e. the few who did well in terms of
growth in earnings from one wave to the next,
Some informal workers have actually did significantly well.
done quite well during the pandemic
Despite the very challenging socioeconomic en- Informal workers positing growing earnings be-
vironment faced by the informal economy in tween waves tend to be somewhat younger, bet-
Mozambique during the COVID-19 crisis and ter educated, less deprived and to live in house-
the negative impact it has overall had on infor- holds with more income earners than the rest.
mal workers livelihoods, some of them have ac- But these differences are not major, and, in
tually fared relatively well during the pandemic. many instances cases, they are not statistically
significant.
The number of informal workers in the sam-
ple reporting positive growth in earnings was Despite the very challenging outlook, in-
very small for the September baseline survey, formal workers overall remain optimistic
only 26, representing 5.8 percent of workers for Despite the difficulties that informal workers
which data on earning was collected during that in Maputo have faced since the beginning of
the COVID-19, they have remained overall op-
Table 10: Informal workers that did better timistic about their situation. This has been
the case even recognizing, wave after wave, that
these positive expectations have consistently not
been met. However, over time, their optimism
has dropped.
Hence, in September 2020, 62.3 percent of in-

formal workers considered things would be bet-
ter or much better during the following month,
despite having experienced a drop in earning
of up to 63.1 percent since the beginning of
the COVID-19 crisis and with the country en-
tering some of the most challenging moments
of the pandemic. A month later, in October
2020, the share of informal workers that felt in
the same positive way had actually increased
Figure 10: Expectations vs. assessments ing to cut back on their food intake and also on
the number of daily meals they take. While the
share of those families going through this situa-
tion has dropped over time, it still remains high.
As displayed in Table 11, at the time of the

baseline survey, in September 2020, almost three
quarters (72.5 percent) of participants in the
CISS study reported having eaten less than
usual during the past month due to lack of
money. Of these, almost four out of five (78.6
percent) indicated they had at least missed
a meal during this same period. These two
shares have been dropping since then, yet for
the March 2021 wave of the CISS survey those
eating less than they should, due to lack of
money still stood at a significant 65.4 percent.
Of these, 60.9 percent had missed at least a
slightly to 64 percent, even though 54.7 per- meal during the past month. Nutritional stress
cent of them thought September 2020 had gone is particularly high among informal workers in
worse than expected, and an additional 21 per- Maputo deemed multidimensionally poor, with
cent felt things had not changed during the past more than 80 percent of these reporting having
month. The share of those remaining optimistic to cut on food and missing meals for all four
in the December 2020 continued to remain high, waves of the CISS survey.
although at a significantly lower proportion of
To obtain more insights on the families’ nu-
53.4 percent. The March 2021 CISS wave seems
tritional dynamics, additional information was
to confirm this slow but downward trend in lev-
solicited on this topic in the March 2021
els of optimism, with the proportion of informal
CISS wave. The information collected suggests
workers considering things would be better or
that families are facing considerable nutritional
much better during the following month drop-
stress. Hence, of those respondents claiming to
ping to a low, but still significant, 48 percent
have missed at least a meal in February 2021,
(see Figure 10).
14.6 percent had done sone more than 20 times
during that month, 16.1 percent eight to 10
6.2. Impacts on informal workers’ times and 38.7 percent, 4 to seven times. These
families problems were not only affecting informal work-
The impact of the COVID-19 socioeconomic cri- ers or other adults in the family: 92.7 percent of
sis on informal workers in Maputo has undoubt- those informal workers reporting having to eat
edly also been felt in their homes and families. less in February indicated that cuts in food were
In some areas, this has led to worrying outcomes also affecting children in the household. More-
and dynamics. In others, families have so far over, families in the informal sector in Maputo
managed to withstand the worst effects of the
COVID-19 socioeconomic crisis. This section Table 11: Nutrition trends in the CISS sample
presents key findings in this area.
Many families are finding it challenging

to meet basic food and nutrition needs
Perhaps one of the most concerning findings
coming out of the CISS study is the impact that
the COVID-19 crisis has had on informal work-
ers’ ability to bring enough food home, with
many informal workers and their families hav-
were not only having to cut on food. Up to 16.8 Table 12: Share of participants that tried to
percent also reported not having sufficient water sell assets
during the month of February to meet house-
holds’ basic water needs.
It is important to note that nutritional stress

