Election Survey Questions and Possibilities
| DOI | 10.1177/2321023019874916 |
| Author | Divya Vaid |
| Published date | 01 December 2019 |
| Date | 01 December 2019 |
| Subject Matter | Notes and Method |
Election Survey Questions
and Possibilities
Introduction
Surveys in general can be seen as the backbone of quantitative research with a wide range of users and
usage. Election surveys, in particular, serve a variety of users—from the immediate use for media houses
and the public interested in election outcomes, to the more sustained use of researchers and academics.
While election surveys in India have a longer history (Palshikar, 2013), among these surveys, the
National Election Studies (NES)1 has played a crucial role in providing data for the study of elections,
electoral processes and in the potential enterprise of theorizing the polity (for some examples, see Yadav,
2000; Yadav & Palshikar, 2008; for a critique see Nooruddin, 2016). The rapid spread and use of election
surveys in more recent times has been much discussed with an emphasis on the concomitant advantages
and limitations of this exercise.2 This research on election surveys in India and outside has highlighted
issues that are more generally relevant to opinion surveys (such as questionnaire design, or sampling
strategies), as well as issues which have a specific relevance to election surveys. In the Indian context,
the size of the geographically dispersed Indian electorate and the nature of the election process itself,
which is spread over various phases across the country, have required a specific survey and sampling
design.3
Election surveys help in understanding the immediate election (either in the period just before the
election—pre-poll) or during and after (exit and post-poll).4 However, the scope of these surveys is much
broader than the simple use of predicting seat and vote shares of political parties. This includes, but is not
limited to, understanding political processes beyond the electoral, how political opinions and attitudes are
understood and articulated and how these opinions relate to broader political and narrower election
1 Begun by the Centre for the Study of Developing Societies (CSDS) with the NES of 1967 and run continuously since 1996 by
the Lokniti programme of CSDS. The Methodological Note for the NES 2009 traces the different ‘generations’ of these surveys
(Lokniti Team, 2009). This also discusses the differential sampling strategies across the three generations of the survey, as well
as specic methodological innovations made in the 2009 survey. Certain major state-election surveys are also part of the larger
election studies conducted by Lokniti.
2 See the August 2016 issue of Seminar on ‘Measuring Democracy’ for a discussion; see especially Kumar, Rai and Gupta (2016).
3 See Heath (2016) for an example of the differences and similarities between election surveys in the UK and in India; see Hillygus
and Guay (2016) for the US experience. The latter also discuss alternative polling methods used.
4 Exit polls are usually conducted at the time of voting and people are interviewed as soon as they have cast their vote. Post-polls
are conducted after the vote is polled and before the declaration of results, and usually involve home visits. In terms of sampling
strategies pre- and post-polls more often use random probability sampling, whereas exit polls do not (see Rai, 2014, p. 15). The
specic context of the NES post poll is discussed in the present note.
Notes on Method
Corresponding author:
Divya Vaid, Centre for the Study of Social Systems, Jawaharlal Nehru University, New Delhi, Delhi 110067, India.
E-mail: divya.vaid.09@gmail.com
Studies in Indian Politics
7(2) 267–273, 2019
© 2019 Lokniti, Centre for the
Study of Developing Societies
Reprints and permissions:
in.sagepub.com/journals-permissions-india
DOI: 10.1177/2321023019874916
journals.sagepub.com/home/inp
Note: This section is coordinated by Divya Vaid, divya.vaid.09@gmail.com
To continue reading
Request your trial