See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/350129158

Impact of COVID-19 Pandemic on Labour Supply, Wages and Gross Value

Added in India

Article in The Indian Economic Journal · March 2021

DOI: 10.1177/0019466221999143

CITATIONS READS

12 1,023

4 authors:

Xavier Estupinan Sargam Gupta

International Labour Organization Indira Gandhi Institute of Development Research
16 PUBLICATIONS 75 CITATIONS 8 PUBLICATIONS 47 CITATIONS

SEE PROFILE SEE PROFILE

Mohit Sharma Bharti Birla

Collaborative Research and Dissemination International Labour Organization
14 PUBLICATIONS 71 CITATIONS 7 PUBLICATIONS 46 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Sargam Gupta on 19 March 2021.

The user has requested enhancement of the downloaded file.

Impact of COVID-19 Pandemic on Labour Supply, Wages and Gross Value Added
in India*
Xavier Estupinan1, Sargam Gupta2, Mohit Sharma3, and Bharti Birla4

[Accepted Version, The Indian Economic Journal]

Abstract

This paper estimates the first order supply shock through labour supply reduction associated
with the containment measures taken by the Government of India to control COVID-19 spread.
We provide the estimates for Lockdown 1.0 and Lockdown 2.0, from March 25 to May 3, 2020,
when India had the highest stringency measures in the world. To get an extensive impact of the
COVID-19 pandemic on the labour market we carry out an in-depth analysis of labour supply
shocks by employment status, industry level and occupation. The workers impacted are those
who work in a non-essential industry and are not able to work from home. To identify jobs that
cannot be done from home we use a novel approach and construct an occupation-based Remote
Labour Index (RLI) for India. Using the PLFS (2017-18) we find that 116.18 million (25 per
cent of the total employed) and 78.93 million (17 per cent of total employed) workers were
affected during Lockdown 1.0 and Lockdown 2.0, respectively. The expected monthly wage
and income loss to workers is estimated to be Rs. 864.5 billion (2017-18 prices). Further, the
reduction of Gross Value Added (2012-12 prices) is estimated at 14 per cent compared to a no-
COVID scenario.

Keywords: COVID-19 Pandemic, Remote Labour Index, Labour Supply Shock, Wage loss,
Gross Value Added, ARIMA Modelling

JEL Codes: E01, J21, J22, J24, J33, J38

* We would like to thank Alankar Gupta, Tejpal Singh Rathod and Anamika Das for their research assistance. We would also
like to thank R. Nagaraj, K.V. Ramaswamy, webinar participants at Strategic Research Unit of RBI, St. Xavier’s College,
Mumbai for their helpful comments. We would also like to thank two anonymous referees for their useful comments. The
views expressed in this paper are those of the authors and do not reflect the opinion or views of the organizations they are
affiliated with. This paper is circulated as a working paper under the title “Impact of COVID-19 Pandemic on Labour Supply
and Gross Value Added in India”.

1 International Labour Organization, Delhi, India. E-mail: estupinan@ilo.org

2 Corresponding Author: Indira Gandhi Institute of Development Research, Mumbai, India. E-mail: sargam@igidr.ac.in
3 Collaborative Research and Dissemination, Delhi, India. E-mail: mohit.mse@gmail.com
4 International Labour Organization, Delhi, India. E-mail: birla@ilo.org

1
1. Introduction

COVID-19 pandemic has unleashed an unprecedented crisis affecting millions of people

around the globe. According to Coronavirus Resource Centre at the John Hopkins University,
the number of COVID-19 confirmed cases worldwide exceeded 18.5 million as of 5th August
2020 (CSSE, 2020). As the pandemic spreads, governments face a double challenge of
controlling the spread of infections and managing its social and economic impact. To flatten
the epidemic curve and contain the virus outspread, measures like lockdown, travel restrictions
and strict social distancing norms are required. But at the same time, these containment
measures taken by the governments can have a detrimental effect on the economy as it puts a
large part of the economy to a halt. Koren and Peto (2020) show how social distancing
interventions can be effective against epidemics but are potentially detrimental to the economy.
According to ILO (2020) June estimates, countries that have included some sort of workplace
closures within the COVID-19 containment measures include around 93 per cent of the global
workforce.

In India, the number of confirmed COVID-19 infections reached 1.9 million as of 5th August
2020 and is increasing rapidly (GoI, 2020). To contain the spread of the virus, the Government
of India issued the first notification on 24th March 2020 which declared a nationwide lockdown
for 21 days till 14th April 2020 [referred to as Lockdown 1.0] (MHA, 2020a). As the number
of COVID-19 confirmed cases kept on increasing, the Government of India came up with
another notification on 15th April to extend the lockdown by 19 more days, from 15th April till
3rd May [referred to as Lockdown 2.0] (MHA, 2020b). Compared to other countries, India had
put in place one of the strictest containment and closure policies in the world, according to the
COVID-19: Government Response Stringency Index developed by the researchers at the
University of Oxford (Hale et al., 2020). 1 While the containment measures came at an
appropriate time and seemed to countervail the spread of the disease, the Indian economy
experienced a severe labour supply shock. Figure 1 aims to provide an overview of the broad
literature on how the containment measures can trigger the supply and demand side of an
economy.2 The supply side gets affected primarily due to the containment measures imposed
by the government as it restricts the movement of its citizens (Rio-Chanona et al. (2020), Hicks
et al. (2020), Dingel & Neiman (2020), McKibbin & Fernando (2020)). According to CMIE,
122 million people went out of the job in April 2020 (Vyas, 2020). Moreover, according to
ILO (2020) in April, 400 million informal workers are at risk of losing their jobs during the

2
crisis. While ILO (2020) estimates do not account for essential industries, the CMIE estimates
are private survey-based and their analysis provided in the public domain is not comprehensive.

The present paper contributes by designing a methodology, a bottom to up approach, to arrive

at labour supply shock estimates due to containment measures taken by the Government of
India to tackle the spread of the virus in Lockdown 1.0 and Lockdown 2.0 using a secondary
database available in the public domain.3 We use the existing PLFS (2017-18) database for the
analysis. Further, we use a novel approach to indigenise the Remote Labour Index (RLI) which
is a crucial component to estimate the labour supply shock. To get an extensive impact of the
COVID-19 pandemic on the labour market we carry out an in-depth analysis of labour supply
shocks by employment status, industry level and occupation.

During a pandemic, the demand side is also affected due to the change in the demand patterns
of consumers as they avoid places and activities which have a high risk of exposure and demand
more health care services (see Baker et al. (2020a), Baker et al. (2020b), CBO (2006),
Muelbauer (2020), OECD (2020)). But the first-round impact of the pandemic on the economy
largely comes from the first order supply shocks through labour supply reductions, as some of
the workers in the non-essential industries are unable to perform their work activities from
home (Rio-Chanona et al., 2020). We find in our analysis that the labour supply reductions due
to lockdown measures led to substantial wage/income and gross value added losses to the
Indian economy as well. Thus the macro-level impact of supply shocks the economy is
primarily Keynesian in nature due to strong aggregate demand-side effects. Guerrieri et al.
(2020) presents the theory of such Keynesian supply shocks and argue that the initial supply
shock could lead to a more than proportional second order fall in aggregate demand during the
current pandemic.

It is important to note that the present paper does an agnostic analysis and provides a
methodology to quantify the estimates on labour supply reductions due to lockdown measures.
Although, for the wage loss and Gross value added loss calculations we use assumptions of
neo-classical theory of labour and output, the macroeconomic level impact on the economy is
primarily Keynesian.