in the CISS sample is something relatively re-
cent and very much coincides with the COVID-
19 pandemic. Thus, the number of meals missed
by informal workers in the CISS study stood at
an average of 1.4 meals in March 2020, just be-
fore the COVID-19 pandemic hit Mozambique,
as reported by survey participants, compared to puto participating in the CISS study that live
an average of 8.4 meals missed in March 2021. In in rented dwellings, a sizeable 43.5 percent had
fact, almost 80 percent of those missing a meal in failed to pay the rent in October 2020, the first
the month prior to the March 2021 CISS wave, time this question was raised in the CISS sur-
had not experience any sort of nutritional stress veys. This percentage has since increased, reach-
a year earlier. ing 53.7 percent in March 2021.
Households are being forced to sell home Households have resorted to savings and
assets and miss key payments credits to mitigate the impact of the crisis
In addition to cutting back on meals, informal With business slowing down and both sales and
workers and their families have sold household earnings dropping, many informal workers have
assets to generate cash. The sale of assets has had to borrow money or use their savings to
been a recurrent feature in the CISS surveys, make ends meet at home and in their businesses.
reaching a high in September 2021 when 21.3 In this regard, figures for participants in the var-
percent of the CISS sample reported having tried ious waves of the CISS study show worrying pat-
to sell household assets during the past month. terns.
This share dropped significantly during subse-
quent waves of the CISS survey, to between 8 Thus, by September 2020, almost half of infor-
and 11 percent (see Table 12). Overall, 32.2 mal workers in the sample (44.4 percent) re-
percent of participants in the CISS study had ported to have used up all their savings since
had at some point since the beginning of the the beginning of the pandemic, in March 2020.
pandemic to sell household assets to generate Another 41 percent did not even have any sav-
additional income. In general, men in the sam- ings to start with, while only 5 percent stated
ple engage more in this type of coping strategy not having used their savings. This points to
than women, but this could simply reflect that a situation of generalised financial stress among
fact that ownership of durable goods and other informal workers in Maputo and their families.
household assets in the households is skewed and
tends to lie with men in Mozambique.31 By groups, it was those most vulnerable or most
affected by the socioeconomic impact of the pan-
Households have not only engaged in the sale of demic, including the unemployed and street ven-
durable goods as a way of coping with the neg- dors, who have resorted to greater extent to us-
ative income shocks brought about by the pan- ing their savings. This is likely to have increased
demic. They have also missed important pay- their vulnerability, as without savings they find
ments, such as rents. Hence, of the approx- it harder invest or purchase merchandise for sale
imately one fifth of informal workers in Ma- and, therefore, to sustain their livelihoods.
31
Data from the 2017 National Census of Mozambique On the other hand, there seems to have been a
helps illustrate this. Hence, that year, up to 30.8 percent
of men owned a mobile phone and 12.2 percent had a surge in borrowing among informal workers in
bank account. For women, these percentages were of Maputo, especially during the initial phases of
22.4 and 6.6 percent, respectively. the crisis, with CISS study participants claim-
Table 13: Share of CISS participants that Figure 11: Children’s support in income
borrowed activities
ing in the September 2020 baseline survey to

have had to borrow money coinciding with the
impact of the pandemic reaching 31.5 percent.
This share has dropped since, although remain-
ing at a considerable 17 to 21 percent of the
sample in subsequent waves of the CISS study
(see Table 13).
Close to two thirds of those informal workers

