Modi Wave Over? Why Pollsters Are Scaling Down Predictions for BJP

CSDS director scales down prediction for BJP, CVoter says Modi’s popularity is down.

Two of India’s most prominent polling agencies – CSDS and C-Voter – appear to be scaling down their predictions for the BJP, after the first phase of polling that took place on 11 April. And this gives us an indication that the last set of opinion polls which predicted Prime Minister Narendra Modi’s National Democratic Alliance either getting a slender majority or stopping just short of the half-way mark, might have ended up overestimating the BJP.

In an article published in The Asian Age on 13 April titled ‘Is It Disadvantage BJP After First Phase of Polling?’ CSDS director Sanjay Kumar makes the following observations:

• “Of the eight Lok Sabha seats in Uttar Pradesh that went to polls in this state in the first round, the turnout went up only in Ghaziabad and Gautam Buddh Nagar seats where the two Union ministers VK Singh and Mahesh Sharma, are, respectively, contesting.”
• “In the remaining six Lok Sabha seats, that have a sizable proportion of Muslim voters, the turnout dropped compared to the 2014 Lok Sabha elections.”

• “My reading is, the BJP may have an advantage only in these two Lok Sabha seats and might be in trouble in the others.”
• “If that happens, the BJP would be down six in UP in the first round. It had won all these eight Lok Sabha seats in 2014 when elections were held there on 11 April.”

Kumar’s observations involve a both qualitative and quantitative change compared to his earlier prediction for the BJP.

In another article in The Asian Age, written exactly a week before this piece, Kumar had argued that it is “advantage BJP”. Within just a week, “advantage BJP ” has changed into “disadvantage BJP” albeit with a question mark at the end.

In terms of seats, let’s compare Kumar’s prediction to the CSDS’ last survey.

In the survey, released just a few days before the first phase of polling on 11 April, CSDS predicted that the BJP could win 32-40 out of Uttar Pradesh’s 80 seats, that is around 40-50 percent of the seats in UP. But in his article on 13 April, Kumar predicts that BJP could lose six out of eight seats in the first phase. If this trend continues in the subsequent phases, the BJP’s tally could end up with a tally of 20-25 seats.

The Quint asked Dr Sanjay Kumar whether he had indeed scaled down his prediction for the BJP. “My earlier piece (on 6 April) was mostly based on the data from the CSDS survey, in which I had said that BJP is at an advantage. The observations in my second article (dated 13 April) are based on the turnout figures for the first phase,” Kumar said.

“BJP supporters tend to be more enthusiastic than others about voting. On the other hand, the Mahagathbandhan parties have fixed votes and they rely purely on arithmetic not chemistry. So if the turnout has fallen, it could mean that BJP isn’t doing as well as expected in Uttar Pradesh.”

Dr Sanjay Kumar, Director, CSDS, to The Quint

Kumar sticks by his assessment that BJP could be in trouble in six out of eight seats in the first phase in Uttar Pradesh.

Using the turnout data, Kumar extends his assessment of Uttar Pradesh to Bihar and Maharashtra as well.

“The turnout didn’t increase in the seats in Bihar and Maharashtra that voted in the first phase, which indicates that the BJP is facing a challenge. In Bihar, the survey had predicted a one-sided sweep for the BJP. But four seats that voted in the first phase, appear to be witnessing a seat-to-seat fight,” he told The Quint.

Even in the case of the Vidarbha region in Maharashtra, Kumar says that things may not have gone as well as expected for the BJP.

In its survey earlier this month, CSDS had predicted a near clean sweep for the NDA in these two states – 28-34 seats out of 40 seats in Bihar and 38-42 seats out of 48 in Maharashtra.

Interestingly, this is the first time in several years that the CSDS had chosen to make seat predictions. Till now, CSDS gave vote share predictions and other key trends in its rather comprehensive surveys but always desisted from predicting seat-shares, which was done by the Chennai Mathematical Institute.

But it isn’t just the CSDS that is scaling down its prediction for the BJP. Another prominent survey agency – CVoter – in its latest survey, has said that the approval rating of the Narendra Modi government has fallen by 19 points in a little over a month. According to the CVoter-IANS election tracker, the Modi government’s approval rating peaked after the Balakot strike on 26 February.

On 7 March, the Modi government’s approval rating was at 62.06 percent. Despite a minor decrease it remained in the 50s till 22 March. But on 12 April, a day after the first phase of polling, the Modi government’s approval rating had fallen to 43.25 percent, a fall of almost 19 percent in about five weeks.

According to CVoter, the Modi government’s, approval rating is now at “pre-Pulwama” levels.

BJP Could Perform Worse Than What Surveys Predict

So based on Dr Sanjay Kumar’s analysis of the first phase turnout and the C-Voter survey’s findings, a few observations can made:

• The Balakot strike had led to a surge in Modi’s popularity for about a month. Naturally as the field work for most pre-poll surveys was done in March, this surge was reflected in the last set of surveys before the first phase of polling.
• But as the CVoter tracker shows, the Balakot surge has now dissipated and as a result, the Modi factor isn’t working as well as expected.
• Even in states where Modi is by far the most preferred PM choice, such as Uttar Pradesh, Bihar and Maharashtra, his ability to trump factors like alliance arithmetic, caste, candidate selection and local factors appear to have greatly weakened.
• While in Uttar Pradesh, the alliance arithmetic seems to be favouring the Mahagathbandhan, other key states like Bihar and Maharashtra are witnessing what Dr Sanjay Kumar calls a “seat-to-seat” battle. This is bad news for the BJP. It means that despite a reasonably favourable approval rating and alliance arithmetic in these two states, the Opposition could pull away a few extra seats due to local or caste factors.
• The last set of surveys before polling predicted the NDA winning 260-280 seats. The final tally is likely to be much lower than that, if the C-Voter tracker and Dr Sanjay Kumar’s observations are any indication.