NDA on top but as poll heat rises, tally begins to wilt

In some of these states, the swing away from the NDA is quite small, while in others, it is significant enough to make a difference to the NDA’s tally.

Since the first poll, the NDA has lost some seats to the Congress alliance in a number of states, particularly in Bihar, where the Laloo Yadav-Ram Vilas Paswan alliance seems to be working.

There have also been smaller losses from the NDA to the Congress alliance in Andhra Pradesh and Maharashtra.

To provide the national picture, it has been necessary to include those states where no opinion polling has been done this time. For those states, the seat projections from the first poll have been used. The projection currently suggests that the NDA alliance will have its work cut out, if it is to gain a majority.

The Indian Express and NDTV jointly commissioned A C Nielsen, a leading market research agency, to conduct fieldwork for a series of opinion and exit polls.

The first one was conducted around mid-March and the first phase of this second opinion poll was conducted about 2-3 weeks later to gauge poltical popularity in the 140 seats that are voting on April 20.

This second round will be covered in four phases—to correspond with the four phases of the elections. This opinion poll covers some questions that will be asked in each of the four phases. The results in today’s report cover the responses of voters of the constituencies going to the polls on April 20. In the next phase of the opinion polls, we shall update these figures to reflect the results of the first two phases of voting.

The fieldwork for this poll was conducted between April 6 and April 13. The sample size of 16,393, spread over 57 of the 140 Lok Sabha constituencies going to the polls. But some words of caution.

Opinion polls measure voter preferences at the time when the poll is conducted and all opinion polls are subject to margins of error. The sampling error in this poll ranges from 1.7% in Andhra Pradesh to 4.5% in Jharkhand (where only six out of 14 constituencies are going to the polls in this round).

It is only in two states that elections are held for all seats in the first phase—Gujarat and Chhattisgarh. Thus, the projections for these two states are for the entire state. For the seven other states, it has been assumed that the same swing in this first phase will also apply to the rest of the state.

Methodology: We went back to the same 57 constituencies that we had surveyed in these nine states. In two of those, namely Gujarat and Chhattisgarh, the entire state is going to the polls. In the other states, we shall conduct another poll next week to cover those constituencies that are voting in the second phase of these elections.

In each of these 57 constituencies, four assembly segments were selected at random. Within each assembly segment, polling booths were selected at random. In each polling booth area, 18-20 interviews were conducted in homes.

For the voting intention question, the respondents were given a mock ballot paper with symbols of the parties. They were asked to mark their preference on the ballot paper and then place the ballot paper in a mock ballot box.

In the three states where assembly elections were also conducted, the respondents were also asked to indicate their voting preference for the Vidhan Sabha polls.