RESEARCH PAPER

SURVEY ON INFORMAL WORKERS

Team name: TEAM DIAMONDS

TEAM MEMBERS: Yechuri Lakshmi Supraja, Ananya Dan, Atasi Sarkar,
Ritika Sharma
SECTION A

ABSTRACT
The measuring of deprivations in the population offers useful data for analysing inequality
and poverty, as well as for planning and implementing economic and public health policies
and interventions. This study aims to examine households' demographic, socioeconomic,
water-sanitation, and sanitary behaviour in the districts of Ranga-Reddy and Sonipat, as well
as quantify poverty using the Multi-Dimensional Poverty Index (MPI) and the relative
contributions of the indicators. A survey based on telephonic as well as field interview was
done based on a questionnaire prepared for 20 households. Around 21% of the households
are ‘deprived’ and on average the people are deprived 51.22% of the weighted indicators.
Consumption or income-based measures are insufficient to describe the multifaceted nature
of poverty and deprivations. Demographic, socioeconomic, health, and nutritional indices all
have a part in defining real living standards.
INTRODUCTION
One of the tragedies of India’s rapidly developing and increasingly progressive society is that
poverty remains prevalent and rampant, and the poor population appears to have grown to be
more vulnerable. A country with a huge population of 1.25 billion sets to become the most
populous country by the year 2022 if the population trend continues to grow at a faster rate
and is also known to be a poverty rich country. Although population growth is decreasing, the
sheer figures suggest: over 18 million people are added each year at a pace of around 1.4
percent. According to official data figures, India's severe poverty rate is roughly 25-30%. 

The counting of the poor goes back to the 19th century in India. Dadabhai Naoroji's "Poverty
and Un-British Rule in India," published in 1867-68 prices, was the first attempt to quantify
the impoverished. He came up with a subsistence cost-based poverty line, ranging from Rs.
16 to Rs. 35 per capita per year in various regions of India, based on the diet prescribed to
"supply the necessary ingredients for the emigrant coolies during their voyage living in a
state of quietude," which includes "rice or flour, dhal, mutton, vegetables, ghee, vegetable oil,
and salt." The National Planning Committee (NPC) then calculated a poverty line of Rs 15 to
Rs 20 per capita per month in 1938. The NPC, like the previous system, based its poverty
threshold on "a minimum standard of living viewpoint in which nutritional needs are
essential."

In 1971, VM Dandekar and N Rath published the first systematic evaluation of poverty in
India, based on NSS data from 1960-61. They claimed that the poverty level in both rural and
urban regions must be determined by spending adequately to provide 2250 calories a day. It
was used during 1960’s and 1970’s at the state and centre level but aroused intense debates
for its low figures. Hence, a task force was made under Dr. Y.K Algah for the estimation of
poverty line where on the advice of the Nutrition Expert Group, 'average calorie needs' were
calculated separately for rural and urban locations across India. 
As a result, the 'Poverty line basket' for urban and rural regions differed. The recommended
calorie intake was 2400 calories for rural regions and 2100 kcal per capita per day in urban
areas.

Expert group of 1993 (Lakdawala) disaggregated ‘All India poverty line’ to “State specific
poverty line’ for the base prices of 1973-74 taking into account of Consumer Price Index-
Agricultural Labour for ‘Rural state specific poverty line and ‘CPI - Industrial workers’ for
urban areas.) Expert Group 2005 (Tendulkar) adopted ‘Mixed Reference Period’ in place of
‘Uniform Reference Period’ and recommended a shift away from basing the Poverty Line
Basket toward target nutritional outcomes. In 2009, the Tendulkar Committee proposed a
new technique for revising poverty lines that accounts for changes in pricing and
consumption patterns by utilising the consumption basket of those teetering on the edge of
poverty. As a result, instead of utilising indices derived from the CPIAL for rural regions and
CPI-IW for urban areas as was done before, the estimates produced in 2009-10 and 2011-12
employ this technique. 

