February 23, 1998

Why do they Vote the way they do?

The fulfilment of basic needs rather than issues like corruption or instability is what bothers people. This is true across the political spectrum, with 47 per cent of the Congress, 41 per cent of the BJP and 40 per cent of the UF supporters saying their daily needs were their paramount concern. Political instability and high rate of turnover in governments were mentioned by a mere 13 per cent and corruption by 14 per cent, with BJP supporters appearing more concerned about these issues than those of the Congress or the UF. Twelve per cent said they were concerned about violence and communal strife. Our respondents clearly did not think much of the two UF governments of the past 20 months. Given a choice, 34 per cent said they would prefer the last Congress government to H.D. Deve Gowda's and I.K. Gujral's. This strident anti-incumbency factor perhaps explains the pronounced swing against the UF.

Can a Late Swing make a difference?

Have you made up your mind?

All figures in per cent

Despite all our efforts to eliminate bias and errors, the forecast presented here is subject to several limitations. Any sample survey of this size carries a 2 per cent margin of error. Thus a forecast of 31 per cent votes for the BJP and its allies means that the BJP's support base could range from anything between 29 per cent and 33 per cent. Under our electoral system, a variation of this size can lead to 20 or more seats changing hands. That apart, the survey was begun when many voters were not even aware who their candidates were, and it was completed with one to three weeks to go before the polls. Almost 13 per cent of our respondents were not even aware that polls were round the corner when our researchers approached them, and as many as 36 per cent were still undecided about whom to vote for. The pattern was uniform among supporters of different parties. It therefore introduces an element of uncertainty in any poll forecast.

Things can change and ideally, another round of surveys, closer to the second and third phase of polling, would have enabled us to give an updated picture. But the Election Commission's order banning the publication of opinion polls after February 14 prevents us from undertaking such an exercise that would take into account the late swings. Finally, a pre-poll survey has no way of foretelling which of the respondents will actually vote on the day of polling. A differential turnout for supporters of various parties can make a critical difference, especially if the turnout is low.

More

The survey was coordinated by social scientists Pradeep Kumar, K. Srinivasulu, Apurba K. Baruah, Rajendra Ravi, Anupama Roy, P.M. Patel, T.R. Sharma, Sandeep Shastri, Gopa Kumar, Ram Shankar, Suhas Palshikar, Bijoy K. Bohidar, Sanjay Lodha, G.K. Prasad, Sudhir Hilsayan, Madhu Sudan Mishra, V.K. Rai, Peter R. deSouza and Surajit C. Mukhopadhyaya. The data analysis and prediction team included Arup Pal, Abhay Bhatt, Suresh Muthuswami, Lakshmi Iyer, Alistair McMillan, Himanshu Bhattacharya, K.A.Q.A. Hilal, Kanchan Malhotra, Praveen Rai, Chitrali Singh, Ashutosh and Rana Dhirendra.

© Living Media India Ltd