The paper proceeds as follows. Section 2 describes the data and methodology used for
estimating the labour supply shock estimates, with subsections 2.1.1 and 2.1.2 providing details
on industry classification and construction of Remote Labour Index (RLI) for India. Section
3.1 provides the aggregate as well as disaggregate (by employment status and informality wise)

3
estimates of the labour supply shock. Employment status wise wage and income loss
estimations are covered in Section 3.2. Section 3.3 shows the impact of the pandemic from a
macroeconomic perspective by providing an industry-wise estimate of the labour supply shock
and gross value added (GVA) loss. Section 4 briefly covers the limitations of the study and we
end this paper with a discussion as covered in Section 5.

Figure 1: Supply and demand side effects of COVID-19 containment measures

Source: Compiled by authors

2 Data and Methodology

2.1 Estimation of the labour supply shock
To carry out estimations for labour supply shock (both aggregated and disaggregated) we have
used Periodic Labour Force Survey-PLFS (2017-18). The estimations are carried out
considering both usual principal and subsidiary employment status of the individuals. 4 To
obtain absolute numbers, all the estimates are adjusted to the projected population numbers
(Mehrotra and Parida, 2019). We obtain absolute numbers by multiplying the PLFS estimates
with the Census Adjustment Multiplier (CAM). CAM is the ratio of the Census projected
population and PLFS estimated population. As the first lockdown was announced by the GoI
on 24th March 2020, we use Census projected population for March 2020 to estimate CAM

4
(Census, 2011). We assume that there is no structural change in employment framework
between 2017-18 and March 2020.

The labour supply shock captures those workers who are employed in the non-essential
industries and are unable to perform their work activities from home (Rio-Chanona et al.,
2020). These are same as the workers at risk of job loss.5 To estimate the labour supply shock
we consider two metrics. The first metric is defined at an industry level where industries are
categorised as ‘essential’ and ‘non-essential’. The second metric captures the potential of work
activity to be carried out from home. The workers impacted are those who work in a non-
essential industry and are not able to work from home. The methodology used to generate these
two metrics is covered in detail in Section 2.1.1 and 2.1.2, respectively.

2.1.1 Essential and non-essential industry classification

To categorise industries as essential or non-essential, we have used the government

notifications issued during Lockdown 1.0 and Lockdown 2.0 (MHA, 2020a and MHA, 2020b).
An industry is categorised as essential if it was allowed by the government to operate during
the lockdown. It is assumed that if the industry comes under the essential category than the
workers working in that industry would not be impacted and thus no resultant job loss would
happen in those industries.

In comparison to Lockdown 1.0, some norm relaxations were introduced in Lockdown 2.0
specifically in the rural area outside the limits of municipal corporations and municipalities
(MHA, 2020b). The essential and non-essential classification is done at the National Industrial
Classification (NIC) 5-digit level separately for both rural and urban areas. PLFS (2017-18)
uses 5-digit classification by NIC (2008) to record the industry of work for the workers
(employed persons). There are a total of 1223 5-digit NIC 2008 industries in which workers
are employed according to PLFS 2017-18. Table 1 shows the number and the proportion of
essential and non-essential 5-digit NIC-2008 industries. As the first lockdown notification did
not distinguish between the rural and urban areas, 37 per cent of 5-digit NIC industries are
identified as essential in both the rural and urban areas. As a result of easing some relaxations
during the Lockdown 2.0, the share of 5-digit NIC industries rose to 77 per cent in rural areas
and 48 per cent in urban areas.

5
Table 1: Number and proportion of essential and non-essential 5-digit NIC-2008
industries
Rural Urban
Essential Non-essential Essential Non-essential
Number % Number % Number % Number %
Lockdown 1.0 453 37 770 63 455 37 770 63
Lockdown 2.0 937 77 286 23 593 48 630 52
Total number of 5-digit NIC industries = 1223
Source: Authors estimation

Further, we find that in comparison to the rural area, a higher proportion of workers were
engaged in non-essential activities in the urban area during both Lockdown 1.0 and Lockdown
2.0 (Figure 2). For instance, in Lockdown 1.0, 62 per cent of workers in urban areas were
engaged in non-essential activity in comparison to 22 per cent in rural areas. The same trend
continues for Lockdown 2.0. As most of the farming activities are concentrated in rural areas
and are classified as essential, we find a lower proportion of workers employed in non-essential
activities in rural areas.

Figure 2: Proportion of workers engaged in essential and non-essential industries

100
12.42
90 22.18
25.21
80 35.43
50.60
percentage of workers

70 61.75
60
50
87.58
40 77.82
74.79
30 64.57
49.40
20 38.25
10
0
Lockdown 1.0 Lockdown 2.0 Lockdown 1.0 Lockdown 2.0 Lockdown 1.0 Lockdown 2.0
All India Rural Urban
essential non-essential

Source: Authors estimation

2.1.2 Remote Labour Index (RLI)

Studies have used different approaches to assess different occupations and activities performed
by workers and to identify the ones that can be carried out from home. Zhang et al. (2020),

6
surveyed 369 adults in 64 cities in China after one month of confinement due to lockdown and
found that 27 per cent of the labour force continued working, and 38 per cent worked from
home. Using O*NET surveys, Dingel and Neiman (2020), under the impact of social distancing
measures, estimate that in the US 37 per cent of the share of jobs can be performed entirely at
home. The authors also apply their occupational methodology to 85 other countries revealing
that lower-income economies have a reduced share of occupations that can be performed at
home. The international comparison does not include India, but it gives outcomes to Sri Lanka
(20.7 per cent), Afghanistan (10.9 per cent), Maldives (36.6 per cent), Nepal (16.8 per cent)
and Pakistan (13.5 per cent). Similarly, Saltiel (2020) uses STEP survey and finds that few
jobs can be done at home, ranging from 5 to 23 per cent across the ten developing economies,
and reports a positive correlation between this share and GDP per capita. In India, Chatterjee
et al. (2020) use O*NET survey data and estimate that about 15.9 per cent of workers across
various occupation types, can work remotely in India. Since O*NET surveys are designed for
work context of jobs in the US it may not appropriately capture the potential for work from
home for Indian jobs well. For instance, using O*NET the percentage of workers in Legal
occupations that who can work from home in India is about 97 per cent (Chatterjee et al., 2020).
Our methodology shows that only 56.25 per cent of Legal professionals can work from home,
which appears more realistic.

In this paper, we create our own RLI for India. RLI measures the degree of work that can be
performed from home. For each activity within an occupation, 𝑅𝐿𝐼 ∈ [0,1], where 𝑅𝐿𝐼 = 0
implies no work activity within an occupation can be performed from home and 𝑅𝐿𝐼 = 1
implies all work activities within an occupation can be performed from home. We use the
National Classification of Occupations, NCO (2015) to map the work activities to the
occupation. 6 As PLFS (2017-18) is based on the NCO (2004) framework, we use the
concordance table given in NCO-2015 to match NCO-2004 occupations. Moreover, as PLFS
collects data at NCO 3-digit level, the mapping of work activities and occupation is done at
that level.7 The mapping from work activity to occupation and industry is summarised in Figure
3.

