February 23, 1998

To gauge voters' mood on election eve, India Today commissioned the Centre for the Study of Developing Societies (CSDS) to conduct a nationwide survey. It was supervised by V B SINGH, principal coordinator, and SANJAY KUMAR, national coordinator. The data was interpreted by YOGENDRA YADAV of CSDS and RAJEEVA KARANDIKAR of the Indian Statistical Institute.

This survey is the first all-India "panel" study over two parliamentary elections, and one of the largest of its kind undertaken anywhere. Between February 4 and 8, our researchers went back to the same voters who were interviewed by CSDS during and after the 1996 Lok Sabha elections. This method allows for more accurate assessment of vote swings among parties than any snapshot survey that has an even bigger sample. For the survey, a target sample of 15,500 voters was selected on the basis of multi-stage random sampling in 452 polling stations in 114 Lok Sabha constituencies spread across 21 states. A total of 8,938 voters were successfully contacted. They included 50 per cent women, 17 per cent Dalits and 9 per cent Adivasis; 77 per cent of the respondents lived in rural areas. Besides the 7,457 who were interviewed in 1996, the sample also included 1,481 fresh respondents, both to compensate for those who could not be contacted as also to record the mood of new voters.

Congress, BJP fighting to a Standstill

The swings for different parties calculated from the survey findings were fed into a sophisticated statistical model to arrive at the number of seats that the various parties could garner. Instead of relying on "crude count" -- based on which party was likely to get the most votes in each constituency -- a probability count was used to reflect each party's chances of winning a seat. This helps in accurately forecasting close contests. For very small states, where either the survey was not held or the sample was too small, assessments of probability were directly fed into the model.

The big question during an otherwise dull election campaign was whether the verdict would end the stalemate in the Lok Sabha. The India Today-CSDS survey shows that it may not. The BJP and its allies will remain the largest bloc in the 12th Lok Sabha. A swing of 6 per cent votes in its favour over its 1996 vote share is likely to increase its tally by 27 seats, but it is unlikely to make the kind of gains it was expected to. The Congress, which appeared to be on its death bed, is showing signs of life again, with a 2 per cent swing and a gain of 21 seats over 1996. On the other hand, a swing of 9 per cent against the United Front (UF) may see its strength in the Lok Sabha come down by as many as 53 seats from the 180 seats it won last time.

The survey indicates that the Congress and the UF together will continue to control a majority of seats in the new Lok Sabha, dashing hopes the BJP may have had of forming the government. As in 1996, government formation may again necessitate the tenuous understanding between these two formations.

For the first time in the history of electoral politics in independent India, the Congress faces serious competition for the position of the largest party or electoral front in terms of votes. The survey shows that the Congress with 32.2 per cent may just remain ahead of the BJP and its allies with 31.3 per cent, though this difference between the two is smaller than the margin of error. About two-thirds of the BJP's gain in vote share comes from its newly acquired allies. The BJP itself may not register more than a 3 per cent swing in its favour.

Normally, a 6 per cent swing should bring in a sizeable number of additional seats but the survey does not predict any such substantial gains. The reason is that the swing in favour of the BJP is not uniformly spread. Thanks to the alliance with the AIADMK, it may get more votes in Tamil Nadu but these may not translate into seats. On the other hand, in states like Maharashtra where a minor swing could have made a huge difference, the party is in fact losing votes. A large proportion of its newly acquired votes are therefore either surplus or wasted. Between the 1991 and 1996 elections, the BJP did not gain votes but improved its tally substantially. This time it may be the other way round.

The regional breakdown of results shows the BJP making some gains in the north. In Uttar Pradesh the divided opposition and a positive vote swing could see the party making gains at the expense of the Samajwadi Party. Bansi Lal faces a rough ride in Haryana.

In the west, the BJP looks set to concede seats to the Congress. In Maharashtra, the Congress-SP-RPI alliance could cause damage to the BJP-Shiv Sena alliance, which could outweigh gains for the BJP in Gujarat and Madhya Pradesh. It is making progress in the east and this may well translate into seats. Mamata Banerjee's Trinamool Congress has succeeded in undermining the prospects of the Congress. Laloo Yadav's RJD is under pressure from the BJP-Samata alliance, but is not out.

For the first time, the BJP and its allies could send a noticeable chunk of the MPs from the south. The BJP is set to gain from its alliance with R.K. Hegde's Lok Shakti, while the Janata Dal (JD) is headed for disaster. Though an anti incumbency factor is evident against the TMC-DMK, it is too insignificant to alter the equation in Tamil Nadu.

Who is the best Prime Minister?

The Vajpayee factor has been the BJP's trump card in this election. Has Atal Bihari Vajpayee met his match in Sonia Gandhi? The poll shows that Vajpayee continues to be the most popular choice for prime minister. The survey figures of 22 per cent for Vajpayee may not accurately reflect the ranking at the time of elections, for peoples' awareness of Sonia's activity is changing rapidly.

None of the other political leaders touches even 5 per cent popularity. Incumbent Prime Minister I.K. Gujral gets the backing of 3.4 per cent, the prestige of office giving him a slight boost, while Congress President Sitaram Kesri is languishing with less than 1 per cent support. Gujral is followed by Jyoti Basu in this respect.

The figures reported here are much lower than those in other surveys because most opinion polls offer people a menu of names from which to pick, whereas this poll just asked respondents to name the person who they would like to be prime minister, without suggesting anyone. Consequently, 39 per cent did not mention any name. This gives a more accurate picture of the actual state of people's preferences. At any rate it may be risky to read too much into the prime ministerial preferences, for this index has tended to be rather fickle.

Who would you like to form the Government?

All figures in per cent
Congress + allies	BJP + allies	United Front	Others

Congress and the BJP are neck and neck in the race to form the next government, with both backed heavily by their partisan supporters. But for the United Front (UF) the picture is more hazy. Nearly a third of UF's supporters do not know which party should head the next UF government, while only 39 per cent state emphatically that a UF government should assume power after the election. This perhaps is a reflection of the decline in the UF's fortunes.

Performance of UF government: Despite being in office for 20 months, the two UF governments seem to have made little impact on ordinary voters. Even among UF supporters, only 17 per cent thought the governments had done very well. Yet, only 23 per cent of respondents in the survey believed that the Congress had been right in bringing down the two governments.

As the election campaign reaches its climax, most responses are along party lines. BJP supporters are equally dismissive of the UF and Congress governments, and think that the Congress was wrong to force an election. As expected, most Congress supporters think that their party did better when it was in government than the UF.

More

© Living Media India Ltd