[To read the full book "Why the Vajpayee Government lost the 2004 Lok Sabha polls", open https://www.amazon.in/Vajpayee-Government-Sabha-polls-analysis-ebook/dp/B08PRW5BNH/ The below is some part of one chapter of it.]
Did ‘every’ exit poll show NDA winning in 2004?
It is often claimed that ‘every poll showed the Vajpayee Government coming back to power’ in 2004, but the actual results defied the polls. It is not true. Exit polls (or at least, some of them) indeed showed the NDA failing to form the Government, only issue was the number of seats.
The NDTV-Indian Express exit poll predicted 230-250 seats for the NDA, showing the ruling coalition short of a simply majority by at least 22 seats, and at most 42 seats. [https://www.rediff.com/election/2004/may/10exit.htm] This poll gave Congress and allies 190-205 seats, Left parties 40-50 and 60-70 to Others. Others included Samajwadi Party which would have been given 25-30 seats, BSP 20 seats (Samajwadi Party finally got 36, and BSP 19, and Left 59).
The final results were- Congress and allies 219, Left 59 (including only the 4 major Left parties of CPM, CPI, RSP and Forward Bloc, not including other LDF allies like JD(S) in Kerala) which gave a clear majority of 219 + 59 = 278 seats. The Samajwadi Party with 36 seats and BSP with 19 seats also supported the Government from outside, to make it 333 Lok Sabha seats for the UPA. Let us see the range of the NDTV’s exit poll. Its upper limit for the Congress and allies was 205, only 14 away from the final number of 219. Adding NDTV’s average numbers of Congress and allies (198), Left (45) and Samajwadi Party (30) it was already a majority for the Congress and allies. Adding the upper range of Congress and allies and Left (205 + 50 = 255), that itself was close to a majority at 255 seats, and only 23 short of what they actually got.
The problem was the bracketing of ‘Others’ as a separate entity and not seen as Congress supporters later. The Left had declared even before the polls were over that it would support the Congress. Had the Left tally been added to the Congress allies’ continuously, it would have been easy to see how close it was to majority.
While discussing this exit poll, NDTV panellists Prannoy Roy, Dorab Sopariwala and Rajdeep Sardesai expressed serious doubt over whether Atal Bihari Vajpayee would be the next Prime Minister. Rajdeep Sardesai said that if the Congress and allies, whose upper limit was given by NDTV as 205 seats, cross the NDA tally, whose lower limit was given as 230 seats, then they would come to power with Left support. That is exactly what happened.
However, Star News-C voter exit poll gave the NDA 263 to 275 seats and the Sahara-DRS survey put the BJP-led combine between 263-278. Both showed the Congress-led combine improving its performance over 1999 with the Star News poll projecting between 174 and 186 seats and the DRS poll giving it between 171 and 181 seats. Even these polls which showed the NDA close to majority, showed a big tally for the Congress and allies, namely 180 seats, and adding Left’s seats they would have been at a healthy 230 seats or so.
Here, while showing the final exit poll results, the Star News journalist (most likely he was psephologist Yashwant Deshmukh) said: “If the NDA’s vote share is just 1% more than predicted in this exit poll, its seat tally will be 301-313. But if it’s only 1% lesser than predicted, its seats will come down to 227-239.”
One must say that all polls do say that there is a range of + or – 3%. If the NDA’s seats decreased or increased by 36 to 38 seats with an increase or decrease of just 1% vote, one can assume that if the vote was 2 or 2.5% lesser than predicted by the exit polls, it’s seats would have been between 185-190, i.e. in the range of the actual 187. That would actually prove any critic of exit polls as correct, that this particular poll at least, had a very big range, from a big win for the NDA to a clear loss for it, with the upper limit being 300+ and lower limit being below 200. However, even this exit poll had indeed warned that even a decline of 1% vote would result in the NDA tally going down to 227-239.
Aaj Tak forecast a hung Lok Sabha giving the NDA 248 seats in the 543-member House, a loss of 54 from 1999, Congress and allies 189 and Others 105. Even here, adding Left’s 50 (say) and Samajwadi Party’s 30 would put the Congress and allies at 269 (i.e. 189 + 50 + 30 = 269), just 3 seats short of majority, and that’s not including other parties like JD(S) which won 2 seats in Karnataka and 1 in Kerala, and parties like BSP with 20 seats. The Zee-Taleem exit poll also projected a hung House with NDA getting only 249 seats and Congress and allies were pegged at 176 while ‘Others’ were expected to get 117 seats.
Where all the polls were indeed wrong was in giving NDA more than 200 seats, at least 230 seats, while it actually got only 187 seats. The state where every single exit poll was wrong was Uttar Pradesh, where the lowest being given to the BJP was 29 out of 80 (by Aaj Tak) while others gave 30-35 seats, and the party itself was expecting 40-45 seats, but the final result was 10 seats for BJP and 1 for ally JD(U) to make it 11.
The problem with these exit polls which showed the BJP-led NDA falling short of majority was that almost everyone who expressed their opinions had thought that they tended to under-estimate the BJP, because that is what they had done in the Gujarat Assembly elections in December 2002 and in the State Assembly elections in Rajasthan, Madhya Pradesh, Chhattisgarh and Delhi in December 2003. Though the exit polls got the winners right in Madhya Pradesh and Delhi, most of them under-estimated the BJP massively everywhere, including in Madhya Pradesh and Delhi. Only one exit poll showed BJP winning 25 seats out of 70 in Delhi while it actually won 20. Some others gave it as low as 11 seats in Delhi. The Star News exit poll showed Congress winning 89 and BJP winning 91 seats out of 200 in Rajasthan, while the BJP actually won 120, which not a single exit poll showed. Nor did any poll show BJP winning 50 seats out of 90 in Chhattisgarh with 37 to the Congress, though India Today’s opinion poll in October/November 2003 showed both BJP and Congress in the range of 38-48 out of 90, saying it was ‘too close to call’.