7
Figure 3: A schematic diagram showing mapping from work activities to industries

To get 𝑅𝐿𝐼𝑖𝑘 , four authors provide their independent and subjective rating, 𝐿𝑖𝑘 , for each work
activity, 𝑖𝑘 , within an occupation, 𝑘, to assess whether a work activity can be performed from
home or not.8 Two scales were used to rate a work activity: binary and a Likert scale of 3. In
the case of binary rating, 0 implies work cannot be performed from home and 1 implies it can
be performed from home. 9 For Likert scale ‘0’ implies work activity is not likely to be
performed from home, ‘1’ implies work activity is somewhat likely to be performed from home,
‘2’ implies work activity is very likely to be performed from home. As the Likert scale would
allow for more variation and it captures the close calls while rating work activities, we use RLI
generated using a Likert scale to estimate the labour supply shock in this paper.10

This method of constructing RLI is similar to Rio-Chanona et al. (2020), although we differ in
two aspects: Firstly, we use Likert scale of three instead of a binary scale to construct RLI, to
account for a profound heterogeneity present in India. For instance, certain work activities like
teaching can possibly be performed from home in urban cities but not in rural areas due to
varying conditions for remote working. Secondly, we catalogue work activities for each
occupation and rate each work activity within an occupation. Rio-Chanona et al. (2020) do not
catalogue work activities for each occupation, rather they consider all possible work activities
and later map them to occupations which can result in bias. For instance, if we consider a work

8
activity like conducting research, it is possible for a mathematician or an economist to work
remotely but may not be possible for a chemist or a biologist as they need access to the
laboratory. Such cases are not captured aptly in Rio-Chanona et al. (2020).

RLI for kth occupation, 𝑅𝐿𝐼𝑘 , using 3 point Likert scale is defined as follows,

4 𝑛𝑘
1 𝐿𝑖𝑘,𝑗 Eq. 1
𝑅𝐿𝐼𝑘 = ∑∑( )
4𝑛𝑘 max {𝐿̅𝑖𝑘 } − min {𝐿̅𝑖𝑘 }
𝑗=1 𝑖𝑘 =1

where, 𝑛𝑘 is the total number of work activities for kth occupation, 𝑖𝑘 is the ith work activity
within kth occupation, 𝐿𝑖𝑘,𝑗 is the rating given by jth rater for ith work activity for kth occupation
and 𝐿̅𝑖𝑘 is the set of possible ratings for ith work activity within kth occupation.11

Figure 4: Average RLI scores at NCO 1- digit level division

0.9
0.8
0.8
0.7 0.7
0.7 0.6
0.6
RLI SCORE

0.6
0.5 0.5
0.4
0.4
0.4
0.3
0.2 0.1 0.1
0.1 0.1
0.1 0.0 0.0 0.0 0.0 0.0
0.0
0
CLERKS
PROFESSIONALS

SKILLED AGRICULTURAL

ELEMENTARY
CRAFT WORKERS
LEGISLATORS/MANAGERS

OCCUPATIONS
TECHNICIANS

PLANT OPERATORS
SHOP, SERVICE &
MARKET

Mean RLI score (Likert 3 point rating) Mean RLI score (Binary rating)

Source: Authors estimation

Our estimates show that about 29.4 per cent and 11.6 per cent of jobs can be done from home
in urban and rural India, respectively, with a country average of 17.5 per cent. 12 We find that
occupations such as Legislators, Senior Officials and Managers, professionals, technicians and
clerks all have an average RLI greater than or equal to 0.5 (Figure 4). On the other hand, low
skilled workers such as service workers, machine operators and elementary occupations have

9
an average RLI score of 0.1, which implies that they have extremely limited means, scope and
ways to carry out work activities remotely from home and are at a greater risk of losing their
jobs (Figure 4).

2.1.3 Measurement of the labour supply shock

The labour supply shock for an occupation 𝑘 within an industry 𝑞, ∆𝐿𝑞,𝑘 , is calculated as
follows,

∑(1 − 𝑅𝐿𝐼𝑘 ) × 𝑍𝑝𝑞,𝑘 ∶ worker 𝑝 working in the non − essential industry

∆𝐿𝑞,𝑘 = {𝑝𝑞,𝑘 Eq. 2
0 ∶ worker 𝑝 working in an essential industry

For a worker 𝑝 working in an essential industry, labour shock would be 0, thus labour shock
for all occupations 𝑘 within these industries would be zero. On the other hand, if a worker 𝑝
with occupation 𝑘 works in a non-essential industry than the labour loss would depend on the
proportion of work that cannot be done from home i.e. (1 − 𝑅𝐿𝐼𝑘 ). If an occupation can be
perfectly done from home than 𝑅𝐿𝐼𝑘 would be 1 and labour shock would thus be 0. To get
labour shock for an industry 𝑞 and a particular occupation 𝑘, ∆𝐿𝑞,𝑘 , we need to aggregate
labour loss for all workers 𝑝𝑘,𝑞 within the industry adjusted by the weighted representation of
a worker in the population 𝑍𝑝𝑘,𝑞 . Aggregating ∆𝐿𝑞,𝑘 over all the occupations 𝑘 would give us

industry-specific labour supply shock, ∆𝐿𝑞 . Aggregating ∆𝐿𝑞 over all industries 𝑞 would give
us economy-wide aggregate labour shock,

𝑛
Eq. 3
𝐴𝑔𝑔𝑟𝑒𝑔𝑎𝑡𝑒 𝑙𝑎𝑏𝑜𝑟 𝑠𝑢𝑝𝑝𝑙𝑦 𝑠ℎ𝑜𝑐𝑘 = ∑ ∆𝐿𝑞
𝑞=1
𝑚𝑞
where industry-specific labour supply shock ∆𝐿𝑞 = ∑𝑘=1 ∆𝐿𝑞,𝑘 , 𝑚𝑞 is the total number of
occupations within an industry 𝑞 and 𝑛 is the total number of industries within an economy.

3 Results
3.1 Labour supply shock estimations

In March 2020, about 465.3 million people were employed in the Indian labour market out of
which 116.18 million (25 per cent) and 78.93 million (17 per cent) workers were affected in
Lockdown 1.0 and Lockdown 2.0, respectively, and were at risk of job loss (Table 2).13

10
Table 2: Region wise number of workers at risk of job loss
Lockdown 1.0 (millions) Lockdown 2.0 (millions)
Region Total number of workers (millions)
(% of total) (% of total)
65.50 53.21
Urban 155.90
(42.01) (34.13)
50.69 25.72
Rural 309.41
(16.38) (8.31)
116.18 78.93
Total 465.31
(24.97) (16.96)
Source: Authors estimation using PLFS (2017-18)

During Lockdown 1.0, around 42 per cent of workers in the urban areas and 16 per cent of
workers in the rural areas were impacted (Figure 5). There are 20 per cent and 6 per cent
workers in the urban and rural areas, respectively, which belong to the non-essential industry
and can work remotely. Moreover, among the workers who cannot work remotely 29 per cent
and 72 per cent in the urban and rural areas, respectively, were employed in essential industries
and thus were not affected (Figure 5). Similarly, one can read Figure 6 to analyse the impact
of Lockdown 2.0.
Figure 5: Percentage of workers impacted during Lockdown 1.0

Figure 6: Percentage of workers impacted during Lockdown 2.0

11
Disaggregating the impact based on employment status shows that the workers who were
significantly impacted due to the lockdown measures are regular and salaried wage earners,
followed by casual and own-account workers (Table 3). In Lockdown 1.0, of the total workers
in each employment status category, 44 per cent regular and salaried employees, 20 per cent of
own-account workers and 22 per cent of casual workers were found at risk of job loss.

