ABSTRACT
In this article, we study how the introduction of a None of the Above (NOTA) option in Indian elections since 2013 has affected voting patterns and election outcomes. First, we examine the relationship between position on the ballot with electoral outcomes by constructing a simple statistic that measures relative performance by position on the ballot or Electronic Voting Machines. Within non-party options, we show that the last option on the EVM ends up being a salient choice, which is an inversion of the usual findings in the literature on order effects, where earlier options are generally preferred. Second, the NOTA option is the salient choice for voters who choose not to vote for a party, and thereby substituting for independent candidates. This share is more than can be explained by ballot position alone. We argue that the existence of a NOTA option is not merely a procedural technicality; instead, voters use NOTA as a strategic option to express their protest.
Acknowledgement
We would like to acknowledge insightful discussions with T.S. Krishnamurthy, Jagdeep Chokhar. V. Suresh, Brinda Viswananthan and Manjushree Hegde.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 For details, see People’s Union for Civil Liberties v Union of India (Citation2013): Writ Petition (Civil) No 161 of 2004, Supreme Court judgment dated 27 September 2013.
2 Specifically, rules 41(2), 41(3) and 49-O. 41(2) and (3) of the Conduct of Election Rules, 1961 directed the voter who took the ballot and decided not to use it to return it to the presiding officer. These ballots were then kept in a separate packet. Rule 49-O catered to the voter who, in the polling booth, chose not to cast vote and put it on record.
3 For more details on the rules see https://indiankanoon.org/doc/152734757/ and https://eci.gov.in/faqs/evm/general-qa/electronic-voting-machine-r2/, accessed on 17 July 2022.
4 Past analyses of elections have included metrics like the Herfindahl index, defined as the sum of the squares of percentage shares (this is minimized for a uniform distribution); and the Kullback-Leibler distance, which is an entropy-based measure of how different a distribution is from the uniform. Our statistic is much easier to calculate, and has the following properties—A candidate with a lower percentage share than the overall average will have a negative DPS value, and a candidate with a higher percentage share than the overall average will have a positive DPS value. The minimum value of the statistic, −1/N, is reached when the candidate in question gets 0 votes and the maximum value, 1−1/N, when he gets all the votes. Further, DPSi values always sum up to zero over all candidates for a given election
5 Subramanian Swamy vs A. Raja on 24 August 2012; In 2017, all the accused in the 2G scam were acquitted by the Special CBI Court Judge O. P. Saini since the prosecution could not prove the case.
Additional information
Notes on contributors
Raghav Kumar
Raghav Kumar works in data science after completing his masters degree from Madras School of Economics, Chennai, India. His thesis focussed on changes in voting patterns after introduction of NOTA in India.
Sudarsan Padmanabhan
Sudarsan Padmanabhan is an Associate Professor in the Department of Humanities and Social Sciences, Indian Institute of Technology Madras, Chennai, India. Sudarsan specializes in social and political philosophy, Indian philosophy and culture. Sudarsan's research focus is on the confluence of law, politics, and ethics in the public sphere.
P. Srikant
P. Srikant is a Visiting Faculty at Madras School of Economics, Chennai, India. Srikant specialises in finance.