borrowing money have resorted to family, friends
Informal workers and their families have
and relatives, and only around 20 percent have
received limited support
benefitted from bank or microfinance financial
Despite the very challenging environment and
support. These loans have mainly gone to ac-
outlook that informal workers in Maputo have
quire merchandise to sell in the market, although
faced since the beginning of the pandemic, the
a not insignificant part has been devoted to buy
support they have received has been very limited
food and other basic goods for informal workers
– and has remained so throughout the various
households, e.g., 20 percent in the March 2021
waves of the CISS survey. This is not only the
CISS survey.
case of the various initiatives established by the
Families have mostly kept children out of government to support business and vulnerable
work families as part of its response to the COVID-19
On a positive note, a main concern at the out- crisis, for instance through the expansion of so-
set of the pandemic – that the impact of the cial protection programmes, the introduction of
COVID-19 socioeconomic crisis would lead to tax breaks and incentives, or the establishment
a rise in child work – has not materialized, at of subsidized credit lines, as described in Section
least not to a significant extent among those in- 3 of this report. It is also true for other poten-
formal workers and their families taking part in tial sources of help, such as that coming from
the CISS study. Thus, when asked whether fam- national and international NGOs, charities, etc.
ilies had had to get their children to help with
In this sense, a very large majority of infor-
their business during the past month, around 90
mal workers in Maputo taking part in the CISS
percent of respondents throughout all four waves
study have consistently reported not receiving
of the CISS survey indicated this had not been
any kind of support or help: over 90 percent in
the case (see Figure 11).
all four waves of the CISS survey. Whatever sup-
port informal workers have received, it has come
mainly from relatives (see Table 14 for details).
Table 14: Assistance to CISS participants The analysis considers as possible predictors
or explanatory variables key individual, house-
hold, business and employment characteristics
discussed throughout this report. These include
gender, age and level of education of participants
of the CISS study. The size and levels of depriva-
tion they face. The number of people earning an
income in their households. The type of activ-
ity they engage in. The number of years they’ve
been working or running businesses in the in-
formal economy. Their employment status, the
location in Maputo in which they operate and
whether they participate or not in Xitiques.
The following sections present the main results

for each of these outcome-based dependent vari-
ables. Table 15 below presents detailed results
6.3. Understanding drivers of impacts for these pooled logistic regressions, as well as
and coping strategies in the CISS the full list of explanatory variables considered
study for this analysis.
Having analysed how COVID-19 has affected in-
Unemployment
formal workers and their families in various ar-
As discussed in Section 6.1, one of the main im-
eas, this section brings together the main ele-
pacts of the COVID-19 crisis on the informal
ments examined in previous sections to explore,
economy in Maputo has been in unemployment
in a multivariate manner, how different factors
levels among informal workers. The results pre-
are associated with key outcomes of the COVID-
sented in Table 15 (column 1) indicate that sev-
19 crisis, as identified in this report. This mul-
eral of the factors or characteristics under con-
tivariate approach allows to take into considera-
sideration in this report are statistically associ-
tion the interrelation between different explana-
ated with unemployment in the sample, in that
tory variables and find statistically significant
presenting any of these characteristics increases
associations between these explanatory variables
the chances of becoming unemployed.
and specific outcome variables of interest.
Some of these factors present a positive associ-
To this end, this section presents the results
ation, meaning that their occurrence increases
of a simple pooled logistic regression analysis
the chances of becoming unemployed. Others, a
that seeks to identify factors associated with
negative association. Among the former, there is
six dependent dichotomous variables capturing
being a female informal worker, or being located,
some of the main outcomes coming out of the
at some point throughout the CISS study, in the
analysis presented in the previous sections: (1)
“Bairro Central”, the central district of Maputo.
whether informal workers are unemployed or
In this latter case, this positive association with
not; (2) whether they participate in Xitiques;
becoming unemployed is linked to the fact that
(3) whether they have experienced positive earn-
