I am going to do a little statistically-illiterate kitchen-table handwaving here.
The idea is to estimate the MINIMUM number of seats BJP will win this time, based partially on the seat-wise outcomes of the past 8 LS elections.
I find the graph above very interesting.
The green line is for "OTHERS". This includes ALL non-national parties: non-INC, non-BJP. It could include coalition partners of NDA and UPA, plus Mahagathbandhan type parties, 4th-Front parties, Left parties, etc.
I have plotted it this way to indicate that the green line is an AGGREGATE of ALL voter motivations NOT to vote for one of the two national parties. Including caste, regional, community etc. loyalties taking precedence over anything else.
Of course to some extent it could imply false positives in that people may vote for a particular party in the green category knowing that it is in either NDA or UPA, i.e. they are giving their votes to it because it is a proxy in their constituency for one of the two national parties, such as Shiv Sena in Mumbai or RJD in Bihar. Yet, the very fact that one of the alliances had to incorporate the party in question shows that in the given constituency, sub-national motivations for voting outweigh allegiance to either of the two "national parties".
So I think I am pretty safe saying: the GREEN line is a trend-line proxy for all seats won on the basis of primarily SUB-national political affinity. Be that caste, region, religion, or whatever.
What is interesting is that the extent of seats won by votes cast on the basis of sub-national affinities is ALMOST constant within a given range, from 1989 to present. It fits in a neat box of 197-260 seats.
The greatest variance in seat share is for the two national parties (blue line = INC, orange line = BJP).
Starting from this point, I'm going to make some predictions as to how many votes MINIMUM BJP is likely to win this time. Remember that as I do this, I am making assuming conditions of MAXIMUM generosity to the opposition (INC, Mahagathbandhan etc.) so that I can estimate the lower limit for BJP seats.
1) How many will INC win?
I am going to make the argument that INC's position this time is roughly analogous (more than to any other data point) to BJP's position between 1989-1991. This is the BEST case for INC.
Why do I say this? Because:
A ) In 1991, INC had India's "tallest leader" with the most pan-India appeal: Rajiv Gandhi. Yes, he was tainted by Bofors, but he was also the son of Indira who had died just 7 years before. Today, NaMo is probably more popular than Rajiv, but for this purpose, let us say he is equally popular as Rajiv was before his death at Sriperumbudur.
B ) BJP in 1991 was in a state roughly analogous to INC in 2019. BJP in 1991 had never proved itself a national party, a "natural party of governance". In 2014, INC was firmly knocked off the pedestal of "natural party of governance" by failing to get even a triple-digit seat count.
However, I'm also being extra generous to INC in making the comparison. Why? Because:
A ) In 1989-91, BJP had many leaders of stature who enjoyed the confidence of the public in many parts of India: ABV, LK Advani etc. By contrast, what does INC from 2014-19 have? Pappu.
B ) BJP in 1990 had kick-started the Ram Mandir movement, which was a true mass-mobilization movement in Indian politics. INC between 2014-19 has done nothing like that. All they have is Lootyens Media lies, "Rafale" rhetoric, "Unemployment" claims, and the "NYAY" vote-buying scheme.
OK. Phir bhi, let us say that
the % increase in seats for INC from 2014-2019 will be equal to the % increase in seats for BJP between 1989-1991. Yes, of course I'm being generous here. But remember, this is a
worst-case scenario estimation.
So what do we get? The rise in seat count for BJP from 1989-1991 was 85 to 120, i.e. ~41%. Let us give INC an increase of 41% in the 2019 polls over their seat count in 2014. This grants INC
62 seats in the 2019 election.
2) How many seats will "Others" win?
This is a tougher question. As mentioned previously, Others have held a range of seats between 197 and 260 in all elections from 1989 to present. This indicates there is a fairly consistent fraction of the electorate, distributed constituency-wise, that votes on the grounds of subnational political affinity rather than for one of the two national parties.
Now if one looks at the seat share of "Others", a pattern emerges. When there is a surge AGAINST BOTH national parties, the green line swings higher. When there is a mandate in favour of one national party, only the die-hard supporters vote for the sub-national parties, and the green line swings lower. But the range is pretty much constant for the last 30 years.
So what I've done (to be generous, again) is consider the median of all the seat shares of the non-INC, non-BJP parties over the last 30 years (this is more generous than the arithmetic or geometric mean). It comes to 231.5.
To this figure I have ADDED
one full standard deviation of ~25.
That brings the figure of seats won by non-INC, non-BJP parties to 256. That is very close to the highest they have EVER won: 260.
...So.
With 62 seats to INC (generous figure) and 256 seats to Others (again, generous figure), what are we left with?
225 seats for BJP.
This is the minimum, the bare minimum, that BJP will win in 2019. It is more than enough to form the govt, given that "Others" includes NDA partners as well. Look at UPA 1 (formed with 145 INC seats), and UPA 2 (formed with 206 INC seats). Even the worst-case scenario for BJP in 2019 is better than either of those figures.
So Conclusion #1: Barring catastrophic contigencies, Modi WILL be the PM from 2019-2024. He will in fact be a STRONGER PM than ANY PM from 1996-2014... even Manmohan Singh/Sonia Gandhi ruled India with only 145 seats in 2004 and 206 seats in 2009. And let me remind you, this is the worst case scenario.
Now:
The worst-case scenario ASSUMES that Modi has been no better, no worse than any other PM for the last 30 years. It takes into account that fluctuations in the global economic environment, and their repercussions in India, have remained within a certain range throughout that time period. It is based on the premise that whatever economic/social pain voters have felt in their daily lives throughout the 1989-2014 period is no better and no worse than what they have experienced in the 2014-2019 period.
However, WHAT IF Modi has in fact been not just ordinary, but transformational?
What if the poorest of the poor have never seen anything like this level of improvement in their daily lives within the last 70 years, let alone 30?
What if nearly all classes of Indians realize that they have, in fact, never had it as good as they have today at any point in living memory? Economically and otherwise?
What if that intangible but incalculable sense of pride (at being Indian, and at being a legate of Hindu civilization) instilled by Modi has no equivalent at all in any of the regimes from at least 1964 to the present?
What if at very long last, there is a deep and satisfying sense of gratitude amongst the people of India that a government has served a full five years without any credible involvement in a scam, or a ghotala, or some other stigmatizing blemish of wanton corruption?
If that is the case, the sky is the limit.
There is NO reason why Modi will not come back with 300+ seats (for BJP alone) this time around. Just look at the graph again. The green line remains, curiously enough, pretty steady between 2009-2014. But the slopes of the orange (BJP) and blue (INC) lines are absolutely unprecedented in this period.
People WANT a National Party running the Union Government. They only vote for the Other (Subnational) parties when they feel equally disgusted by both National parties. There is every reason to believe that will not be the case this time around. Pappu will bring at best 20 more seats to the INC tally (and it is quite possible that he will arrive at a score lower than 44). Others may very well hit something close to their low-water-mark of ~180 seats.
Looked at from a historical perspective, it is very much possible that BJP by itself will win 300 or even 320 seats this time around.