The lockdown exacerbated the situation, causing major sectors to fail and go down, stranding
millions of migrant workers. According to a telephone poll conducted by the Azim Premji
Foundation (2020) in partnership with the Centre of Civil Society of 4000 employees across
12 states, 80% of urban workers reported job loss, while 50% of the remaining workers
claimed income reductions or even no salary distribution. The negative impact was amplified
by the fact that the industrial sector has been dropping in growth and value since 2017–2018.
Non-farm workers (11.6 million) have lost the most employment, followed by salaried
employees (4.2 million), daily wage earners (4.2 million), and businesspeople (3 million).
However, job loss is simply one component of a larger picture of poverty and
impoverishment. Further investigation of the state of employment and subsequent re-
employment is required.

METHODOLOGY AND DATA 

A sample-based survey was conducted, on a sample of 20 households residing in different
districts of Telangana, Andhra Pradesh and Haryana. The first district is Ranga-Reddy where
10 of the informal workers work in one of the manufacturing companies owned by one of the
teammate’s fathers and the rest are from different villages of Sonipat nearby like Akbarpur,
Barota, Bandepur, Jagdishpur who worked on the main campus of O.P Jindal Global
University.

A proforma was created that included a pretested questionnaire for gathering information
from the informal workers with different employment types who were interviewed for the
data collection through either telephonic conversation or field survey. This questionnaire was
later posted on the application Open Data Kit (ODK) specifically designed for the data
collection and submission using Android mobile services. The questionnaire incorporated
information regarding demographic and socioeconomic characteristics of the family
members, availability and usage pattern of safe drinking water for drinking purpose,
affording amenities, having a bank account, sanitation practices, basic education, monthly
expenditure, impact of the pandemic on household (both waves). The group verified the data
entry, compilation, data cleaning, and data analysis and Microsoft Excel for Windows was
used to evaluate the data.

MULTIDIMENSIONAL POVERTY INDEX (MPI)

Multidimensional Poverty Index gives us a poverty statistic of a specific population using the
choices of variables, a higher value indicating a higher level of poverty. Following the Alkire
& Foster’s ‘counting technique”, this figure examines both the fraction of the population
classified as poor and the 'breadth' of poverty experienced by these 'poor' families. MPI, as
proposed by Oxford Poverty and Human Development Initiative (OPHI) is segregated into
three dimensions: health, education, standard of living and measured by ten indicators
(nutrition, availability of free grains, years of schooling, school attendance, sanitation, water,
electricity, housing, floor, amenities, cooking fuel etc.) The three dimensions were given
equal weight, and indications within each dimension were likewise given equal weights. The
cross dimensional cut-off is used once the weighted deprivations have been added up. The
cross dimensional cut-off was set at 1/3, or 33%. As a result, if a household's weighted
deprivation was more than or equal to 33 percent, it was classified as multi-dimensionally
poor. The final MPI number was calculated using the product of the incidence of
multidimensional poverty (H) and the average intensity of poverty among the poor (A).

MPI = Incidence of poverty (H) * Intensity (A) of poverty

Choice of Dimension: 
As Amartya Sen correctly points out, choosing dimensions, deprivations, and abilities for any
poverty indicator should be a value judgement rather than a scientific task. Dimensions
should be of great relevance to society and social impact, with a specific emphasis on public
policy. Health, education, living standards, empowerment, job, the environment, and safety
from violence are all conceivable aspects. We have taken the first three dimensions for this
study.

Choice Of Indicators
We have taken these indicators for calculating MPI: 2 for education, 2 for health, 7 for
standard of living.

EDUCATION

Within the education dimension, the MPI uses two indicators that complement each other:
one looks at completed years of schooling for household members, and the other looks at
whether children are attending school. Years of education serve as a substitute for household
members' knowledge and comprehension. Both years of schooling and school attendance are
imperfect proxies as they don’t inform about the quality of education, the amount of
knowledge attained or the skills gained. The MPI mandates that at least one member in the
family has finished six years of schooling and that all children of school age are enrolled in
grades 1 through 8. One thing to observe is that due to the nature of MPI,a member of
household has completed six years of schooling, he is said to be not deprived even though he
may not be educated. Similarly, someone who lives in a house where even if the child has
finished his education, he is deemed disadvantaged in this indication if he is not in school.

