Gaon Connection Rural Survey Analysed by Lokniti-Csds
Gaon Connection Rural Survey Analysed by Lokniti-Csds
Gaon Connection Rural Survey Analysed by Lokniti-Csds
METHODOLOGY
Two months into the Coronavirus-induced lockdown, Gaon Connection in consultation with
the Lokniti programme of the Centre for the Study of Developing Societies conducted a sample
survey among rural households located in all zones of India to find out how lives and livelihoods
in villages had coped during the lockdown. The survey was conducted in 179 districts spread
across 20 States and 3 Union Territories. The States where the survey was conducted are
Rajasthan, Punjab, Haryana, Himachal Pradesh, Uttarakhand, Uttar Pradesh, Bihar, Jharkhand,
West Bengal, Sikkim, Assam, Arunachal Pradesh, Manipur, Tripura, Odisha, Kerala,
Maharashtra, Gujarat, Madhya Pradesh and Chhattisgarh. The surveyed Union Territories are
Jammu and Kashmir, Ladakh and the Andaman and Nicobar Islands. A total of 25,371
respondents were interviewed between May 30 and July 16, 2020 during the exercise. All
respondents interviewed were main earners of their households, and thus primarily Men.
In addition to the interview with the main earner in every household, a supplementary
interview was also conducted in those households where a household member had returned
from a city or a town during the lockdown. This interview was taken after the completion of
the main earner’s interview. A total of 963 supplementary interviews of such migrants were
completed. The aim of interviewing migrants was to find out their experience and struggle in
the city/town at the time of the lockdown and their experience/ordeal in travelling back home
during the lockdown.
The sampling method adopted was purposive, as districts chosen for the survey were those
that had a presence of the Gaon Connection team in them. Because of this limitation and given
the extraordinary situation of the epidemic/lockdown and the movement restrictions that came
along with it, the sampled districts (and the villages surveyed within them) in some States are
not necessarily spread out evenly. For instance in some states, some of the districts chosen for
the survey were geographically contiguous or located close to each other. Moreover, the
number of surveyed districts varies from State to State - some big States have relatively fewer
districts in the sample and some smaller States have relatively more districts. This also meant
that some big States with fewer districts in the sample had a more clustered sample as they saw
many more interviews being conducted within a district in order to meet the sample target.
A total of 25,000 interviews were targeted across the country at the outset of the exercise.
The sample target for the each of the States and UTs where the survey was to be conducted
was first determined on the basis of the proportion of their rural population in the overall rural
universe (the total rural population of States and UTs combined) and then adjusted/boosted in
some of the smaller States to achieve a good enough sample that would allow for robust State-
based claims. The achieved sample in some of the states fell short of the target due to Covid-
related restrictions/limitations. It must be stressed here that while making overall claims for the
country as a whole, the data reported in this report has been weighted by the State/UT
proportion – i.e., each surveyed State’s and UT’s share in the total sample is the same as the
share of their rural population in the total rural population of all the surveyed States and UTs
combined. However, while making any specific State-centric claims unweighted data has been
relied on as it ensures inferences based on robust/larger samples for the smaller States.
Union Territories
Andaman and Nicobar Islands 1 82 9
Jammu and Kashmir 10 1608 324
Ladakh 2 50 8
Total 179 25371 25371
*Based on State/UT’s rural share in total rural population of all surveyed States/UTs
Profile of sample in terms of key socio-demographics
% of Unweighted Sample % of Weighted Sample
Male 79.6 80.8
Female 20.1 18.9
Others 0.3 0.3
The fieldwork for the survey was conducted by Gaon Connection’s members in each district.
All interviews were conducted face-to-face at/outside with residence of the interviewee. During
the interview the interviewers from Gaon Connection made sure that they were wearing a
mask and maintained a distance of 1 meter between themselves and the interviewee while
asking the questions.
The survey was conducted using a structured and standardized interview schedule that was
administered via a specially designed mobile phone application. The interview schedule was
designed by researchers at CSDS for the Gaon Connection. The main interview took about
30-35 minutes to complete on average whereas the supplementary interview took about 15-20
minutes.
The survey data was analysed by researchers at Lokniti, CSDS for Gaon Connection.
Given that the sampled locations were not spread out evenly in most States/UTs and were
selected by non-probability sampling methods owing to logistical and Covid-related issues, we
are not in a position to provide reliable estimates of sampling error.
While analyzing the data 2 Classifications and 3 indices were used to make sense of the data.
Their details are given below.
Classifications
2. District Classification
• Districts where the survey took place have been classified as Green-Orange and Red
Zone districts based on the Government of India list released on May 1, 2020. Out of
the total 179 districts surveyed, 67 are Green districts (those least affected by the
Coronavirus and having no active cases at the time of ‘Lockdown 3’ on May 1), 37 Red
districts (the worst affected) and 75 are Orange districts (neither too badly affected and
nor unscathed). While their color coding may have changed when the fieldwork was
eventually conducted, a classification of sampled districts helps in an analysis of the
certain questions. District classification-wise analysis presented in the report is
unweighted.
Indices
The Economic Class Index and the resultant class categorizations have been arrived at on the
basis of whether certain select were assets owned by the respondent’s household. A scoring
system was used to construct the index. If a respondent’s household owned a car/jeep/van, an
air conditioner, a computer/laptop or an I-pad, a tractor and had a fixed line internet
connection, then they were given a score/marking of 4 for each. Households with a washing
machine and a pumping set were given a score of 3 for each item. Those with a scooter/mo-
bike, a fridge and a toilet inside the house were given 2 points for each. Finally those who had a
fan/cooler, a television set and an LPG gas connection were given 1 point for each. Those who
did not own any particular item were scored 0. The scores of all 13 asset related questions
were summed up. The resulting summated scores for each respondent ranged from 0 to 35.
The summated scores were distributed across four categories that indicated economic classes.
Respondents/households with summated scores ranging from 0 to 5 were categorized as being
‘Poor’. Those with summated scores that ranged from 6 to 9 were categorized as being from
the ‘Lower Class. Respondents with summated scores ranging from 10 to 20 were labeled as
being ‘Middle Class’. Finally respondents with scores ranging from 21 to 35 were considered as
being ‘Rich’.
The Income Change index was based on respondent answers to two questions in the survey –
one that asked them about their total monthly household income a month before the lockdown
and a question soon after that sought to find out their total monthly household income during
the lockdown months. A respondent who reported a lower (or no) income in the second
question compared to the previous question was categorized as a respondent whose
‘Household income decreased during the lockdown’. On the other hand a respondent who
reported the same income or greater income in the second question was categorized as
someone who’s ‘Household income stayed same or even increased during the lockdown’.
There were around 5,300 respondents that refused to divulge their income details in both the
questions. Such respondents were excluded from the index/analysis.
3. Index of whether money was received from the government during the
lockdown
The Index of whether money was received from the government during the lockdown has been
constructed using three questions - one that sought to know from Kisaan Samman Nidhi
scheme beneficiary households if they had continued receiving money in their bank accounts
under the scheme during the lockdown, another that sought to know whether any woman
member of the household had benefitted from the lockdown scheme of the Central
government to deposit Rs 500 in every women Jan Dhan account holder’s account for three
months, and a third that sought to know generally whether a household member had received
any amount from the government in their bank account during the lockdown. Respondents
who answered in the affirmative to any of these three questions were considered as having
‘received money in the bank account from the government’ and those who answered in the
negative to all three questions were considered as having ‘received no money from the
government’.