It is concerning that out of 116.18 million workers who were found at risk of job loss during
Lockdown 1.0 around 60 million (more than 50 per cent) are casual and own-account workers.
The casual labour market mainly consists of people from economically poorer households,
engaged in irregular work, compensated on a daily basis and with low levels of education and
skills (Rani and Belser, 2012). The incidence of such labour is high among socially
disadvantaged groups. 14 Casual workers are usually under-employed as working cycles are
irregular and they are compensated only for days worked (ILO, 2018). These characteristics of
workers engaged in casual labour put them in the most vulnerable category amongst the
compensated workers. With a little or no job and income security, they are likely to suffer the
most upon any economic shock such as the one we are facing. Moreover, most of the own-
account workers are also home-based workers and home workers and their income fluctuates
the same as casual workers. The biggest challenge is that they are undercounted and
unrecognised (ILO, 2018).

Table 3: Employment status wise number and proportion of workers at risk of job loss

Job loss (in millions)

Total workers
Employment status (percentage loss)
(in millions)
Lockdown 1 Lockdown 2
34.24 24.61
own-account worker 168.43
(20.33) (14.61)
1.44 0.87
employer 9.53
(15.13) (9.16)
5.64 3.71
contributing family members 60.53
(9.31) (6.12)
49.78 36.97
regular and salaried employee 113.96
(43.68) (32.44)
25.08 12.78
casual worker 112.85
(22.22) (11.32)
116.18 78.93
Total 465.31
(24.97) (16.96)
Source: Authors estimation using PLFS (2017-18)

12
Another concerning matter is that the Indian economy is characterised by widespread
informality in economic activities and the labour market.15 About 80.7 per cent of workers in
all economic activities in India are found in the unorganised sector with only 0.5 per cent under
formal employment contract within them. Even in the organised sector, the share of workers
with informal employment is about 53 per cent. Overall, India has about 91 per cent of informal
workers in the labour market. 16 Workers with informal employment contract have no job
security, income security, health or pension benefits and they are likely to be most affected by
aggregate economic shocks (see ILO (2020), Dev and Sengupta (2020), Mehrotra and Parida
(2019)).

Table 4: Employment status and informal category wise proportion of workers at risk of
job loss
Lockdown 1.0 Job loss (millions)
(percentage loss)
Employment status
FOS IOS FUS IUS FEH IEH Total
0.33 33.91 34.24
own-account worker
(0.96) (99.04) (100.00)

0.07 1.37 1.44

Employer
(4.99) (95.01) (100.00)

0.12 5.52 5.64

contributing family member
(2.09) (97.91) (100.00)

10.71 13.24 0.78 20.59 0.10 4.35 49.78

regular and salaried employee
(21.52) (26.59) (1.56) (41.37) (0.21) (8.75) (100.00)

0.13 6.47 0.03 17.71 0.00 0.74 25.08

casual worker
(0.51) (25.79) (0.12) (70.61) (0.00) (2.97) (100.00)

11.24 19.82 0.81 79.10 0.10 5.10 116.18

Total
(9.68) (17.06) (0.70) (68.08) (0.09) (4.39) (100.00)

Lockdown 2.0 Job loss (millions)

(percentage loss)
Employment status
FOS IOS FUS IUS FEH IEH Total
0.26 24.35 24.61
own-account worker
(1.07) (98.93) (100.00)

0.04 0.83 0.87

Employer
(5.05) (94.95) (100.00)

0.09 3.62 3.71

contributing family member
(2.41) (97.59) (100.00)

8.37 9.96 0.57 13.61 0.10 4.35 36.97

regular and salaried employee
(22.64) (26.94) (1.53) (36.83) (0.28) (11.78) (100.00)

0.08 2.20 0.03 9.72 0.00 0.74 12.78

casual worker
(0.62) (17.22) (0.24) (76.10) (0.00) (5.82) (100.00)

8.76 12.25 0.60 52.13 0.10 5.10 78.93

Total
(11.09) (15.52) (0.76) (66.04) (0.13) (6.46) (100.00)
Source: Authors estimation using PLFS (2017-18)

13
We have provided job loss estimates using the dualistic framework of informality where we
account for the possibility and intersection of formal/informal employment and
organised/unorganised sector. According to the dualistic framework of informality, there is a
possibility of a worker employed in the organised sector and informally employed and vice
versa. For this paper, we have come up with the following six groupings for informality:

1. FOS = Formal employment in organised sector,

2. IOS = Informal employment in organised sector,
3. FUS = Formal employment in organised sector,
4. IUS = Informal employment in unorganised sector,
5. FEH = Formal Employer’s household,
6. IEH = Informal Employer’s household

Among the six grouping IOS, IUS and IEH together constitute informal employment and FUS
and IUS together constitute employment in the unorganised sector. Workers employed in the
IUS category form the most vulnerable group. We find that, among the workers who are at risk
of losing jobs during Lockdown 1.0, 68 per cent (approx.) belongs to the IUS category. Among
all the workers who were at risk of losing jobs during Lockdown 1.0, we find that 89.5 per cent
(approx.) workers are informally employed and 72 per cent belong to the unorganised sector
(Table 4). In absolute numbers, out of 116 million workers who were at risk of losing jobs, 104
million workers were informally employed. Around 79 million workers belonged to vulnerable
IUS category and thus had a higher risk of losing their jobs in Lockdown 1.0 (Table 4).
Interestingly, among the total regular and salaried employees, who were at risk of losing a job,
77 per cent (approx.) were informally employed (Table 4). This implies that these regular and
salaried workers receive no social security benefits, and are more vulnerable than those who
are formally employed.

3.2 Wage and Income loss estimations

After estimating job loss for different categories of workers we proceed to measure how much
workers have lost in terms of wages and income. According to marginal productivity theory of
wages, under perfect competition wages of workers are value of the marginal product of labour.
If a worker is unable to supply his/ her labour then the value of the marginal product would be
zero and so would the wages be. In the current COVID-19 pandemic scenario a worker who is
unable to work from home (due to low RLI), would not add any value to the output and can be
denied wages or given wage cuts depending on the employment contract. Moreover, if the
14
supply shock is not transitory and persists for a while they could lose their job as well. The halt
in major part of the economy, due to lockdown measures, thus put workers at risk of job loss
and a possible wage and income too. In this anticipation, the Government of India on March
29, 2020 ordered the mandatory payment of wages to alleviate the suffering of those impacted
(MHA, 2020a.7). But later, the Supreme Court overruled this order and said that the employers
who are unable to pay full wages will not be prosecuted (Rautray, 2020). Given this backdrop,
while estimating wage loss, we thus assume that employment of informal nature would have
disruptions in regular payment of wages. To identify workers (wage earners) who will have a
higher likelihood of wage disruptions we follow the paid leave criteria, so that we do not
overestimate the wage loss. Paid leave is one of the benefits, along with social security benefits
which a worker with the formal employment contract gets. If a worker (wage earner) is not
eligible for a paid leave then we assume that he/she is informally employed and do not receive
wages due to labour supply disruptions. Therefore, the workers (wage earners) who were at
risk of job loss and are eligible for paid leave have been assumed to receive wages and thus not
considered for wage loss estimations.17 However, for the estimation of income loss of own-
account workers, we have not used paid leave eligibility. 18 We calculate the expected
wage/income loss estimation for Lockdown 1.0 and Lockdown 2.0 at 2017-18 prices.