many of these informal workers have been moved
ings’ growth (4) whether they have borrowed
out from Central Maputo, without local author-
throughout the crisis; (5) whether their house-
ities providing alternative locations in which to
holds have had to cut on food; and (6) whether
operate. The chances of being unemployed also
they have had to sell household assets. It uses
increase with the number of deprivations faced
a pooled logistic regression approach to use the
by households in the informal sector. In addition
full set of data collected in the various waves of
to these three factors, having technical educa-
the CISS study and examine these issues.
tion and having been operating or working in the
informal sector for up to 10 years are also posi-
tively associated with becoming unemployed.
On the other hand, factors negatively associated Factors with statistically positive association
with unemployment, for which their occurrence with participating in a Xitique include having
or magnitude decreases the chances of becom- a greater number of members in the household
ing unemployed, include the number of members earning an income, running an informal business
earning an income in the household. In other for more than 5 and 10 years, as opposed to a
words, the larger the number of members of the year or less, which is the comparator age of the
household earning income the less likely an in- firm.
formal worker is to become unemployed. This
could perhaps be explained as the result of “net- The first of these associations is self-evident: the
work effects”, by which informal workers are able greater the number of income earners in a house-
to tap into family business and work networks hold the more savings or surplus cash they are
to remain employed. It also includes participat- likely to be able to generate and, hence, the like-
ing in a Xitique. The likely explanation behind lier it is that these households continue to put
this negative association is that, if people are part of these savings into a Xitique. The sec-
participating in a Xitique, it’s because they are ond of these associations is perhaps the result
still earning an income or have money saved that that older businesses are likely to be more es-
they don’t need and, therefore, they can put into tablished in the line of business in which they
a rotational savings scheme, meaning that they operate and hence they perhaps perform better
are unlikely to be unemployed. Another factor and are better able to withstand business shocks
that lowers the probability of unemployment is such as that caused by COVID-19.
having a primary education level compared to
no formal education at all. Factors negatively associated with participating
in a Xitique include age of informal workers,
Finally, the non-significant coefficients for the the number of deprivations faced by households
Wave dummy variables presented in Table 15 in the sample, as well as being unemployed.
would confirm that, on average, we cannot ob- In other words, the older an informal worker
serve any trend in unemployment vis-à-vis the and the more deprived is his/her household, the
baseline survey. lower the chances of participating in a Xitique.
The same happens to those unemployed. These
Overall, the above results underscore the fact results make sense, considering that participat-
that it is the most vulnerable who are more ing in a Xitique can be taken as a proxy of being
exposed to the unemployment phenomena de- relatively well off.
tected in the Maputo informal sector during the
COVID-19 pandemic. It is women, deprived and Finally, the negative and statistically significant
multidimensionally poor households and families coefficients found for all three CISS follow-up
with fewer income earners that are likelier to be- surveys (Wave 1, 2 and 3) confirms that Xitique
come unemployed during COVID-19. participation has followed a downward trend
throughout the pandemic. That is, the situa-
Xitiques Participation tion has consistently worsened over time on this
As indicated in previous sections, Xitiques are a particular performance metric.
popular form of rotating savings scheme among
informal workers in Maputo and are widely Altogether these findings point to the fact that
used to manage surplus, savings and investment it is the more vulnerable – deprived households
needs. As with the analysis of unemployment, with fewer income earners or the unemployed –
the pooled regression analysis of Xitique partici- who have been more affected by the impact of
pation yields statistically significant associations the pandemic and less able to continue partici-
in both directions, both positive and negative, as pating in Xitiques which, in the context of the
in decreasing or increasing the changes of infor- Mozambican informal economy, is a key invest-
mal workers participating in Xitiques during the ment and cashflow management mechanism.
pandemic (see Table 15, column 2).
Increases in earnings The only statistically significant variables asso-