HEALTH

Although they are linked, the MPI utilises two health measures that differ greatly from
standard health indicators.
The first indication examines the nutritional status of family members. Indicator being if the
person has skipped meals to balance the expenditure in the duration of lockdown. If anyone
in a person's family (for whoever information is available—children, mothers, or other
adults) has skipped, the MPI classifies them as nutritionally deprived. The second indicator of
health used here is free grains provided by the government enough for the entire household. If
there has been not enough provided even for one person in the household, it is deemed to be
deprived in the MPI. 

STANDARD OF LIVING

Seven indices of living standards are considered by the MPI. It comprises three standard
MDG indicators relating to health and living conditions that are especially important to
women: access to safe drinking water (including the time for fetching it), better sanitation,
and the use of clean cooking fuel.

WATER:

If the water source is piped water, public tap, borehole or pump, protected well, protected
spring, or rainwater, and it is within 30 minutes' walking distance, the individual has access
to clean drinking water (roundtrip). If it fails to meet these requirements, the household is
regarded to be water deprived.

SANITATION (ACCESS TO TOILETS):

If a person lives in a home with a flush toilet or latrine, or a vented improved pit or
composting toilet that is not shared, they are deemed to have improved sanitation. If a home
does not meet these requirements, it is deemed sanitary deprived.

COOKING FUEL:

If the household cooks using dung, charcoal, or wood, the person is deemed to be in need of
cooking fuel.
It also contains two indicators that are not MDGs: access to power and bank account.

ELECTRICITY:

If a person does not have access to electricity, they are deemed impoverished.

BANK ACCOUNT:

The cornerstone to financial inclusion for previously unbanked families is the possession of a
bank or post office account. If a person doesn’t have an account opened on its name, they are
considered to be deprived.

The last two variables which indicate the standard of living for the household are: conditional
structure and amenities which can be afforded. 

AMENITIES:

If the person has more than 1 amenities that he/she can afford, then it means he/she is not
deprived or vice-versa.

CONDITIONAL STRUCTURE:

If the house of a household is built with bricks, it is considered to be not deprived and if it is
constructed with other materials, then it is deemed deprived.

UNITS OF ANALYSIS
 
As a unit of analysis, we used households. People in these settlements lived in families and
shared communal resources, which explains why Individuals living in these families would
be deprived in these indicators if their households were impoverished in these indicators. The
same reasoning applies to MPI.

CUT-OFFS AND WEIGHTS FOR EACH INDICATOR

The cut-offs and weights for each indicator is given below in the table - 
After plotting the table, we come to the conclusion that there are 5 households that are
deemed ‘deprived’: Household 2,4,7,8 and 9.

CALCULATION OF MPI 

The MPI combines two important pieces of data: (1) the proportion of people (within a given
population) who suffer from various deprivations (H) , and (2) the severity of their
deprivation: the average proportion of (weighted) deprivations they experience (A).

HEAD COUNT RATIO

It indicates the incidence of poverty. It helps in the determination of what percentage of

people are multidimensionally impoverished. The formula for the same is given as:- 

                                                 
Here, q is the number of persons who are multidimensionally impoverished and n is the entire
population. In our study, q = 5 + 7 + 3 + 4 + 2 = 21 and n = 11 + 5 + 8 + 7 + 5 + 6+ 3+ 4 + 2
+ 4 + 2 + 3 + 4 + 4 + 6 + 9 + 5 + 4 +4 + 4 = 100
H = q/n = 21/100 = 0.21

INTENSITY OF POVERTY

It is the average proportion of deprivations encountered by multidimensionally poor persons.

It is the sum of all multidimensionally impoverished people's deprivation scores. The second
partial index in the MPI's construction, A, addresses the question, "How poor are the poor?"
The formula for the same is given as :- 

                                      
where ci(k) is the censored deprivation score of individual i and q is the number of people
who are multidimensionally poor.
  
