The unthinkable is now being predicted. Prime Minister Modi could be on a treacherous wicket in 2019.
The Lokniti-CSDS-ABP Mood of the Nation Survey, published a fortnight back, threw up a faint prospect of the ruling party’s defeat. A few of its topline findings are astonishingly contrarian (click here for more details):
- Modi’s government is about as unpopular right now as the UPA was in July 2013, nine months before its electoral debacle in 2014 – “nearly half (47%) of the total 15,859 respondents are of the opinion that the Modi government does not merit another opportunity”.
- While minorities like Muslims, Christians and Sikhs are overwhelmingly against the government, the majority Hindu community is virtually split down the middle over its support/opposition.
- Over the last 12 months, “BJP’s popularity is down seven percentage points, if this declining trend continues then the ruling party may well dip below the 30 percent mark in the next few months.”
- Congress could “net about one in four votes (25 percent) nationally”; and the erstwhile UPA would secure 31 percent of the votes across the country.
- Remember, this does not include the Congress’s new found allies, viz Mayawati’s Bahujan Samaj Party (BSP), Akhilesh Yadav’s Samajwadi Party (SP) and HD Deve Gowda’s Janata Dal (Secular), which could add another 11 percentage points to the “new UPA’s” tally
To read the story in hindi, click here.
Crucial, Stronger, More Empirical Validation!
I was now very agitated about these findings. I kept on restlessly asking myself: is there anyway we could get an authentic empirical validation, based on real, not polled, data?
Then a brainwave hit me. Since January this year, there have been a large number of bye-elections scattered all over the country. What if I compared this treasure trove of actual voting data with what CSDS’s poll captured?
Fortuitously, the time period had a perfect match, from January through May this year.
And while CSDS’s 16,000 odd respondents were selected through a “scientifically random” process, the bye-election data is “circumstantially random”, since no systematic bias gets created when legislators die or resign.
Here, then, are the contours of our “real world sample”: after the Gujarat Assembly Elections in December 2017, we’ve had 10 parliamentary and 21 assembly bye-elections, spread over 15 states, in which over 1.25 cr people have actually cast their votes for nearly 19 political parties:
For those who delight in numbers, do click through to see the granular worksheet.
Clearly, comparing the two data sets is academically defendable (even if not pristinely consistent).
And I daresay, a highly rewarding exercise, because the results are dramatic. Gaze at this graph for a while – take it all in slowly – now read it again, before moving further:
True Data Is Stranger Than “Sampled Fiction”!
The perfect overlap between the bye-elections’ real voting data and CSDS’ survey percentage is spooky! It gives a whole new meaning to the adage that truth is stranger than fiction:
- BJP+allies got 37 percent of the votes in the survey; they’ve got 36 percent of the actual votes cast (normalised) in the bye-elections since Gujarat Assembly Polls.
- Congress+allies got 31 percent in the survey, and 32 percent of the real votes.
- And while BSP+allies got 10 percent in the CSDS survey, they notched up 13.3 percent of the actual votes polled in the bye-elections.
I daresay that this “cent-percent validation” gave me much more confidence to go back and read the granular predictions again (which I had earlier dismissed as “fanciful”, since these sounded so incredibly against the current media-created hype).
While the following may still be erroneous or turn out to be exaggerated, they certainly enjoy a stronger ring of truth after the bye-elections’ polling data:
- Modi is now only marginally ahead of Rahul in voter support; his 17 percentage points lead has fallen to only 10 percentage points.
- An equal 43 percent like both Modi and Rahul; and since fewer people dislike Rahul, his “net likeability” is actually better than Modi’s.
- Rahul has also managed to convince nearly 30 percent of his “naysayers” into becoming “supporters”; conversely, Modi has converted 35 percent of his earlier supporters into opponents.
- Rahul’s biggest gains have come among middle-aged and elderly voters (those with a higher propensity to go out and vote); Modi’s fall is sharpest among middle and lower class voters.
- Confirming the above trend, Congress is recovering quicker in towns and small cities; and beginning to show early traction in big cities.
- Shockingly, over 60 percent feel that Modi government is corrupt; over 50 percent have heard about Nirav Modi’s scam, and two-thirds of them are dissatisfied with the actions taken, or not taken.
- Congress has staged a remarkable recovery amongst Dalits and Adivasis, nosing ahead of the BJP by 1-2 percentage points
- Farmers are deserting Modi at an alarming rate – a fall of 12 percentage points over one year – and the bulk of these gains are accruing to the non-Congress regional parties.
- Except for the North, Modi has lost support everywhere, most sharply in South, West and Central India.
- GST, the goods and services tax, is becoming an albatross around Modi’s neck, its unpopularity getting worse, from 24 percent to 40 percent (January to May)
- And this one is impossible to fathom: there isn’t a single issue on which the Modi government is rated positively now!
2019 Race Is Wide Open!
While a lot could change once Modi hits the campaign trail and cranks up his legendary skills to spin a narrative of hope for the electorate, a few things are clear:
- Rahul is a push-over only among Modi bhakts (devotees); for all other voters, he is slowly but surely emerging as an option they could pin their faith on.
- Modi has to re-invent himself and his pitch; the jury is out on what that could be.
(At The Quint, we are answerable only to our audience. Play an active role in shaping our journalism by becoming a member. Because the truth is worth it.)