As outlined in section 6.1, despite the very large ciated with borrowing are that of age of respon-
negative impact that COVID-19 has had on dents and their unemployment status. Age has a
the informal economy in Maputo, some informal negative impact on the probability of borrowing.
workers have managed to come out relatively un- In other words, older workers are less likely to
scathed, reporting to have seen their earnings in- borrow money. Unemployment also negatively
crease at some point since the beginning of the influences the probability of borrowing. The rea-
pandemic. Table 15, column 3, presents the re- son behind this might be that without any in-
sults of a pooled logistic regression of the odds come, unemployed informal workers cannot af-
of positing growing in earnings at some point ford to borrow or, simply, cannot find a mon-
during the various waves of the CISS study that eylender or financial institution willing to give
seeks to identify participants’ characteristics as- them a loan.
sociated with this outcome.
The three dummy variables capturing each of
As can be seen from the results presented in this the three follow-up waves done after the base-
table, several variables present statistically sig- line survey are also statistically positive, with
nificant coefficients: being a female (negative), all three presenting a negative coefficient. This
having secondary or technical education (nega- indicates that overtime, borrowing has consis-
tive), the number of contributing members to tently been less frequent during the pandemic.
household income (positive), having worked in This is important, in that this declining trend in
the central district of Maputo (positive), plus borrowing may be hampering informal workers’
all three trend – Wave 1, Wave 2 and Wave 3 – capacity to invest in their businesses or obtain
variables (positive). working capital with which to purchase mer-
chandise to sell.
We can interpret these results as indicating that,
other things equal, reporting increases in earn- All other variables under consideration for this
ings is associated with being a man, not hav- logistic regression of borrowing patterns in the
ing secondary or technical education, living in sample, such as the age of businesses, household
a family with a larger number of income earn- size or the number of people earning an income
ers and not having worked in central Maputo. in the household, are statistically not significant,
Conversely it can be said that women, those meaning we cannot establish a statistically ro-
who have worked in the central district of Ma- bust association between these variables and in-
puto or those living in more “income vulnera- formal workers borrowing behaviour.
ble” households are less likely to report increases
in earnings. Again, this points to the fact that Nutrition in the household
those people who are or tend to be more vulner- As described in section 6, a considerable number
able are those fairing worse on this performance of informal workers and their families had to re-
metric. Finally, the positive coefficients on the sort to negative coping strategies in response to
trend variables simple indicate that the number the COVID-19 crisis, including cutting back on
of workers reporting increases in earnings has meals and on the amount of food eaten at home.
increased over time. Column 5, in Table 15, presents the results of
our logistic regression for this coping strategy.
Borrowing
As discussed in section 6.2., many informal Not surprisingly, the more deprived respondents
workers in the CISS study have had to borrow and their families are, proxied by the variable
money in order to deal with the loss of profit and capturing the number of deprivations faced by
income during the COVID-19 pandemic. The lo- informal workers’ households, the higher the
gistic regression results on borrowing behaviour likelihood of cutting back on food. The re-
presented in Table 15 (column 4), which in- sults also suggest that families with fewer people
clude mostly the same variables as for unem- earning money have greater likelihood of cutting
ployment and Xitiques, present who is likelier back on food. Other variables, such as the size
to borrow in the sample. of the household, the maturity of business or the
age of the worker show no statistically significant Finally, it is important to note that the positive
impact on the probability of borrowing. and statistically significant coefficients found for
all three CISS follow-up surveys (Wave 1, 2 and
Importantly, two of the coefficients for the 3) in this logistic regression confirms that the
dummy variables capturing each of the three sale of household assets has followed an upward
follow-up surveys done for the CISS study, those trend throughout the pandemic. That is, the
for Waves 2 and 3, are negative and statistically situation has consistently worsened on this par-
significant, meaning that over time, and espe- ticular duress metric, with the sale of assets be-
cially since December 2020, problems of nutri- coming more common over time among informal
tion and lack of food in informal workers’ house- workers.
holds have, on average, become less of an issue.
Selling household assets 7. Conclusion: Main findings and pol-