 
 A = (33.4 *5) + (64.2 * 7) + (50.1 * 3) + (47.5 * 4) + (59.5 * 2) = 51.22% (in percent)
                                          21

This indicates that the average poor person is deprived in 51.2% of weighted indicators.

MPI as discussed above is the product of H and A

MPI for the study = 0.21 * 0.5122 = 0.107562 or 10.75% 

In this study, 21% of the people are ‘deprived’. They lack at least a) all signs of a single
dimension or b) a mixture of indicators across dimensions, such as living in a family with a
starving individual, no clean water, a filthy floor, and unimproved sanitation. We also
observe that - on average - the poor here are deprived in 51.22% of the weighted indicators
because each deficiency is input according to its relative weight, these indicators are referred
to as "weighted" indicators.

INCOME SHORTFALL

The ratio by which the poor's mean income falls below the poverty level. The poverty line is
defined as half of the whole population's median family income. The poverty gap provides an
indication of a country's poverty level, which helps clarify the poverty rate. The formula for
the same is:

    
                             
Where gi is the distance between the poverty line (in the absence of established assistance) z
and household income y is equal to gi.

g represents an overview of these differences for the population impacted by the

circumstance under consideration.

q: The population affected by the current circumstances.

z: Average of poverty (or lack of support).

RESULTS

Ameniti es
1
Deprivation score

0
H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20
Household NUMBER

Cell phone Television Radio Bicycle Motorbike Animal cart Others

(Figure 1)
In Figure 1, we explain and observe the amenities that are at disposal or are owned by each
household. The presence or absence of the blue line indicates if a household owns a cell-
phone or not accordingly. Similarly, red line indicates television, yellow line indicates radio,
green line indicates bicycle, orange line indicates motor-bike, light blue indicates animal cart
and pink indicates any other amenity that is owned by a household accordingly. So, in this
graph, we observe that most of the households own a cell-phone. Thus, it would be safe to
say that almost all the households that we observe in this data-set consider a cell-phone to be
a necessity. Taking a further look at the graph, we observe that a “radio” is only owned by
one household, whereas a television is owned by multiple households. We also observe that
in most of the households under observation (with the exception of only 3 households), either
a “motorbike” or a “bicycle” is owned, which means that we can either see an orange line or
a green line in most households. We also observe that household 3 does not own any
amenities while household 18, not only owns most of the listed amenities, but also owns
“other” amenity, which indicates that is more well-off than all the other households that we
observe in this data-set. By standards, a household is considered “deprived” if it does not own
more than of these assets: bicycle, radio, television, telephone, computer, animal cart,
motorbike or refrigerator; and does not own a car or a truck. Keeping this in mind, we can
conclude that in terms of amenities, most of the households are “not deprived” since most of
them own more than one of the above-mentioned commodities.
 Deprivation score
Cooking fuel

H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20
Household number

Woods/Shrubs Cow dung Charcoal/Coal LPG/ Natural gas

Electricity Biogas others

(Figure 2)
Moving on, in Figure 2, we are observing the primary sources of cooking fuels in the
numerous households that we are observing.
A household is considered “deprived” if the primary source of cooking fuel is dung,
agricultural crops, shrubs, wood, charcoal or coal. Improved or safe sources of cooking fuels
include electricity, LPG/ natural gas or biogas. Just the presence of safe sources of cooking is
not sufficient for a household to warrant the status of “not deprived”. A household’s
PRIMARY i.e., most used cooking fuel should be one of the safe sources of cooking fuel. To
make things clearer, if a household has a connection of LPG, but still the most used form of
cooking fuel is, let us say, wood coal, even then, the household will be considered
“deprived”.
Coming back to the graph, we notice that most of the households have an LPG connection,
which is represented by the “yellow line” in the second graph. We also notice that one of the
other abundantly owned sources of cooking fuel in the households in question is none of the
mentioned sources, aka it falls under the “other” category. Furthermore, we also see that only
one household, that is household 3 uses woods/ shrubs as cooking fuel.
Thus, it is safe to conclude from this graph that most of the households have access to a “safe
source” of cooking fuel.
 Age Composition
80

