COVID -19 Fallout on Poverty and Livelihoods in Bangladesh: Results from SANEM’s Nation-wide Household Survey (November-December 2020)

Selim Raihan, Mahtab Uddin, Md. Tuhin Ahmed, Mir Ashrafun Nahar, Eshrat Sharmin

Research output: Book/ReportBook

301 Downloads (Pure)

Abstract

Executive Summary
Until the onset of COVID-19 in March 2020, Bangladesh made an impressive reduction in the
poverty rate from as high as 56.7% in 1991-92 to 20.5% in 2019. Despite this remarkable
alleviation, most of the people who graduated remained close to the poverty line income –
thus remained as the vulnerable poor. In the pre-pandemic situation, nearly half of the
population in the country were within the threshold of vulnerable poverty. Given this context,
any major economic shock, such as the COVID-19 pandemic, is obvious to leave dents on the
progress achieved in alleviating poverty over the past decades. A thorough assessment is
warranted to tackle the pandemic’s multi-dimensional ramifications on the economy,
particularly on Poverty, Inequality, and Employment (PIE). Understanding the dynamics of PIE
in the pre-COVID and the post-COVID situation is critical to achieving inclusive economic
growth as per the agenda of the SDGs, the 8th Five Year Plan (8FYP), and the Perspective Plan.
SANEM, through a nationwide survey in November-December 2020, aimed to fulfil this
objective.
The study investigates poverty, income and employment scenarios from pre-COVID to postCOVID. The 2020 survey is built on a survey conducted by SANEM in 2018. SANEM, in
collaboration with the General Economic Division (GED) Planning Commission, conducted a
nationally representative survey of 10,500 households in 2018. To understand the impact of
the pandemic on PIE in the pre and post COVID-19 periods, SANEM attempted to reach all
10,500 households from the 2018 survey. Given the ongoing pandemic situation, SANEM
surveyed over the phone in November-December 2020. Among the 10,500 households,
SANEM successfully interviewed 5,577 households from 500 Primary Sampling Units (PSUs)
distributed across eight divisions and 64 districts. The survey non-response was 10%. The
team could not reach 37% of the households due to network conditions, language barriers,
and wrong numbers. Close attention was given to analyse any systematic bias in the
responses or success rate given such attrition.
A careful checking for the bias was done based on several observable characteristics of the
households such as sample distribution by divisions and regions, sex of the household head,
household head’s primary occupation, household’s main income sources, distribution of the
households by income deciles, and education level of the household head. A comparison was
made for the households covered in 2020 with those who were not covered, and the overall
distribution of households surveyed in 2018 based on the observable characteristics. The
covered households’ attributes appeared the same as the non-covered households in all the
parameters without any statistically significant difference showing no systematic bias.
The 2020 survey questionnaire included questions pertinent to households’ basic
characteristics, education, employment, COVID-19 led major challenges and coping
strategies, social protection, health, migration, and remittances, along with pre-COVID and
during-COVID household income and expenditure information. For better comparison, the
pre and post-COVID-19 impacts on PIE for these 5,577 households were analysed.
Using the Cost of Basic Needs (CBN) method, the upper and lower poverty lines for 20 strata
(eight rural, eight urban, and four metropolitan areas) were calculated based on the 2018
xi
survey dataset. Each of the poverty lines was then updated to 2020, adjusting for inflation
following a systematic approach. In updating the poverty lines for changes in inflation rates
between 2018 and 2020, rural, urban and metropolitan areas were given differentiated
weights. The updated upper poverty line (UPL) per person per month for rural areas ranged
from Tk. 2246 (Barisal) to Tk. 2936 (Dhaka). For the urban areas, it ranged from Tk. 2604
(Khulna) to Tk. 3322 (Dhaka Metropolitan). The rural lower poverty line (LPL) ranged from Tk.
1912 (Barisal) to Tk. 2561 (Dhaka), while the urban LPL ranged from Tk. 1953 (Rajshashi) to
Tk 2800 (Sylhet).
Based on the updated poverty lines, it was found that the upper poverty rate almost doubled
from 21.6% in 2018 to 42.0% while the lower poverty rate tripled from 9.4% to 28.5%. The
poverty rate expanded faster in urban than in rural areas. In the urban areas, the upper
poverty rate more than doubled from 16.3% to 35.4% while in the rural areas the rate climbed
up from 24.5% to 45.3%. In the case of lower poverty, the rate tripled in both rural (33.2%)
and urban (19%) areas compared to the respective rates in 2018. A regional pattern also
emerged: the western divisions registered higher poverty rates than the eastern divisions.
The highest poverty rate was observed in Rangpur (57.3%), followed by Rajshahi (55.5%),
Mymensingh (46.2%), Khulna (41.8%), Dhaka (38.4%), Chattogram (35.1%), Sylhet (35%), and
Barisal (29.3%).
Given the panel dimension of the dataset, the dynamics of new poor were further delved -
who fell back and who graduated out of poverty. Of extreme poor households in 2018, 46.2%
of them remained extreme poor in 2020. Interestingly, 15.8% of these households graduated
to upper poverty, 17.7% moved to the vulnerable poor category (where the vulnerable
poverty line is defined as 1.25 times the UPL), and the rest moved to the non-vulnerable nonpoor category. Contrastingly, among the moderate poor households in 2018, 41% of them fell