Table 5: Wage and Income Loss (2017-18 prices) for Lockdown 1.0 and Lockdown 2.0
(40 days)

Regular and salaried

Casual worker Own account workers
employee
Wage/ Total
Wage loss Wage loss Income loss
Income Avera Avera (Rs
quartiles Rs ge Rs ge Rs billions)
Averag
per cent (billions RLI per cent (billions RLI per cent (billions
e RLI
) ) )
Quartile 1 0.26 40.6 81.8 0.17 16.6 22.7 0.23 15.3 31.1 135.7
Quartile 2 0.27 36.2 111.3 0.16 18.9 33.1 0.23 15.8 63.6 208.1
Quartile 3 0.37 20.2 110.2 0.16 22.7 48.5 0.26 19.1 98.0 256.7
Quartile 4 0.54 2.7 29.9 0.15 27.8 79.8 0.36 17.7 154.3 264.0
Total 0.35 15.3 333.3 0.16 22.6 184.1 0.27 17.4 347.0 864.5
Source: Authors estimation

According to our estimations, total wage/income loss due to lockdown measures is Rs. 864.5
billion (Table 5). In proportional terms, this is 15.3 per cent, 22.6 per cent and 17.4 per cent
reductions in the total wage and income earnings for regular and salaried wage employee,
casual and own-account workers respectively. Interestingly, for regular and salaried employees
the proportion of workers who are at risk of losing jobs, as well as the proportion of wage loss,

15
decreases with increasing wage quartiles. In other words, those with lower wages (lower
quartiles) have a higher risk of job loss and more wage loss due to lower RLI, which makes
them most vulnerable. This result is consistent with the findings of Adams-Prassl et al. (2020)
who surveyed both UK and US citizens in March 2020 to identify that higher wage occupation
workers can work more from home. However, we find no such evidence for casual and own-
account workers. This shows the vulnerability of casual and own-account workers across wage
quartiles to deal with the uncertain economic situation caused due to COVID-19 pandemic.
To understand the severity of wage/income loss due to labour supply shock it suffices to
mention that the wage/income loss for Lockdown 1.0 and 2.0 (40 days) of Rs. 864.5 billion is
1.4 times the annual union budget allotted for employment guarantee scheme MGNERGA in
2020-2021 (Union Budget, 2020). This large scale wage/income loss can dampen the effective
demand in the Indian economy. From a macroeconomic perspective, the first order supply
shock could fuel into second order demand shocks as millions might lose their wages and
income. Perhaps, efficient provisioning of social security net comprises of the minimum wage
guarantee and/or an employment guarantee scheme is important in this context.

3.3 Labour supply shock by industry and GVA losses

In this section, we aim to understand the impact of lockdown measures from a macroeconomic
perspective. To do this, firstly we look at the industrial group wise labour supply shock and
then we proceed to calculate relevant gross value added (GVA) loss. As elaborated earlier,
labour supply shocks are driven by the proportion of workers employed in the non-essential
activity and the extent to which they cannot work from home. Table 6 provides an industrial
group wise proportion of workers who can work remotely and are engaged in non-essential
activities. In Lockdown 1.0, top 3 industrial groups with the highest proportion of non-essential
activities are Manufacturing, Trade, hotels, transport communication & services related to
broadcasting and Public administration, defence and other services. Among them,
Manufacturing has the lowest RLI.

The trend provided above follows the job loss figures provided in Table 7. In Lockdown 1.0,
the most affected sectors were Manufacturing in which 64 per cent of the workers were at risk
of job loss. The other most affected sectors were Trade, hotels, transport communication &
services related to broadcasting (45.38 per cent of workers are affected) and Public
administration, defence and other services (36.35 per cent of the workers are affected).

16
Table 6: Industrial group wise proportion of workers in non-essential activities and can
work remotely
per cent of workers employed in per cent
non-essential activity who can
Industry work
Lockdown 1.0 Lockdown 2.0 remotely

Agriculture, forestry & fishing 0.04 0.02 5.71

Mining & quarrying 22.11 5.13 10.69

Manufacturing 83.40 57.22 23.30

Electricity, gas, water supply & other utility services 0.00 0.00 21.78

Construction 31.52 12.34 4.30

Trade, hotels, transport communication & services 63.17 39.12 28.28

related to broadcasting
Financial, real estate & professional services 46.62 46.36 47.90

Public administration, defence and other services 70.04 70.04 42.19

Total 35.43 25.21 17.54

Source: Authors estimation

Moreover, in Lockdown 2.0, the same industrial groups remain the most affected. However, a
disaggregated analysis shows that proportions of workers at risk of job loss reduced
significantly in Construction and wholesale trade. This is because many activities belonging
to these sectors were allowed to operate, especially in rural areas, during Lockdown 2.0.
Moreover, the sub-industrial groups which were significantly affected during Lockdown 1.0
and 2.0 are accommodation, transport, real estate, arts and entertainment and other services.19

It is pertinent to note that in both the lockdowns one of the least affected sector was Agriculture,
forestry & fishing. Although we do acknowledge that there have been supply chain disruptions
in the agriculture sector during initial periods of lockdown, as emphasised in Narayan and Saha
(2020). The present study does not account for the supply chain disruptions as we only
considered the first order supply shocks to the economy.

17
Table 7: Industrial group wise proportion of workers at risk of job loss
Total
Workers at risk of job loss as per cent
number
of total workers
Industry of
workers
Lockdown 1.0 Lockdown 2.0
(millions)
Agriculture, forestry & fishing
193.29 0.04 0.01
Mining & quarrying
1.93 20.51 4.43
Manufacturing
59.81 64.30 43.19
Electricity, gas, water supply & other utility services
2.93 0.00 0.00
Construction
53.89 29.28 11.36
Trade, hotels, transport communication & services related to
broadcasting 87.98 45.38 28.79

wholesale trade
49.61 38.13 9.51
transportation
23.92 57.22 55.62
accommodation
9.20 73.36 73.36
information
5.24 10.89 10.61
Financial, real estate & professional services
16.76 22.96 22.87
financial activity
5.39 0.26 0.00
real estate
1.10 33.30 33.30
professional activity
4.30 26.85 26.85
administrative
5.97 38.71 38.71
Public administration, defence and other services
48.71 36.35 36.35
public admin
7.98 0.00 0.00
education
18.32 25.52 25.52
human health
6.05 0.00 0.00
arts/entertainment
1.41 57.49 57.49
other services
9.42 74.45 74.45
act of hhs
5.51 94.36 94.36
act of extra
0.02 28.20 0.00
Source: Authors estimation

To comprehend the industry-wise labour supply shock in value-added terms we provide Gross
Value Added (GVA) loss (2011-12 prices) for the first two lockdowns. The GVA loss is
estimated using the following equation,20

