Indian Politics & Policy Volume 3, Number 1, Spring 2020 | Page 43
Do Issues Matter in Indian Elections?
the issues. In 2019, Congress’s manifesto
prioritized employment, Minimum
Income Support Programme or Nyuntam
Aay Yojana (NYAY), and national
security. However, for the BJP, the issue
of national security topped the list.
Months before the general election, the
attack in Pulwama (India) and India’s
retaliation against Pakistan, led by the
incumbent government, steered the
election narrative towards national security
and seemed to have helped them
electorally. 27
In 2014, the BJP fought the election
on price increase, employment,
and corruption (which were the top
three issues according to voters in the
NES survey as well) and Congress prioritized
the issues of secularism, Foreign
Direct Investment, low inflation,
minimum wage laws, and employment.
Corruption as an issue was at the bottom
of Congress’s manifesto for the
general election of 2014. In retrospect,
it is evident that in the general election
of 2014, the BJP’s manifesto better reflected
the voters’ choices of issues, or
perhaps the party successfully made
them the issues of voters.
III. Data and Variables
The paper takes data from the NES
of 2019 and 2014, as collected by
Lokniti-CSDS. 28 For the purpose
of understanding the voter’s behavior
related to issues, the two largest national
parties, Congress and the BJP were
selected for analysis, excluding their
allies at the time of election. The sample
for the present analysis consists of
only those parliamentary constituencies
where there was a contest between
Congress and the BJP.
Dependent Variable
The dependent variable is the response
to the question Whom did you vote for?
The dependent variable has the values
of 1=Voted for BJP and 0=Voted for
Congress. Any respondent who mentioned
voting for any other party except
for Congress and BJP was left out of the
analysis.
Independent Variable
Responses to the question According to
you, which was the single most important
issue while voting? were the independent
variable. The analysis consists of the top
four election issues, mentioned by the
respondents/voters, with reference to
“any other issue” mentioned by the voters.
Any respondent who did not mention
an issue was left out of the analysis.
Further, some demographic variables
are also considered in the model,
such as gender (coded as male and female),
religion (coded for Hindus, Muslims,
and other religions), caste (coded
for scheduled castes, scheduled tribes,
other backward classes, and all other
castes), monthly household income of
voters as reported by the respondent
(coded into two variables of below and
above Rs. 5000), locality of residence of
the respondent (coded into rural areas
and urban areas, which include towns,
cities, and metropolitan cities), age of
the respondent (coded into youths and
others, as youths and first time voters
are expected to vote differently than
others; those who are between 18–29
39