back to extreme poverty. Another 18.7% of these households moved up to the vulnerable
poor group while 22.9% graduated to the non-vulnerable non-poor category. The most
significant dip in poverty is observed for the vulnerable poor households in 2018: 34.8% fell
back to extreme poverty, while another 14% fell back to moderate poverty. Amongst the nonvulnerable non-poor households, 20% fell below the extreme poverty line, 12% fell below
moderate poverty, and 18% became vulnerable poor.
The aforementioned ‘falling back to poverty’ dynamics is primarily linked to the households’
sharp income/expenditure falls in 2020. A large number of the households experienced a fall
in their per capita household expenditure, in absolute terms, in 2020 compared to the
respective levels in 2020. The most significant fall in per capita expenditure was observed for
the extreme poor households (45%) followed by moderate poor (29%) and vulnerable poor
households (17%). Conversely, non-vulnerable non-poor households had an increase in per
capita expenditure by 6%. The extreme poor and moderate poor households cut through their
food expenditure (30% and 15% respectively) and their non-food spending (63% and 49%
respectively). While the vulnerable poor households also cut in both food and non-food
expenditures (17% and 2% respectively), the non-vulnerable households increased their food
expenditure (in absolute terms) by 17% compared to 2018.
To better understand the new-poor, households were categorised as “old-poor” and “newpoor” depending on whether they were already poor before the pandemic or fallen below
xii
the poverty line during the pandemic. The findings show that in the “old poor” household
category, 37% of household heads were self-employed, 20.5% were wage-employed, and
39.5% were day labourers. In contrast, in the “new poor” households, 42.3% of household
heads were self-employed, 23.9% were wage-employed, and 30.2% were day labourers. For
the primary source of income, among “old-poor” households, 43.4% relied on agriculture,
5.2% on the industry, 46.5% on service, and 3% on remittances. In contrast, among “newpoor” households, 36.6% relies on agriculture, 6.4% on the industry, 51.2% on service, and
3.2% on remittances.
The change in inequality has been observed with the Gini coefficient. The consumption
expenditure Gini coefficient increased from 0.31 in 2018 to 0.33 in 2020. Such an increase in
inequality primarily originated from the fall in income (expenditure) for the poorer income
(expenditure) groups compared to the richer groups. The ratio of income shares between the
richest 5% and poorest 20% households increased from 2.05 in February 2020 to 2.45 in
November 2020. Correspondingly, the ratio of expenditure share of the richest 5% to that of
the poorest 20% increased from 1.34 in 2018 to 2.15 in 2020. The expenditure share of the
richest 5% households increased by 1.02 percentage points even weathering this pandemic,
whereas, for the poorest 20%, it declined by 3.13 percentage points. One critical point to
remember is that since most ultra-rich households could not be included in the survey, the
actual impact on inequality might be much more significant than found in this survey.
The rise in inequality didn’t limit to the income dimension only. There was a widening gap in
investment in human capital (education and healthcare). Overall, the average per capita
education expenditure fell for all households between 2018 and 2020. However, the fall was
as high as 58% for the extreme poor households, followed by moderate poor households
(41%) in contrast to non-vulnerable non-poor households who cut it down only by 9%. Also,
while the average per capita health expenditure increased for all households, the least
increase was for the extreme poor (only 3%). The largest increase was for the non-poor nonvulnerable households (104%). Not to mention, the poor households spent only a fraction of
the expenditures incurred by non-poor-non-vulnerable households on education and
healthcare.
There appeared a digital divide too. The access to online/TV education was also largely
heterogeneous. Only 21% of the households reported that their children could participate in
online/TV education. The gap between the rural and urban areas is noteworthy - 19% and
27%, respectively. The digital divide by poverty status is also clearly evident. In oppose to 26%
of the non-poor households, only 15% of the poor households reported that their children
participated in some form of online/TV education. Nevertheless, less than a third of the
respondents mentioned online classes as effective. Regarding the reasons behind not joining
the online/TV classes, the respondents mentioned the unavailability of online classes (49.1%),
no access to technological devices (6.1%), insufficient access to devices (5.3%), inadequate
access to an internet connection (5.4%), inability to bear the cost of internet connection
(6.5%), amongst others.
Alarmingly, around 3% of the households responded that they were not sure about continuing
their currently enrolled children (rural 3.7%; urban 1.4%). The rate was the highest for Sylhet
(4.71%), followed by Khulna (4.7%), Barisal (3.4%), Dhaka (2.9%), Chattogram (2.8%), Rangpur
xiii
(2.8%), Mymensingh (2.7%), and Rajshahi (1.5%). Reasons for not continuing education
included unaffordability of the households to continue (national 68%; rural 67%; urban
73.7%), being already involved in economic activities (national 17.2%; rural 17.5%; urban
15.8%), and being married (national 9%; rural 13.6%; urban 5.3%), amongst others.