Another trend observed during the pandemic icy recommendations
among informal workers in the CISS study is
the sale of household assets. For many, this 7.1. Main findings
has served o generate additional cash to be used
by households in a context of extreme hardship. The COVID-19 Informal Sector Survey (CISS)
The results of the logistic regression analysis pre- study has provided a rich set of data of informal
sented in Table 15 (column 6) shows the cor- workers in the city of Maputo and their families,
relations between trying to sell household assets as well as of the socioeconomic impact of the
and selected explanatory variables. COVID-19 crisis in this particular group of the
population.
Age has a negative relationship with the prob-
ability of selling household assets. That is, The results of this study indicate that the pan-
younger people are more likely to sell house- demic has had a very significant and negative
hold goods than older ones. A gender differ- impact on informal workers in Maputo and their
ence comes to light once again – women are less families. The impact has been felt across many
likely to sell their assets. The reason behind lies areas: on employment, on hours worked, on
probably in the much lower ownership of goods earning and savings, and hence the ability to in-
among women than men in Mozambique and in vest. Also, on an increasingly challenging busi-
the CISS sample, in the first place. Thus, while ness and market environment, with rising prices
53 percent of female participants do not pos- for goods that informal workers purchase to re-
sess at least three durable goods (among radio, sell, falling demand and fewer clients, or restric-
tv, computer, smartphone, gas or electric stove, tions on business hours. Families of informal
fridge, car, motorbike, or bicycle), this only ap- workers have also suffered, using up family sav-
plies to one third of the men in our sample. In ings to make ends meet, cutting back on meals,
addition, the results of the logistic regression or selling household durable goods, among oth-
analysis also suggest that the higher the level of ers. In addition, in middle of the pandemic,
deprivation a household suffers, the higher is the many informal workers have had to deal with
probability it sells durable assets. This is a con- the added challenge of having to move business
cerning insight, underlining the risks of the poor location, as part of the municipal government’s
becoming even poorer during the pandemic. urban regeneration efforts in the central district
of Maputo. Many, in fact, have become unem-
The results also show a positive relationship be- ployed as a result of these relocation measures.
tween the age of businesses and the likelihood
of selling household goods. In other words, the Data collected in the various rounds of the CISS
longer a worker has been in the informal econ- survey indicates that this impact was particu-
omy the more likely s/he is to sell household larly intense during the initial stages of the cri-
assets. Worryingly, this means that even people sis, from March to September 2020, yet has con-
working for years have had to resort to selling tinued to unravel over ensuing months. Hence,
household assets. while the situation seems to have stabilised to-
wards the last quarter of 2020, on many ac- the way for post-pandemic recovery in the infor-
counts the overall situation of informal workers mal economy.
has continued to deteriorate, albeit at a slower
pace. This, of course, doesn’t mean that all in- 7.2. Policy recommendations
formal workers have been negatively impacted
by the pandemic. A few have managed to see The findings coming out of the various rounds
their earnings increase at some point during of the CISS survey point to a number of areas
the pandemic, with these numbers growing over where action could contribute to ameliorate the
time – but still remining small in comparison to situation of informal workers in Maputo in the
those negatively affected by the COVID-19 cri- context of the pandemic. It also provides useful
sis. Against this background, informal workers findings that can help inform how best to sup-
in Maputo remain surprisingly optimists about port informal workers and their families going
future business prospects. However, this opti- forward and how best to respond to crisis such as
mism has waned over time. the COVID-19 pandemic. This section outlines
some of these findings and recommendations.
The results of this study suggest that informal
workers and their families were already consid- • One of the main findings coming out of the
erably vulnerable group of the population going study is that there is a significant number
into the pandemic. The COVID-19 crisis, if any- of informal workers who have become
thing, seems to have exacerbated these vulnera- unemployed coinciding with the pandemic.
bilities. It is important to note, in this regard, Many of these find themselves in this situa-
that our analysis indicates that it is those who tion after being relocated from their regular
were already more vulnerable who have fared places of work by the local authorities in
worse during the pandemic. It is women and Maputo without being provided alternative
informal workers in families with fewer income locations. The situation of these people
earners and suffering from greater deprivations could be improved by providing new spaces
who have been harder hit by unemployment. for them to operate, for instance, in the
It is older people, the unemployed and, again, new market area in Laulane or in other lo-
those suffering from greater deprivations, who cations in existing markets around Maputo.
are not able to save or borrow and, therefore,
to invest in a way out of crisis. It is the most • Consistently, throughout the various
deprived families and those with fewer bread- rounds of the CISS survey, informal
winners who are having to cut back on meals workers in Maputo have identified rising
and the amount of food they bring to the plate. prices for the merchandise they sell as
And it is these same families who are finding one of the main difficulties they’ve faced
themselves selling household assets. Against this throughout the pandemic. This finding
highly challenging context, very few of the infor- is consistent with that of other studies.32
mal workers taking part in the CISS study have It is unclear what might be driving these
received any support or help, whether from gov- price dynamics. During the initial stages
ernment, relatives or local NGOs and charities. of the COVID-19 crisis global supply chain
This has been the case since the first baseline disruptions might have been behind these
survey was undertaken in September 2020, all dynamics, but these disruptions seem
the way to the last wave, in March 2021. to have now been solved. The relevant
government authorities could perhaps
Over time, as the pandemic unfolds and contin- investigate this issue further.
ues to affect informal workers and their families
in Maputo in the coming months, these dynam- • The impact of the pandemic has resulted
ics carry the risk of chronifying these people’s in reduced earnings for informal workers
vulnerability, making it harder for them to over- and has forced many of them to use their
come this situation. In this sense, action is re- 32
See, for instance, TechnoServe (2020). “COVID-19
quired to mitigate the situation currently faced Impacts on Micro-Retail Businesses”. Women in Busi-
by informal workers in Mozambique and pave ness (WIN) Programme. Maputo, April.
savings and, in the extreme, sell household • More generally, the experience of the infor-
assets. This has left them with no working mal sector in Maputo suggests there might
capital with which to purchase merchan- be a need to rethink how best to assist, from
dise, or with which invest, for instance in a social protection perspective, those in
in new workspaces (e.g. in building new need in emergency situations. Hence, estab-
market stalls). The very sharp decline lished social protection schemes typically
in Xitique participation attests to this. address the needs of “standard” vulnera-
In this respect, part of the COVID-19 ble groups: the elderly, single parent fami-
response could go to provide liquidity or lies, people with disabilities, etc. However,
subsidized loans to informal workers, for emergency situations may give rise to new
instance through local microcredit schemes. groups of vulnerable people, people whom,
moreover, might require urgent assistance.
• Given the difficulties faced by informal In this context, existing social protection
workers in accessing the government’s cash- programmes may not be best suited to ad-
based social protection scheme that has dress these needs, especially if requirement
been established to support the pandemic complex administrative processes and logis-
response, alternative support approaches tics (bank accounts, mobile phones, etc.).
could be considered. For instance, soup or Other social protection approaches might,
community kitchens could be established in this regard, provide a more effective way
in partnership with local CSOs, CBOs or to respond to these emergency situations.
charities targeting specific residential areas
where the more vulnerable segments of the
population in Maputo live. This could help
mitigate some of the difficulties identified
in the CISS study that families in the
informal sector face in regularly having the
recommended three meals per day.
• Similarly, local authorities or local char-

