NEW DELHI, 25 April 2004 — There is one exit poll no politician ever forgets: It is written in the eyes of a public meeting. You can bring a crowd to the water, but you can’t make it drink. Cheerleaders might go hoarse, but politicians know the difference between a roar of the crowd and a gurgle. There are tell-tale signs: People get restless, or leave mid-speech. Note that most of the crowd has been assembled by the party machine, so is partisan or inclined toward the party platform. So if the faithful begin to shuffle before the oratory ends, the story is virtually over. Congress leaders who were with Mrs. Indira Gandhi in 1977 recall that she got her first intimation of bad news when at the end of her meetings crowds were less than full-throated when she asked them to repeat after her, “Bharat Mata ki jai!”. In better times, there would be all sorts of “jais” through the speech.
A view has been nurtured by psephologists that the opinion polls they conduct are “scientific” while the message in the eye of the voter is erratic. It is a self-serving view. The eye, in fact, might represent a higher science, for chemistry is more human than mathematics.
There are merits in both, as well as pitfalls. Companies like ORG Marg and AC Nielsen, probably the two best in India, make it a point to bang a huge drum when they get the end-figure right, and change the subject when they get it wrong. Last year’s state elections will suffice. ORG Marg (working for Aaj Tak and India Today) gave the BJP between 55 and 65 seats in Rajasthan, and the Congress between 120 and 130. AC Nielsen gave the BJP 67 seats and the Congress 113 in Outlook. The results? The BJP got 120 seats and the Congress 56. Talk of being spectacularly wrong.
ORG gave the BJP between 30 and 40 seats in Chhattisgarh and the Congress between 46 and 56 seats. Nielsen said the two would be dead even at 43 each. The final tally? BJP got 50 seats and the Congress 37. Opinion polls permit themselves a 3 percent comfort zone margin of error. In Rajasthan it was more like 300 percent. Such science is not infallible.
Throw your mind back to the Aaj Tak-India Today polls before the Gujarat elections. One done before the election process got properly started gave the BJP a huge victory. A second, done on the eve of the vote, pared this to a paper-thin edge. Other opinion polls, as for instance in the Week actually predicted victory for the Congress. We all know what the results were.
An interesting conclusion emerges. The voter does not care a damn about opinion polls. The controversy over whether polls should be published during the electoral process is therefore a non-issue. The Indian voter is not a headless chicken. He does not rush about from one direction to another at the last minute. His decision matures over time, and once made rarely shifts. The BJP voter in Rajasthan last year did not fall into deep depression after seeing Aaj Tak and walk towards the political funeral pyre to commit sati. He had made up his mind and voted accordingly. He may not even have misled the pollsters. There is so much weightage given to methodology that it can often send projections askew. Different barometers promise different squalls and storms. Nielsen did the fieldwork for both The Asian Age and the NDTV polls this month. But Nielsen gave NDA 25 seats in Bihar, while working on research by the same company NDTV gave NDA only 14. Sometimes the methodology is inexplicable. This week one opinion poll extrapolated the results of the first round of voting to a national conjecture. This is absurd. Regions within a state have their own dynamics, and vote virtually independently of each other; and often at cross purposes. For instance, the vote in Andhra Pradesh in the first round was concentrated on Telangana, where there is a strong sentiment for separation from Andhra and therefore an anti-Chandrababu Naidu wave. Naidu’s strategy is to win seats outside Telangana vote by rejecting the separatist demand.
Politicians, naturally, take no chances: they double-check what they get from “scientists” with what they see on the trail. Mrs. Sonia Gandhi, for instance, ignored Karnataka in the last days before the second round even though one poll thought that S.M. Krishna might pull off an assembly win despite losing the Lok Sabha battle. She clearly did not see much hope in the eyes of those who came to her meetings in Karnataka. She concentrated, instead, wisely, on Andhra Pradesh, where the Congress has a good chance of following up the body blow it delivered in Telangana with an upper cut elsewhere. It was interesting that Mrs. Gandhi chose to spend two vital days in her own constituency, Rae Bareli, instead of campaigning elsewhere. That seems like misplaced apprehension. Surely Amethi and Rae Bareli should be the two safe seats for the Congress in Uttar Pradesh.
Prime Minister Atal Behari Vajpayee concentrated on Maharashtra and Bihar in the last phase of the second round, two “swing” states whose results could determine the size of the NDA victory. As they say, size matters.
This is the appropriate moment to remove a misconception that persists, and continues to be promoted by opinion polls. The meaning of “victory” has altered. Once, when the Congress was the dominant player, it meant getting a simple majority in a popular mandate. But since the formation of the P.V. Narasimha Rao government in 1991, victory belongs to the party that is most capable of winning a confidence vote in the Lok Sabha. The Congress did not win a majority in 1991, but was sworn in. Rao skillfully retained his artificial majority through five years of maneuvering, helped by the fact that the opposition was too divided over the central issue of the time, the Babri-Ram mandir dispute, to challenge him.
Since 1996, coalition governments have become routine, and a precedent has been established by the president of India: the largest single party in the Lok Sabha gets the first chance to form a government. This explains why Vajpayee was prime minister for about a fortnight the first time he was sworn in. Both the president and the prime minister knew that the BJP would not survive, but the principle was affirmed.
Vajpayee will therefore be asked to form the next government in May since the BJP will be the largest single party. There is no doubt about that. But some doubts have arisen as to whether the NDA will get a comfortable majority. I say this not on the basis of variable and varying opinion polls, but my own observation. Chandrababu Naidu is not going to deliver the number of MPs he did to the coalition in the last election. That is clear. The BJP will also dip in Andhra because it had MPs from Telangana. The question is: How many MPs will Naidu deliver? Ten? Twenty? Much will depend on that. The NDA is also in trouble in Tamil Nadu, where Jayalalitha is paying for a few accumulated political sins, in particular her vengeful decision to send Vaiko to jail. Her friends report that she is beginning to turn the tide; her adversaries are gleeful that she will be decimated. We will know what the voters say in a couple of weeks.
But Vajpayee will get the support he needs, if not from inside then from outside. The prime minister’s personal popularity and the confidence that the country has in him is undisputed. That will be the rallying point. In fact, the rallying has begun already. The prime minister has already said, quite specifically, that the constituents of the ruling alliance will increase after the elections. George Fernandes, chairman of the NDA, has been harping on the welcome he will provide to his “good Lohiaite friend” Mulayam Singh Yadav. It is no accident that Yadav is contesting for a Lok Sabha seat. Nor would Mayawati, also heading for Parliament, be averse to a place in the Cabinet. K. Karunanidhi and Sharad Pawar might have been allies of the Congress in the states, but would they be allies of Sonia Gandhi at the center?
The future lies not with exit polls but with an exit line. They have all fought separately in the states to unite at the center.
