Gujarat 2022: Why A Recent Pre-Poll Survey Doesn’t Exactly Pass The Smell Test

The results of a pre-poll survey for Gujarat were released by a leading media house this week, soon after Prime Minister Narendra Modi inaugurated the Bharatiya Janata Party’s (BJP) election campaign in the state.

The survey has four principal findings:

● The BJP will win a record 135-143 assembly seats in a house of 182.

● The Congress will be reduced to 36-44 seats, and a historically low vote share of 32 per cent.

● The Aam Aadmi Party (AAP) will surge to 18 per cent, dent the Congress badly by cutting heavily into its votes, but still not win more than 0-2 seats.

● The BJP’s sweep, the Congress’ decline, and the AAP’s surge, will follow roughly-similar patterns in all the four regions of the state — Saurashtra-plus-Kutch, north Gujarat, central Gujarat and south Gujarat.

Here are the survey results along with those of the 2017 Gujarat assembly elections:

And this is a map of the region-wise survey results.

How Representative Is This Survey?

While its seat projections conform broadly with this writer’s calculations and ground reports, the survey’s predicted vote shares do not. There are multiple reasons for this assessment.

First, a decline of the ‘Others’ vote by over half (from 9.4 per cent in 2017 to 4 now), and most of it going to the AAP, is difficult to digest.

This is because the ‘Others’ have consistently polled around, or over, 10 per cent in Gujarat right from the first assembly election.

This has been the case particularly in the past three decades, when the contest in Gujarat became rigidly bipolar between the Congress and the BJP.

Second, a secular swing of nearly 10 per cent votes directly from the Congress to the AAP is not supported by ground reports.

On the contrary, the news reports of the past six months have been of a steady exodus to the BJP by sitting Congress MLAs and party functionaries.

In May 2022, Congress MLA Ashwin Kotwal for Khedbrahma seat (a reserved tribal constituency and a solid Congress hold) resigned from the assembly and joined the BJP.

On 4 October, Harshad Ribadiya, the sitting Congress MLA for Visavadar (another seat where the BJP has performed poorly in the past few elections) resigned from the assembly.

He is expected to join the BJP shortly.

Third, according to the ABP-C-Voter survey, the Congress is set to lose 37 seats to the BJP while losing 9.4 per cent vote share to the AAP.

This doesn’t add up, even if the basic concept does — that the AAP cuts the Congress vote to the BJP’s advantage.

The reason is that a detailed victory margin analysis of the 2017 assembly elections showed that the Congress won 31 seats with a victory margin of under 5 per cent, and a narrow, average win margin of just 2.5 per cent.

Consequently, there is a next-to-nil chance of the Congress losing ‘only’ 37 seats even after its vote share plummets by nearly 10 per cent.

In other words, if the Congress’ vote share reduces by 10 per cent as the survey predicts, it will lose a lot more than 37 seats (all to the BJP).

Equally, if the survey predicts that the Congress will lose 37 seats, then the associated vote share decline (to the AAP) has to be much less. But it can’t be both at the same time.

Four, the survey predicts that the BJP’s vote share will reduce by 2.1 per cent, with most of it going to the AAP.

This is a significant figure by Gujarat standards. Now, we wouldn’t normally dispute such a finding, but there are inherent contradictions.

Readers may remember that in 2017, when the Congress unleashed three caste cards to give the BJP a minor jolt, the BJP’s vote share still went up by 1 per cent.

Many of those factors have either been neutralised, or subsumed into the BJP (like Alpesh Thakore and Hardik Patel).

In addition, and as mentioned earlier, a clear shift of votes and leaders from the Congress to the BJP is actively ongoing at present, meaning that the BJP will benefit.

In that case, how are we expected to agree with a survey which doesn’t reflect this key aspect, and instead, blithely shows votes going from everyone to only the AAP?

Fifth, these numerous incongruities noted in the survey at the state level, and its attendant, inexplicable electoral math, are seen to manifest themselves in the four main regions of Gujarat as well. That raises a number of additional flags.

In Saurashtra and Kutch combined, where the Congress and the BJP were neck-to-neck on vote shares in 2017 (45.5-45.9) and the Congress actually won more seats than the BJP (30-23), the survey predicts a debilitating 13.5 per cent vote swing from the Congress to the AAP, along with a 2 per cent swing from the BJP to the AAP, and nearly 3 per cent from ‘Others’ to the AAP.

Yet, at the same time, it predicts 11-15 Congress wins in this region.

How is that ever going to be possible, when the average victory margin in these 30 Congress wins is only 9.4 per cent?

Obviously then, either the vote swing from the Congress to the AAP is grossly overestimated, or the number of Congress victories is. But it can’t be both at the same time.

It is the same issue in north Gujarat, where, too, the Congress did better than the BJP in 2017, as a result of its caste ploys.

Further, the ‘Others’ vote is predicted to reduce from 10 per cent to just 3 per cent, with all these ‘Others’ votes going obediently to the AAP.

Anyone who knows this area also knows that the probability of the ‘Others’ vote behaving in this manner is close to zero.

These survey contradictions manifest themselves in central Gujarat as well, with the ‘Others’ vote going down to an infinitesimal 2 per cent, and everything to the AAP.

So too in south Gujarat:

It actually beggars incredulity that fully 5 per cent of the ‘Others’ vote is projected to shift to the AAP at a time when, and in precisely that region where, a successful, prominent third player, the Bharatiya Tribal Party (BTP), announced that it will not be entering into an alliance with the AAP.

“No alliance with topi-wallahs’, said the BTP’s maverick leader, Chotu Vasava, making his dislike for the AAP amply clear.

Therefore, in conclusion, while the AAP may make a splash in Gujarat later this year by cutting into the Congress’ vote base, while the BJP will do quite well (perhaps even exceedingly well), and while the Congress will be singed, suffice here to say without belabouring the point any further, that one would be hard-pressed to concur with the findings of this survey.

Two months ago, Swarajya offered a friendly word of caution to mainstream media houses and polling agencies: their surveys would have to be transparent and logical if they wished to be taken seriously, for, if they didn’t, their credibility would be affected, and their efforts would be interpreted as unobjective narrative-building more, and as psephology less.

There is still time left for them to heed these words.

 All data from Election Commission of India website.