ities, CSOs and NGOs could consider
establishing cash-for-work programmes for
those workers in the Maputo informal sec-
tor who have become unemployed during
the pandemic, which as shown in the CISS
survey are a considerable number.
Table 15: Pooled logistic regression results

(1) (2) (3) (4) (5) (6)
Positive Cutting
Becoming Participating Sale of HH
earnings Borrowing back on
unemployed in Xitique assets
growth food
Age 0.00346 -0.0163** -0.005 -0.0326*** 0.000478 -0.0433***
(0.00673) (0.00570) (0.008) (0.00620) (0.00573) (0.00859)
Female 1.020*** 0.0831 -0.671*** 0.212 -0.0194 -1.078***
(0.147) (0.118) (0.157) (0.123) (0.114) (0.160)
Base: No education
Primary education -0.562* 0.379 -0.103 0.367 -0.493* 0.128
(0.266) (0.262) (0.304) (0.285) (0.245) (0.487)
Secondary education 0.0452 0.392 -0.680** 0.481 0.267 0.438
(0.297) (0.284) (0.339) (0.305) (0.265) (0.502)
Technical school 1.149* -0.413 -0.896* 0.808 0.0376 0.0733
(0.454) (0.458) (0.503) (0.457) (0.416) (0.685)
University 0.837 0.637 -0.557 0.0797 -0.381 0.132
(0.565) (0.490) (0.700) (0.561) (0.436) (0.721)
Nr. of deprivations 0.217*** -0.220*** -0.120 0.0467 0.670*** 0.323***
(0.0583) (0.0526) (0.078) (0.0536) (0.0579) (0.0750)
Size of HH 0.0407 0.0249 -.0130 0.0216 0.0290 -0.0154
(0.0222) (0.0191) (0.026) (0.0203) (0.0189) (0.0270)
Nr. of HH contributors -1.393*** 0.133* 0.188** -0.0885 -0.211*** -0.144
(0.155) (0.0672) (0.080) (0.0695) (0.0627) (0.102)
Worked in Central 1.534*** -0.0147 -0.806*** 0.0188 0.198 0.539**
(0.133) (0.125) (0.179) (0.136) (0.126) (0.175)
Base: Business age: <1
year
Business age: <3 years 1.347*** 0.362 -0.293 0.461 -0.154 0.821*
(0.282) (0.245) (0.361) (0.254) (0.240) (0.343)
Business age: <5 years 0.703** 0.765** 0.127 0.115 -0.0310 0.764*
(0.269) (0.234) (0.323) (0.253) (0.233) (0.349)
Business age: <10 years 0.803** 0.410 -0.261 0.377 -0.0641 0.863*
(0.258) (0.223) (0.328) (0.237) (0.220) (0.336)
Business age: >10 years 0.449 0.642** 0.125 0.323 0.0649 0.817*
(0.238) (0.211) (0.295) (0.229) (0.213) (0.319)
Base: Baseline Wave
Wave 1 -0.153 -1.913*** 2.472*** -0.869*** -0.203 -1.245***
(0.178) (0.148) (0.275) (0.162) (0.153) (0.210)
Wave 2 -0.0488 -1.776*** 2.460*** -0.603*** -0.412** -0.900***
(0.173) (0.146) (0.266) (0.154) (0.148) (0.188)
Wave 3 -0.309 -1.925*** 2.574*** -0.589*** -0.491*** -1.340***
(0.179) (0.148) (0.265) (0.153) (0.148) (0.212)
Xitique participation -1.157*** 1.117*** -0.0144 -0.164 -0.174
(0.138) (0.163) (0.122) (0.115) (0.157)
Baseline activity: Street
0.127 0.0590 -0.119 -0.163
merchant
(0.183) (0.210) (0.204) (0.260)
Baseline activity: Fixed
0.277 -0.114 -0.236 -0.168
merchant
(0.164) (0.200) (0.191) (0.245)
Baseline activity: Services 0.191 -0.227 -0.132 0.121
(0.264) (0.305) (0.282) (0.337)
Unemployed -1.057*** -0.832*** 0.302 -0.0326
(0.158) (0.203) (0.186) (0.244)
Constant -1.354** 0.862 -2.638 -0.182 0.260 -0.413
(0.519) (0.489) (0.667) (0.527) (0.487) (0.768)
Observations 2071 2071 1455 2071 2071 2071
Standard errors in parentheses
Note: Each regression is a Pooled logistic regression with robust standard errors.
* p < 0.05, ** p < 0.01, *** p < 0.001

