Sampling and Categorization of Households for Research in Urban India
| Author | Tarun Arora,Katie Pyle |
| DOI | http://doi.org/10.1177/23210230221137633 |
| Published date | 01 December 2022 |
| Date | 01 December 2022 |
| Subject Matter | Original Articles |
Sampling and Categorization
of Households for Research
in Urban India
Tarun Arora1 and Katie Pyle2
Abstract
Conventional sampling methodologies for citizens/households in urban research in India are constrained
due to the lack of readily available, reliable sampling frames. Voter lists, for example, are riddled with
errors and, as such may not be able to provide a robust sampling frame from which a representative
sample can be drawn. The Jana–Brown Citizenship Index project consortium (Janaagraha, India; Brown
University, USA) has conceptualized a unique research design that provides an alternative way on how
to identify, categorize and sample households (and citizens within) in a city in a representative and
meaningful way. The consortium consists of the Janaagraha Centre for Citizenship and Democracy,
based in India, and the Brown Center for Contemporary South Asia, part of Brown University, USA.
The methodology was designed to enable systematic data collection from citizens and households
on aspects of citizenship, infrastructure and service delivery across different demographic sections of
society. The article describes how (a) data on communities that are in the minority, such as Muslims,
scheduled castes (SC) and scheduled tribes (ST), were used to categorize Polling Parts to allow for
stratified random sampling using these strata, (b) geospatial tools such as QGIS and Google Earth were
used to create base maps aligning to the established Polling Part unit, (c) the resulting maps were used
to create listings of buildings, (d) how housing type categorizations were created (based on the struc-
ture/construction material/amenities, etc.) and comprised part of the building listing process, and (e)
how the listings were used for sampling and to create population weights where necessary. This article
describes these methodological approaches in the context of the project while highlighting advantages
and challenges in application to urban research in India more generally.
Keywords
Sampling, sampling frame, urban, methodology, governance
Original Article
1 Christ (Deemed to be University), Bengaluru, Karnataka India
2 Janaagraha Centre for Citizenship and Democracy, Bengaluru, Karnataka, India
Corresponding author:
Tarun Arora, Christ (Deemed to be University), Bengaluru, Karnataka 560029, India.
E-mail: tarun.arora@christuniversity.in
Studies in Indian Politics
10(2) 254–274, 2022
© 2022 Lokniti, Centre for the
Study of Developing Societies
Reprints and permissions:
in.sagepub.com/journals-permissions-india
DOI: 10.1177/23210230221137633
journals.sagepub.com/home/inp
Arora and Pyle 255
Background
One of the most significant challenges that India faces in the twenty-first century is the governance of its
cities. Primarily rural thus far, India will be increasingly urban in the coming years and decades. Cities
are, moreover, centres of innovation, opportunity and growth. But their full potential can only be achieved
if they are well governed. In any democracy, and especially in one as diverse as India’s, the quality of
governance is inextricably tied to the quality of citizenship. With this problem in mind, Janaagraha Cen-
tre for Citizenship and Democracy (Janaagraha) and the Brown Center for Contemporary South Asia, at
Brown University, USA, formed a partnership in 2011 and developed a research project exploring urban
governance and citizenship. The project aims to gather systematic and robust data on the relationship
between citizenship, basic services and infrastructure delivery in 17 cities across India.
The project methodology includes focus groups, key respondent interviews, and a large and compre-
hensive household survey in each city. Approaches in the methodology of the latter are the focus of this
article. The methodological approach focuses on the systematic identification, categorization and sam-
pling of households (and citizens within) in a city in such a way that helps in achieving a representative
coverage of traditionally undercounted, informal settlements (slums) and groups which form a smaller
proportion of the total city population (e.g., scheduled castes, scheduled tribes and Muslim citizens). So
far, surveys have been administered in ten cities and analysed in eight2 while preparations are underway
for fieldwork in another seven3 cities. The sample size in each city varies from 1,000 to 4,000, depending
on the population size of the city.
Difficulties in Undertaking Urban Research
There is a growing consensus in the research community that conducting urban research in India is a
challenge for multiple reasons. First and foremost, Panman (2019) argues that it becomes particularly
challenging to research cities because there is no clear consensus on what we call ‘urban’ in the Global
South. For example, there are two well-established, different definitions of ‘urban’ in India. One is the
‘administrative definition’, and the other is the ‘Census’ definition. According to the administrative defi-
nition, the default classification of a settlement is rural, and it becomes urban only if the state govern-
ment converts it through requisite legal processes, including notification (Tandel et al., 2016). The
second pertains to the Census definition of urban. Here, ‘urban’ includes all places with a municipality,
corporation, cantonment board or notified town area committee. Additionally, the definition consists of
all other places with at least 5,000 inhabitants, at least 400 people per square km, and at least 75% of the
male main working population engaged in non-agricultural pursuits (Census of India, 2011). Thus, the
administrative definition of urban is a subset of the census definition (Tandel et al., 2016).
The non-standardization of political and administrative boundaries in Indian cities adds an extra layer
of complexity. When considering a study of governance, citizenship and service delivery, for example, it
is desirable to orient sampling frames around particular boundaries of governance and administration to
ensure coverage of citizens/households being served by the different units. However, urban local bodies,
civic agencies and others use different geographic ways to split up the same city. For example, in
Bengaluru, the urban local body (The Bruhat Bengaluru Mahanagara Palike, BBMP) divides the city
2 Bengaluru, Mumbai, Chennai, Hyderabad, Ahmedabad, Vadodara, Bhavnagar, Kochi, Mysuru (not yet analysed), Shivamogga
(not yet analysed).
3 Delhi, Kolkata, Bhubaneswar, Lucknow, Bhopal, Ajmer and Jalandhar.
To continue reading
Request your trial