Tuesday, December 18, 2007

NDTV MAKES A MOCKERY OF EXIT POLLS

The recent assembly elections in Gujarat had almost the whole nation riveted. On one side was Narendra Modi and on the other were almost all others, including major elements of the Sangh parivar like the VHP and the RSS and disgruntled local BJP leaders.

When the first phase of polling took place on December 11, 2007, everyone awaited the results of the exit polls which were being conducted by various agencies, including leading news channels. Predictably, all exit polls showed a swing away from Narendra Modi resulting in a significant loss of seats to the BJP. But, all except NDTV showed the BJP still ahead of the Congress in the first phase, which saw ballots being cast for 86 seats.

Prannoy Roy, at one time the most respected Indian psephologist, and his assistant, Dorab Sopariwala, on the other hand, sprang a major surprise, showing on NDTV that the Congress had actually defeated the BJP in the first phase, by winning an astounding 43 seats, a jump of 13 from the seats it had won in 2002. Roy took pains to inform the viewers that the elections were held in a tense atmosphere and that Modi did not know the first name of Mahatma Gandhi. He also did not forget to inform the nation that Rahul Gandhi had campaigned extensively in the area for the Congress party. For Roy, the Rahul Gandhi factor was working again, just like he had hyper excitedly thought it had in the UP elections.

After the second and final phase of polling on December 17, 2007, when the interest in the outcome of the polls had reached an almost feverish pitch, all exit polls, including the one conducted by NDTV showed that Narendra Modi would, in all probability win the elections again, though with a reduced majority. That was not a surprising deduction, given the fact that most of the BJP leaders who had rebelled against Modi had their spheres of influence in the parts of Gujarat that went to polls in the first phase.

After giving out the precise number of seats likely to be won by each party in the first phase, along with statistical explanations, NDTV was expected to be equally precise in its forecast in the second phase too, considering that it had the largest sample size and the expertise of Prannoy Roy. But, surprisingly, what we saw was an unbelievable joke.

NDTV gave the Congress party, believe it you must, between 27 and, yes, 52 seats in the second phase! In percentage terms, this represents a variation of an absurd 92 percent above the minimum seats projected! The BJP was projected to win between 50 and 70 seats, a variation of 40 percent. Based on these figures and those of the first phase, the BJP was given a generous band of 90 to 110 seats while the Congress was projected to win between 70 and 95 seats, a whopping 35.7 percent cushion!

The table below explains it all nicely:-

-------------------------------------------------

BJP…………………...Seats……….. Variation

Phase 1……………….40…………………-

Phase 2……………….50-70………… 40%

Total……………………90-110………..22%

--------------------------------------------------

Congress……………Seats……….. Variation

Phase 1……………….43………………..-

Phase 2………………27-52……………92%

Total…………………..70-95…………..37%

---------------------------------------------------

Others………………Seats………….. Variation

Phase 1………………3…………………….-

Phase 2………………0-2…………………!

Total…………………..3-5……………….66%

----------------------------------------------------

What happened between the two phases that rendered NDTV helplessly incapable of giving as specific and clear a projection in phase 2 as it did after the first phase? Why did Roy not even try to explain the ‘scientific’ reasons behind what was clearly a mockery of exit polls after the second round, despite the claim of basing the projections on the largest sample size?

Is there any parallel anywhere in the world where a supposedly sophisticated statistical model throws up a variation of 92 percent from the lower end of the seats projected?

If we look back to the UP elections, there too Roy was clearly subjectively influenced by the impact of the Rahul Gandhi factor. There too he kept telling the nation right till the very end that the percentage share of the Congress party had gone up significantly. There too, NDTV had projected that the Congress would win between 35 and 45 seats, more than the projection of all other exit polls. In the event, the Congress won only 22 seats out of 403 and even suffered a drop in vote percentage!

The unexplainable volte face by NDTV in its exit poll after the second phase of the Gujarat elections is deeply disappointing and lends further credence to the widely held belief that the channel is almost a party organ of the Congress. In Tamil Nadu, when you tune into Jaya TV or Sun TV you know in whose political camp you are. Perhaps it is time for national parties to similarly openly ‘own’ their own channels and then propagate their agenda. At least viewers will be spared the sophisticated deception of objectivity that they are being subjected to now.

NDTV’s mockery of exit polls has one more deduction that refuses to go away. The way even the pretence of a fight by the Congress was hastily shed by Roy and Sopariwala after phase 2, shows that even NDTV is convinced that Modi is headed for a much bigger win than what the channel had initially anticipated but concealed. Sopariwala had even given a hint to this effect by saying softly that a two percent variation in the projected vote share either way would lead to a landslide victory for the BJP or the Congress.

Wait till the 23rd.