70
70 70
60 65 65
60 60 62
55 56
50
50 50 50 50 48
40 45 45 45 46 46 45 4644 44 45
Age

40 40 40 40 42 40
35 37 35
3535 35 35 36 37 37
303234
33 33 34
32 32 33
30
26 2828 25 25 25 26
20 23 22 22 22 2221 22
19
19 20 18 18 19
17
16 16 17
16
10 12 13 12 14
13 14
12 12 13 12 13
11
10 11 11 9 119 10 108 10
75 5 6 75 8 6
0 3 3 2
H1 H1 H1 H2 H3 H3 H4 H5 H6 H7 H8 H9 H10H11H12H13H14H15 H16 H17H18H19H20
Household number

(Figure 3)
Now, if we look at Figure 3, that is the third graph, it shows the age composition of a
household. Just by looking at the graph, a simple observation can be made that a lot of the
observed households have an older household member. It is also a noteworthy observation
that during the process of data collection, it was found that most of the adults in the
households are uneducated. On further observation, we can also see that some households
live in a nuclear family, that is 2 adults along with their children (typically 2 or 3), whereas a
lot of households live in a more traditional way, where 2 or 3 generations (including the
children) live in one household, where, most of the time, the adult(s) (not the child or the
older member of the household) is responsible to shoulder the earning of the entire
household. It is not feasible to allocate a status of “deprived” or “not deprived” to a
household, based on the number of members or the age composition of a household. Thus, no
such concrete comment can be made about that by observing this particular graph.

OBSERVATIONS ON THE FIELD

As the survey was being conducted, we were observing our informants’ behaviour and
reactions to our question. Most of them displayed discomfort and reluctance when answering
sensitive questions such as their salary, cast and whether they skipped meals during the
pandemic. Almost all of them come from families in less than desirable positions. One of our
informants has a husband who turned alcoholic due to losing their job during the lockdown,
and some have family members who were the breadwinners but got badly injured and hence,
had to stop working, meaning that our informant had to take up a job to continue earning for
the family.

A lot of our informants in Sonipat were surveyed during the day meaning that we were
interrupting their them when they were in the middle of work. Since they were paid at the end
of the day, we tried our best to mindful about their time and not take too much of their time
especially when they were getting nothing in exchange for taking the time to answer our
questions. Furthermore, they were curious as to why they were being asked these questions
and what we were planning on doing with their personal information. A handful of potential
informants we approach were suspicious of our intentions and refused to take part in our
survey which is understandable.

OBSERVATIONS FROM INDICATORS USED FOR MDPI

Before discussing the indicators, let us observe the various household sizes that we have
collected data from, through the manner of survey. We see that the highest household that we
have under observation is a household containing 11 members. Second highest being 9, then,
8, 7. The lowest number of members in a household that we have is 2. 4 seems to be the most
prominent size of household throughout the survey. Moving on, let us shine light upon the
indicators that we have taken for the calculation of MDPI.
We have used three indicators to calculate the MDPI, which are:
1) Health
For health, we have taken 2 further 2 sub-indicators to find the MDPI.
 If you have skipped meals to cut expenses
In this, “0” means that the households did not need to skip and “1” indicates
that the household had to cut meals.
Under this category, we can observe that apart from household 4, none of the
other households had to skip meals to cut expenses, which implies that
household 4 is worse off as compared to other households, at-least under this
sub-indicator.
 Were the free grains enough
In this, “0” indicates that free grains were enough and “1” indicates that the
free grains were not enough for the household. Keeping this indicator in mind,
we can observe that under this category, most of the households found the free
grains to be sufficient. Among the 20 observations, we can see that there were
6 households that did not find the food grains to be sufficient, which makes
up-to 31% households finding the food-grains to be insufficient.

We divide (1/3) by “2” because of there being “2” indicators under the
category of “health”, which gives us a 16.7% each weightage for this category.