The impact on employment was not homogenous for all households. Among the surveyed
households, 55.9% responded that despite being employed, the household’s primary earner’s
income had fallen since March 2020. Around 8.6% of the households claimed that they lost
work during March-November 2020, 7% claimed that working hour was reduced, and 33.2%
reported that their work stopped at least for a while during the outbreak. Only 17.3% of
households responded that they were involved in economic activities without any disruption.
Between February and October 2020, the primary income earners across all employment
categories experienced a fall in average incomes: the decline was 32% for self-employed, 23%
for wage-employed, 29% for day labourers, and 35% for other categories.
The occupational mobility across industries was also observed between 2018 and 2020. In
2018, agriculture was the primary source of income for 26% of the households, followed by
the services sector (46.4%), industry (17.4%), and remittances (8.6%). In 2020, 29.4% of the
households relied on agriculture as the main source of income, while the dependence on the
services sector and the remittances declined to 44.7% and 4.9% respectively. Moreover, in
2018, 57.3% of the households' main earners were self-employed, which fell to 45.1% in 2020.
Compared to 2018, in 2020, the main earners’ occupation share in the wage-employment
category increased by nine percentage points to 27.6%.
The study further delved into the apparent paradox of remittance inflow in 2020. The official
foreign remittance receipts soared even during the pandemic. However, in this survey, 82.1%
of the foreign-remittance receiving households claimed that they received fewer remittances
during the months between March and November 2020 compared to a similar period a year
ago. Only 0.3% of the households reported experiencing a rise in remittance incomes. A fall
in the amount of internal remittances was also observed: 64% of such remittance-receiving
households claimed that they received less during most of the months in 2020 compared to
what they received in the pre-pandemic months. A possible explanation for this paradox is
that a substantial amount of remittance was received through informal channels before the
pandemic. Since those channels had been blocked and there had been incentives from the
Government of Bangladesh, a large proportion of sent remittances took the formal channels
diverting from the informal routes (like Hundi). Moreover, many workers lost their jobs in
the overseas markets, faced pay cuts, many could not repatriate back to work due to travel
bans, amongst other challenges.
More than two-thirds of the households responded that they faced several critical challenges
during the pandemic. Among these households, around 1.3% responded that their family
suffered due to COVID-19 infection or death of any family member due to coronavirus (rural
0.95%, urban 1.9%). Serious illness or death of an earning member of the family (not from
COVID-19) was a major challenge for 5.7% of the surveyed households. Nearly half of the
households responded unusually high price of daily necessities as a major challenge (rural
50.1%, urban 48.7%). Amongst other major challenges faced by the households included:
income of the main earner of the family stopped (national 15.6%; rural 14.1%, urban 18.3%),
xiv
and distraught due to floods, landslides or river erosion (national 13.25%, rural 16.1%, urban
7.6%).
In reaction to the crisis, households adopted a variety of coping strategies, often from
multiple sources such as borrowing (48.7%), reliance on savings (32.4%), reduced expenditure
on non-food items (27.3%), involuntary change in dietary patterns (27%), donations from
friends/relatives (16.7%). Alarmingly, 7.5% of the households responded that they could not
cope with the problem at all.
Regarding getting support from private or public organisations during the pandemic, 32.9%
of households from the poorest expenditure quantile reported receiving some forms of
support (cash or in-kind) from private organisations. In comparison, 25.9% received benefits
from government initiatives. For the richest expenditure quantile, the figures were 24% and
15.54% respectively. When the households were further asked whether they found the
government supports as sufficient, only 22.1% of them perceived such support measures as
enough. About the ability to cope with the COVID-19 induced crisis and return to normalcy,
only 27.2% expressed optimism.
This survey comes up with first-hand numbers from the field that the policymakers can take
on the table to adequately revise the strategies and devise short- and long-term policies
where required. For example, five key suggestions emerged from the respondents: (i) better
management of the crisis (ii) increasing social safety net coverage, including direct cash
transfer to the poor, (iii) price stability of essential products (iv) reduction of corruption, and
(v) creating employment opportunities. In conclusion, this survey provides the necessary
evidence for recalibrating the policymaking process towards an effective recovery.’
Original languageEnglish
Place of PublicationDhaka, Bangladesh
PublisherSANEM Publications
ISBN (Print)978-984-35-1314-4
Publication statusPublished - Sep 2021

Keywords

  • Poverty dynamics
  • Covid-19 pandemic
  • Covid-19 and Inequality

Research Beacons, Institutes and Platforms

  • Global Development Institute

Fingerprint

Dive into the research topics of 'COVID -19 Fallout on Poverty and Livelihoods in Bangladesh: Results from SANEM’s Nation-wide Household Survey (November-December 2020)'. Together they form a unique fingerprint.

Cite this