18
 𝑌𝑞,𝑡 Eq. 4
∆𝑌𝑞,𝑡 = 𝛼𝑞 × × ∆𝐿𝑞,𝑡
𝐿𝑞,𝑡
𝑌𝑞,𝑡
where 𝛼𝑞 × 𝐿 is the marginal product of labour, 0 < 𝛼𝑞 < 1 is the industry-specific labour
𝑞,𝑡

income share in output and 𝐿𝑞,𝑡 is labour input to production for industry 𝑞.21 To get industry
wise share of labour income 𝛼𝑞 , in Eq. 4, we use KLEMS (2019) database for Indian economy.
The labour income share in GVA provided for 2016-17 is used here. ∆𝐿𝑖,1 and ∆𝐿𝑖,2 refers to
the labour supply shock in Lockdown 1.0 and Lockdown 2.0, respectively. We approximate
the industry-specific output level, 𝒀𝒒 , using GVA at basic prices (2011-12 prices) and use
quarterly GVA data for eight major sectors of the economy from Q1:2011-2012 till Q4:2019-
2020 for our analysis. The data from Q1:2011-2012 to Q3:2019-2020 is sourced from EPWRF
(2020) database and for Q4:2019-2020 from the latest estimates of national accounts statistics
in MOSPI (2020). To get a no-COVID-19 scenario, we forecast the series for each sector using
ARIMA modelling. 22 Then using Eq. 4, we estimate GVA loss at an industry level 𝑞 for
Lockdown 1.0 and 2.0. The estimates of GVA loss presented here are based only on first order
supply shock. However, in the medium run, the actual losses could be higher as the output will
subsequently be affected by second order supply shocks and demand shocks. Assuming that
the effects of first order supply shock will dominate initially, we provide GVA loss estimates
for a very short period from March 25, 2020, to May 3 2020. Moreover, the second order supply
and demand linkages might be levelled off with the presence of inventories in firms and savings
with the consumers in the initial periods.

Table 8 shows the industrial group wise loss in GVA at basic prices (2011-12 prices) for the
first two lockdowns. The baseline model assumes that the labour and capital are imperfect
substitutes with output following a Cobb-Douglas production function. Due to the aggregate
labour supply shock, there has been a GVA loss (2011-12 prices) of Rs. 2069.2 billion during
first two lockdowns. That is about 14 per cent reduction in GVA (2011-12 prices) from a no
COVID-19 scenario. 23

The most impacted industrial groups are Trade, hotels, transport communication & services
related to broadcasting, Public administration, and defence and other services and
Manufacturing. These three major industrial groups account for 72.5 per cent of the total GVA
loss in the economy (2011-12 prices). Within these three major industrial groups, Trade, hotels,
transport communication & services related to broadcasting are affected the most, with GVA

19
loss percentage of 22.12 followed by Public administration, defence and other services and
Manufacturing with 19.63 per cent loss and Manufacturing with 17.58 per cent loss.

Table 8: GVA loss at basic prices (2011-12 prices) by industry

GVA at Basic prices (2011-12 prices) loss (Rs. million)
Industry Percentage loss due to lockdown from the no-COVID scenario
Lockdown 1.0 Lockdown 2.0 Total

226.16 78.68 304.83

Agriculture, forestry & fishing
(0.02) (0.01) (0.01)

13720.00 2557.54 16277.54

Mining & quarrying
(5.64) (1.22) (3.59)

283445.07 16875.58 452200.90

Manufacturing
(20.78) (13.96) (17.58)

0.00 0.00 0.00

Electricity, gas, water supply & other utility services
(0.00) (0.00) (0.00)

136793.56 47434.05 184227.61

Construction
(22.48) (8.72) (15.99)

Trade, hotels, transport communication & services 372171.38 278787.45 650958.83

related to broadcasting (23.79) (20.22) (22.12)

185892.40 181979.71 367872.11

Financial, real estate & professional services
(10.69) (10.66) (10.68)

210524.52 186845.39 397369.90

Public administration, defence and other services
(19.65) (19.61) (19.63)

1202773.09 866438.63 2069211.72

Total
(15.15) (12.03) (13.67)
Source: Authors estimation

Although, the Manufacturing has low RLI and a higher share of non-essential industry in
Lockdown 1.0 than the other two sectors, the GVA loss during lockdown is higher for Trade,
hotels, transport communication & services related to broadcasting and Public administration,
defence and other services and manufacturing. This is due to the following reasons, firstly
according to the KLEMS database the labour income share, 𝛼, in manufacturing industrial
group is 0.32, which is lower than the labour income share of 0.49 and 0.68 in Trade, hotels,
transport communication & services related to broadcasting and Public administration,
defence and other services, respectively.24 Secondly, the share of the non-essential industry for
Public administration, defence and other services stayed at 70.04 per cent in both the
Lockdown 1.0 and Lockdown 2.0, whereas for manufacturing it reduced from 83.40 per cent
in Lockdown 1.0 to 57.22 per cent in the Lockdown 2.0 (Table 8). The perceptible decline in
the share of the non-essential industry for manufacturing in Lockdown 2.0 resulted in a
relatively lower loss in GVA than Public administration, defence and other services.

20
According to CBO (2006) the severely impacted industries from the first order demand shock
in a pandemic, like 1918-1919 Spanish flu outbreak, are likely to be arts and recreation,
accommodation and transportation (rail, air, transit). Note that these industries are sub-
industries within the broad industrial groups: Trade, hotels, transport communication &
services related to broadcasting and Public administration, defence and other services. As
discussed above, these are also the most affected industries with highest GVA losses from first
order supply shocks. Therefore, it is likely that the GVA losses for these industries will sustain
as the demand for their output will remain low even if the supply recovers.

4 Limitations of the study

We believe that to get a comprehensive impact of lockdown measures supply chain linkages
need to be considered. However, the present study does not account for the supply chain
disruptions. Further, while aggregating the ratings at an occupation level we assume that all the
work activities within an occupation have an equal weightage. The estimates for GVA loss,
due to labour supply reduction, assumes that the labour productivity of a worker remains the
same while ‘working from home’ or ‘working in office’.

Additionally, for the wage loss estimations, we only consider those workers whose usual
activity status matches the current weekly status. This implies that our estimation of wage and
income loss is an underestimate of the actual wage and income loss. Finally, it is important to
note that the labour supply disruptions can have sequential second round effects on both the
demand and supply side of the economy.25 To get the comprehensive effect of a pandemic both
the first-order and second-order effects needs to be accounted for. While this paper only
focusses on the first round supply effects during the pandemic, the income/ wage losses due to
job losses can lead to severe aggregate demand reductions.26 The limitations briefly described
give us a direction and possibility for future research work.

5 Discussion

The present paper distinctly shows that the impact of COVID-19 pandemic on the Indian
economy and especially on the labour market is substantial and non-trivial. The present shock
has potentially affected the Indian economy more profoundly than the Financial Crisis of 2008-
09, thereby accentuating the already saddened state of the economy existing even before the
COVID-19 pandemic started (Kumar, 2020; Patnaik, 2020). The support extended by the
government of India, claimed to be 10 per cent of the GDP, is hardly enough to match the

21
extent of losses (Adhikari et al., 2020; Ghosh, 2020; Ray and Subramanian, 2020). This
precarious situation is compounded by the informal nature of the Indian economy in general
and labour markets in particular.

As shown in the paper, about 90 per cent and 88 per cent of all workers who suffered job loss
under Lockdown 1.0 and 2.0, respectively, are informally employed. Most of these informal
workers are unprotected, while most of the social security benefits lean towards the formal
employees. For instance, under Atal Bimit Vyakti Kalyna Yojna (ABVKY) formal workers
registered with Employee State Insurance Corporation (ESIC) can claim 50 per cent of their
salary for 3 months if they have lost their jobs anytime between 24th March and 31st December
2020. These workers are also likely to enjoy paid leave and have a higher probability of
returning to their jobs once the conditions improve. Moreover, the Atma Nirbhar Bharat
package aimed to boost the MSME sector, in practice, is mainly subsidising the private
corporate sector employing formal workers and not the home-based and micro-units which
employs a significant proportion of the informal workforce (Nagaraj and Vaibhav, 2020).
Therefore, it is urgently required that the government should redirect a comprehensive relief
package such as access to the credit and special tax exemptions towards the home-based and
micro-enterprises. Additionally, income support measures for informal workers such as social
protection transfers, unemployment benefits and temporary wage subsidies assume
importance.