2) Education
Secondly, we have health. Again, this has 2 sub-indicators.
 If the years of schooling is less than 6 years for the adults
Under this category, we see that most of the adults have not received 6 years
of education: “0” indicates that the adults have 6 or more years of education
whereas “1” indicates that the adults in the household have received less than
6 years of education.
 For children dropped out of 8th
 “0” indicates that the children have not dropped out of school before or during
8th standard of schooling and “1” indicates that the children have dropped out
of school before or during the 8th year of schooling. There is one more
indicator – “weight added to above”, this means that there are 0 children in
that household, so this particular indicator does not apply to that household.

Therefore, both of the sub-categories in the “education” add up-to 16.7% each
of the MDPI. It is calculated in the same way as the “health” indicator, that is,
by dividing (1/3) / 2 to get the results and the weightage that this indicator
holds for the calculation of MDPI.

3) Living conditions
Lastly, we come to “living conditions”. We have 7 sub-categories in this particular
indicator, which are –
 Cooking fuel
 Access to toilets
 Electricity
 Amenities
 Conditional structure
 Bank account
 Improved quality of water

In all of the above-mentioned sub-indicators, “0” indicates that they are “not deprived” while
“1” means that they are “deprived”. Since we have 7 categories to define this indicator,
therefore, we divide (1/3) / 7 to calculate MDPI. Each category contributes 4.7%.
After that, taking into account each category and sub-category, we calculate the household
deprivation score of each household which indicates how “poor” a particular household is.
We can notice that in our entire data-set, the lowest household deprivation score is held by
household 13. Household 13 stands at a household deprivation score of 4.7%. the highest
household deprivation score of 16.2% is held by household 4.
Now, a household is considered to be “poor” if the household deprivation score is less than or
equal to 33.3%. keeping this in mind, we have deemed households to be poor or not. If we
look at the dataset, we observe that out of 20 households that were surveyed, 15 households
are not poor and 5 households are poor.
Finally, to calculate the MDPI, we multiply the headcount ratio by intensity. For our dataset,
our MDPI comes out to be approximately 10.76%

KEY OBSERVATIONS
During the survey, we noticed that if a household falls under the category of “deprived” or
“poor” while calculating the MDPI, does not necessarily mean that that household will be
considered “deprived” or “poor” according to the income of that household. For example, in
our dataset, we can observe that household 1 falls under the category of “not poor” in the
MDPI table, whereas, at the same time, it falls under the category of “poor” in the income
index of the households. From our observation, this is so because, in the MDPI table, we have
categorised the various households into the category of “poor” and “not poor” after
meticulous calculation and data observation, taking into consideration multiple factors that
influence the result of the data. Whereas, in the second case, there is only a single factor
which decides if a household falls under the category of “poor” or not, which is the monthly
income of that household.
If we were to put it in a precise way, the monthly income of a household is a subset of the
MDPI. Bringing into consideration other factors such as health and education status of a
household, it can change the status of the household. Thus, we can say that the monthly
income of a household is only a quantitative measure in itself, while the MDPI takes into
consideration the qualitative factors, basically meaning that it defines the living standard of a
household through various factors, out of JUST income.

EFFECTS OF LOCKDOWN
The first nationwide lockdown occurred in March, 2019. It was extremely difficult for people
below and on the borderline of poverty to be able to get through their monthly expenses.
Among the people that we surveyed, we noticed that a majority of the households did not
own a job card. Thus, as a result, a lot of the households had to rely on only 1 member of the
household to shoulder the entire responsibility to earn for the household, which became a
herculean task since the job opportunities, especially unskilled work was not available, as it
is, under normal circumstances. This proved to be even more difficult for some of the bigger
households, which comprised more than 5 or 6 members.
The government provided free food for people with ration cards, which made it easier. Some
of the households had to sell some assets or take up loans during the first and the second
lockdown in order to keep up with their monthly expenses. We can observe these effects in
more detail if we look at the survey that was conducted.

-----------THANK YOU---------