So these polls underestimated the BJP’s performance. As a result, when these polls showed BJP-led NDA short of a majority, or barely touching majority, it was felt by many that they were under-estimating the BJP. The fact that they could be over-estimating the BJP and/or allies by the same margin was not expressed by anyone prominently on TV.
After some exit polls were broadcast in the initial phases of the 2004 Lok Sabha polls, which did show NDA losing ground, the exact above thoughts were mentioned by many, like Arun Nehru (1944-2013) on Aaj Tak. Organiser (2 May 2004) reported in an article by R Balashankar:
“NDA the clear winner
But psephologists prop a Congress ‘catch up’
Exit
polls have given a clear edge to NDA, but they have generously consoled the
Congress of catching up. In modern day politics, psephology has replaced
astrology for clairvoyance. Psephologists like astrologers, only give us the
trend, the sign board, it may or may not take up to the destination.
In the coalition era, election is more arithmetic and less chemistry. But ask
the psephologist. He will give you, like an astrologer, an explanation for any
result, real or perceived. But there are some clear patterns. If we analyse
all the opinion polls and exit polls in the recent past, the prediction for the
BJP was always less than what they actually got, and the projection for
Congress was always higher than what they finally got. Why does this happen?
There is a huge lobby in Indian politics that thrives on instability, hung outcome and uncertainty. The common pattern that emerges in all exit polls point to this: a carnival of this hung lobby….
A quick analysis of the track record of exit polls indicates that in the last round of assembly elections, none of the exit polls saw the clear victory of BJP in Chhattisgarh and Rajasthan. The landslide BJP victory in MP was predicted at best as ‘BJP has an edge’. The only state where they saw a clear verdict was Delhi, where incidentally the Congress was winning. It becomes obvious that wherever the NDA alliance was winning, the exit polls kept the final verdict hanging. They suggested neck and neck, close, tough fights. And the ‘winner is’ hung house!...
To polls-Phase-II
Congressmen never tire of pointing out that the exit polls for the last assembly elections were off the mark and so it would be this time. The early trend of the exit polls indicates a NDA comeback, though all polls contradict one another substantially.
Opinion polls and exit polls may be scientific and correct elsewhere in the world, where there is homogeneity in socio-political parameters. In India they vary from region to region.
In the last three elections, six weeks before polls the pollsters pegged NDA at the top of the chart and the battered Congress at the bottom. Midway through, the Congress graph climbs, the NDA plunges, with the forecasts suggesting (only suggesting) the NDA slowing down, peaking early and Congress fast catching up. This pitch is maintained until the final results are out. Polls-exit or opinion are like the astrology columns in daily newspapers. Read it for a tickle and leave it in the dustbin. And like astrologers they have those wide elbow room-the undecided and the 5 per cent error margin. In a rational analysis psephology is so much astrology as astrology is so much science. Time and other data are crucial for both. And they will predict ‘other things remaining the same’-this will be the result. We all know other things never remain the same…
Like astrologers pollsters are known to give predictions to please their client. Taking the veil of 3-5 per cent error margin and undecided they go about browbeating the political marketplace. How else does one explain the huge variations in the exit poll figures from channel to channel. If it is science such variations have no place...”
https://www.organiser.org/archives/dynamic/modulesc9f4.html?name=Content&pa=showpage&pid=21&page=2 or https://organiser.org/2004/05/02/27252/general/r9c4aac1a/
Although NDTV’s exit poll showed Congress & allies, Left and Samajwadi Party winning a majority of the seats, there were discussions that in case the NDA fell short of a majority, Congress allies who were ex-NDA allies too could return to the NDA fold. Rajdeep Sardesai said on NDTV that “DMK, MDMK, PMK were with NDA till December 2003. They may be the first to return to the NDA in case it is found short of numbers.” A correspondent of Star News said that except Congress, Left, RJD and Muslim League, any party can support the BJP.
As a result, even though the NDTV exit poll actually showed something else, the feeling that Atal Bihari Vajpayee will become Prime Minister again remained. On NDTV the panelists also felt that the Telangana Rashtra Samiti (TRS) which they showed as winning 5 seats contesting in alliance with Congress, could support the BJP if the NDA fell short of numbers, as ‘the BJP has ideologically always supported Telangana.’ Though it is a fact that the BJP has ideologically always supported Telangana, it had to drop that demand due to pressure from TDP, with which it entered into an alliance, which was against bifurcation of the state of Andhra Pradesh.
After the last phase of polling, one BJP leader Yashwant Sinha (1937-) was asked on TV about his prediction of the final results, and he said: “240 seats for the BJP, and 340 seats for the NDA”. This appeared to be a genuine belief on the face of it, because India Today’s survey in the ‘Mood of the Nation’ poll of January 2004 had shown the NDA winning 330-340 seats. While all exit polls were showing the NDA at 230 to 278, this leader (like many others) expected that they under-estimated the NDA by 60-70 seats at least. They actually over-estimated the NDA by that number of seats.
[Above is a part of the book “Why the Vajpayee Government lost the 2004 Lok Sabha polls- An analysis” This book on the 2004 Lok Sabha polls exposes a lot more media propaganda, particularly before the 2004 Lok Sabha polls, and before the 2002 Gujarat State Assembly polls. To read the full book "Why the Vajpayee Government lost the 2004 Lok Sabha polls", open https://www.amazon.in/Vajpayee-Government-Sabha-polls-analysis-ebook/dp/B08PRW5BNH/ ]
No comments:
Post a Comment