These income support measures are particularly important given that the total wage and income
loss of Rs. 864.5 billion (2017-18 prices) occurred during just the first two lockdowns, as
shown in the paper, would subsequently dampen the aggregate demand. If no social protection
strengthening is introduced, households in the medium term may turn to their savings or debt
over time steepening the problem with further larger falls in aggregate demand. National Rural
Employment Guarantee Act (NREGA) is one of the schemes which has the potential to provide
social protection buffer in times of crises like the one we are facing today. Although the
government has enhanced the budgetary allocation of NREGA to 1 trillion rupees, much is still
desired. Srivastava (2020) points out that in addition to seasonal migrants, long-term circular
migrants in the urban informal economy have also been hugely impacted. However, the
NREGA scheme targets only the rural ambit and our estimates show that urban informal
workers are the most impacted. This calls for a similar urban intervention or shock absorber
as the government could act as an employer of last resort. Many distinguished economists,

22
social scientists and activists have also suggested the idea of an urban employment guarantee
law by adjusting the NREGA to the urban environment (Vij, 2020). Alternately, we could think
of a digital public works scheme for semi or skilled workers to make use of remote work
capacity aligned to the social distancing norms. Moreover, linking of the NREGA wages to the
Minimum wages act is urgently required as most of the states are observed to have NREGA
wages lower than the corresponding state’s minimum wage (Aggarwal, 2017). Credible
enforcement of minimum wages in both rural and urban areas can be another effective tool
which can feed into the aggregate demand without overburdening the government budget.

To sum up, to arrest the depressing tendencies in the economy it is imperative for the
government, in addition to providing immediate relief to the most vulnerable sections of the
society, to strengthen social security framework for the marginalized and at the same time work
towards feeding into the purchasing power capacity of lower and middle-income class to boost
aggregate demand in the economy.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship,
and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of
this article.

Notes

1 Although countries like the USA, Russia, Brazil, the United Kingdom, Spain, and Italy had
highest cases of infection, India had the highest stringency index of 100 during the entire period
of Lockdown 1.0 and 2.0 in the world.

2 We aim to provide an overview of some of the important studies to contextualize the present
study. However, the review may not be exhaustive.

3 For present analysis we do not consider labor supply reductions due to mortality and
morbidity as the percentage share of total deceased to total employed is extremely low.

4 We have considered the working age population (>15 years) for all our estimations.

5 Note that labour supply shock and workers at risk of job loss have been used interchangeably
throughout the paper.

6 Occupation-wise details on work activity are given for NCO (2015) and are unavailable for
NCO (2004).

23
7 This paper assumes that labour productivity of a worker remains the same while ‘working
from home’ or ‘working in office’.

8 Following is assumed: i) the occupation and related work activities are independent of other
occupations with similar or different work activities; ii) all the work activities within an
occupation carry an equal weightage for all occupational groups.

9 For binary scale we consider an activity can be performed from home if 3 or more raters have
agreed on it.

10 As a robustness check we find correlation between the RLI scores using binary and Likert
scale to be 87 per cent.

11 The rating is given by four raters from possible values {0,1,2}. Since the possible ratings
do not vary for raters and the possible set of ratings is uniform for all 𝑖𝑘, , the value of
𝐿𝑖𝑘,𝑗
max {𝐿̅𝑖 } − min {𝐿̅𝑖 } would be 2 for all 𝑖𝑘, . Note that the ratio
𝑘 𝑘 max {𝐿̅𝑖𝑘 }−min {𝐿̅𝑖𝑘 }
normalises rating for each rater 𝑗 for a given 𝑖𝑘, to a scale of [0,1].

12 RLI at NCO 3-digit level is available in online supplementary technical appendix D.

13 To do a robustness check for our methodology we estimated job loss numbers by applying
a common methodology on PLFS 2017-18 and PLFS 2018-19. We estimated that according to
PLFS 2018-19 job loss is 114.72 million which is 1.25 percentage lesser than the estimate of
116.8 million using PLFS 2017-18. We can see that the difference is insignificant and thus our
methodology is robust. We believe this happens because there has not been a major structural
change in employment framework between 2017-18 and 2018-19 and the economy did not face
any major economic shock during this period.

14 46.3 per cent of casual workers belong to schedule caste and schedule tribe group. 38.6 per
cent belong to the ‘other backward class’ category (PLFS, 2017-18).

15 The definition used to measure informality is provided in online supplementary technical

appendix A.

16 Authors estimation. For details refer online supplementary technical appendix A.

17 For a detailed methodology of wage loss estimation refer to online supplementary technical
appendix B.

18 Only 0.1 per cent of the own account workers are eligible for paid leave according to PLFS
2017-18.

19 It includes activities of member organizations, repair of household goods and other personal
services.

20 Refer to the online supplementary technical appendix C for details and GVA loss for the
first four lockdowns.

21 Note that αq is independent of time as it is assumed that the labour share in income does not
change over a short period of time considered for analysis here.
24
22 Fitted ARIMA models are provided in online supplementary technical appendix C.

23 The GVA loss increases to 18.68 per cent from a no COVID-19 scenario when we assume
perfect complementarity between capital and labour in the manufacturing industry. The labour
share in income in manufacturing would be 1 in this case. For details refer to online
supplementary technical appendix C.

24 Although KLEMS database provides data on labour share income for the disaggregated
manufacturing sector, we use aggregate level shares due to data unavailability of quarterly
GVA series.

25 Inoue and Todo (2020) showed that due to strong supply chain linkages, production shut
down in one part of the country can lead to more than a proportional loss in production for the
entire country. Using simulations, the authors find that a lockdown in Tokyo for a month can
result in 86 per cent reduction of production in Japan. Although, the production shut down in
Tokyo only accounts for 21 per cent of total production in Japan.

26 Guerrieri et al. (2020) show that the losses in the aggregate demand due to initial supply
shock could be larger than the shock itself.

25
References

Adams-Prassl, A., Boneva, T., Golin, M., & Rauh, C. (2020). Inequality in the impact of the
coronavirus shock: Evidence from real time surveys. IZA DP No. 13183 (April 2020).

Adhikari, A., Goregaonkar, N., Narayanan, R., Panicker, N., & Ramamoorthy, N. (2020).
Manufactured Maladies: Lives and Livelihoods of Migrant Workers During COVID-
19 Lockdown in India. The Indian Journal of Labour Economics, 1-29.

Aggarwal, A. (2017). Fairness of Minimum Wages for MGNREGA. Economic and Political
Weekly, Vol. 52, Issue No. 44, 04 Nov, 2017.

Baker, S. R., Farrokhnia, R., Meyer, S., Pagel, M., & Yannelis, C. (2020a). How does
household spending respond to an Epidemic? consumption during the 2020 Covid-19
pandemic. NBER. Working Paper 26949.

Baker, S., Bloom, N., Davis, S., & Terry, S. (2020b). COVID induced economic uncertainty.
NBER Working Paper 26983.