UNDP MZ Socioeconomic Impact

Uploaded by

Copyright:

Available Formats

UNDP MZ Socioeconomic Impact

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UNDP MZ Socioeconomic Impact

Uploaded by

Copyright:

Available Formats

THE SOCIOECONOMIC IMPACT OF

COVID-19 ON THE URBAN INFORMAL

Maputo 12th November 2021

Ministry of Economy and Finance

ASSOTSI – AEIMO (Associação da Economia Informal de Moçambique)

UNU Wider – Inclusive Growth Programme

2 COVID-19 and the informal economy – an international perspective 5

3 The impact of COVID-19 and the informal economy in Mozambique 7

4 The COVID-19 Informal Sector Survey (CISS) 11

5 Characteristics of the sample – a highly vulnerable segment of the population 13

6 The socioeconomic impact of COVID-19 on the informal economy in Maputo 17

7 Conclusion: Main findings and policy recommendations 30

1. Introduction vey and three follow-up waves – implemented be-

Overall, ILO estimates that almost 1.6 billion in-

Even though vaccination programmes have

It is important to note that the impact of

3.1. The Socioeconomic impact of

As elsewhere around world, the COVID-19 pan-

A nationally representative survey of 1,333

In response to these impacts, the government

3.2. The informal economy in Table 3: Number of informal workers aged 7+

The informal economy in Mozambique is esti-

First, as an intrinsic part of the urban land- 4.2. Methodological considerations

Table 5: Profile of participants 5.2. Business and employment

5.3. Characteristics of households in the

Table 7: CISS household characteristics Figure 4: Asset ownership in CISS households

In general, living conditions for families in the

Figure 5: Deprivations and multidimensional 6. The socioeconomic impact of

Having drawn the characteristics of informal

The September 2020 baseline survey for the

This is the case, for instance, of the number of

Figure 8: Contributions to Xitiques (MZM) Informal businesses and workers have

When asked about the reasons for switching ac-

Hence, in September 2020, 62.3 percent of in-

As displayed in Table 11, at the time of the

Many families are finding it challenging

It is important to note that nutritional stress

ing in the September 2020 baseline survey to

Close to two thirds of those informal workers

The following sections present the main results

Increases in earnings The only statistically significant variables asso-

Selling household assets 7. Conclusion: Main findings and pol-

• Similarly, local authorities or local char-

Table 15: Pooled logistic regression results

You might also like