CBO (2006). A Potential Influenza Pandemic: Possible Macroeconomic Effects and Policy
Issues. Congressional Budget Office.

Census. (2011). Population Projections for India and States 2011-2036. Report of the technical
group on population projections, November 2019.

Chatterjee, P., Dey, S., & Jain, S. (2020). Lives and Livelihood: An Exit Strategy from
Lockdown. Economic and Political Weekly, vol lV no. 22.

CSSE (2020). COVID-19 Dashboard. Center for Systems Science and Engineering (CSSE) at
Johns Hopkins University (JHU). Retrieved from https://coronavirus.jhu.edu/map.html

Dev, S. M., & Sengupta, R. (2020). Covid-19: Impact on the Indian economy. IGIDR Working
Paper, WP-2020-013.

Dingel, J. I., & Neiman, B. (2020). Who Can Work at Home? Journal of Public Economics
189 (2020) 104235.

EPWRF (2020). EPWRF India Time Series. National Accounts Statistics of India. Retrieved
from http://www.epwrfits.in/TypesOfNAS.aspx

Ghosh, J. 2020. A Critique of the Indian Government’s Response to the COVID-19 Pandemic.
Journal of Industrial and Business Economics 47: 519–530.

GoI (2020). COVID-19 Dashboard. Government of India. https://www.mygov.in/covid-19/

Guerrieri, V., Lorenzoni, G., Straub, L., & Werning, I. (2020). Macroeconomic implications of
Covid-19: Can negative supply shocks cause demand shortages? NBER Working Paper
26918.

26
Hale, T., Petherick, A., Phillips, T., & Webster, S. (2020). Variation in government responses
to COVID-19. Blavatnik School of Government Working Paper, 31.

Hicks, M. J, Faulk, D & Devaraj, S. (2020). Occupational Exposure to Social Distancing: A

Preliminary Analysis using O*NET Data. Ball State University. mimeo.

ILO (2018). India Wage Report: Wage Policies for Decent Work and Inclusive Growth.
International Labour Organization.

ILO (2020). COVID-19 and the world of work. ILO Monitor. International Labour
Organization Third and Fifth Edition. 29 April and 20 June 2020.

Inoue, H., & Todo, Y. (2020). The propagation of the economic impact through supply chains:
The case of a mega-city lockdown to contain the spread of Covid-19. Covid
Economics, 2, 43-59.

KLEMS (2019). Measuring Productivity at the Industry Level. The India KLEMS Data Base.
https://www.rbi.org.in/Scripts/KLEMS.aspx

Koren, M., & Pető, R. (2020). Business disruptions from social distancing. arXiv:2003.13983.

Kumar, A. (2020). Macroeconomic Consequences of a Lockdown and Its Policy Implications.

Economic and Political Weekly, Vol. 55, Issue No. 39, 26 Sep, 2020.

McKibbin, W. & Fernando, R. (2020). The Global Macroeconomic Impacts of COVID-19:

Seven Scenarios. Centre for Applied Macroeconomic Analysis, ANU Working Paper
19/2020.

Mehrotra, S., & Parida, J. K. (2019). India’s employment crisis: rising education levels and
falling non-agricultural job growth. Working Paper. Azim Premji University,
Bengaluru.

MHA (2020a). Ministry of Home Affairs announcement of Nationwide lockdown dated 24th
March 2020, Order No.40-3/2020-DM-I(A). Ministry of Home Affairs, Government of
India.
- MHA. (2020a.1). Annexure dated 24.03.2020.
- MHA. (2020a.2). First Addendum dated 25.03.2020.
- MHA. (2020a.3). Second Addendum dated 27.03.2020.
- MHA. (2020a.4). Third Addendum dated 02.04.2020.
- MHA. (2020a.5). Fourth Addendum dated 03.04.2020.
- MHA. (2020a.6). Fifth Addendum dated 10.04.2020.
- MHA. (2020a.7). MHA Order restricting movement of migrants and strict
enforcement of lockdown measures - dated 29.03.2020.
MHA (2020b). Revised Consolidated Guidelines MHA order dated 15th April, 2020, No. 40-
3/2020-DM-I(A). Ministry of Home Affairs, Government of India.
- MHA. (2020b.1). First addendum dated 15th April, 2020
- MHA. (2020b.2). Second addendum order dated 15th April, 2020.

27
 - MHA. (2020b.3). Third addendum order dated 15th April, 2020
- MHA. (2020b.4). Clarification dated 15th April, 2020
- MHA. (2020b.5). Fourth addendum dated 24.04.2020
MOSPI (2020). Press Note on Provisional Estimates of Annual National Income 2019-2020
and Quarterly estimates of Gross Domestic Product for the Fourth Quarter (Q4) of
2019-2020. Dated 29.05.2020. National Statistics Office, MOSPI.

Muellbauer, J. (2020). The coronavirus pandemic and U.S. consumption. VoxEU.org, 11 April.

Nagaraj, R., & Vaibhav, V. (2020). Revising the Definition of MSMEs: Who is Likely to Benefit
From it?. The Indian Journal of Labour Economics, 1-8.

Narayan, S. & Saha, S. (2020). Urban food markets and the lockdown in India. IGIDR Working
Paper, WP-2020-017.

NCO (2004). National Classification of Occupations - 2004. Government of India, Ministry of

Labour & Employment. Retrieved from https://www.ncs.gov.in/
NCO (2015). National Classification of Occupations-2015, Vol-I and Vol-II. Government of
India, Ministry of Labour & Employment.

NIC (2008). National Industrial Classification-2008. Central Statistical Organisation, Ministry

of Statistics and Programme Implementation. Retrieved from www.mospi.nic.in

OECD (2020). Evaluating the initial impact of Covid-19 containment measures on economic
activity. 14 April 2020.

Patnaik. P. (2020). Macroeconomics of a Lockdown. Economic and Political Weekly, Vol. 55,
Issue No. 38, 26 Sep, 2020

Rani, U., & Belser, P. (2012). Low pay among wage earners and the self‐employed in India.
International Labour Review, 151(3), 221-242.

Rautray, S. (2020). SC says no action for now against employers who do not pay full wages
during lockdown. Economic Times, 16 May.

Ray, D., and S. Subramanian. 2020. India’s Lockdown: An Interim Report. NBER Working
Paper No. 27282.

Rio-Chanona, R. M., Mealy, P., Pichler, A., Lafond, F., & Farmer, J. (2020). Supply and
demand shocks in the COVID-19 pandemic: An industry and occupation
perspective. Oxford Review of Economic Policy, Volume 36, Issue Supplement_1,
2020, Pages S94–S137.

Saltiel, F. (2020). Who can work from home in developing countries? Covid Economics,
7(2020), 104-118.

Srivastava, R. (2020), “No Relief for the Nowhere People” The Hindu, May 4, 2020.

28
 Union Budget. (2020). Union Budget 2020-2021. Government of India.

Vij, S. (2020). “What an ‘urban NREGA’ should look like” ThePrint, September 14, 2020.

Vyas, M. (2020). India has a jobs bloodbath as unemployment rate shoots up to 27.1%.
Business Standard, May 4, 2020.

Zhang, X. S., Wang, Y., Rauch, A., & Feng, W. (2020). Unprecedented disruption of lives and
work: Health, distress and life satisfaction of working adults in China one month into
the COVID-19 outbreak. Psychiatry Research. 288, 112958.

29

View publication stats