India 2014

[jaichind]

Dhubri seat is being contested by Badruddin Ajmal, the leader of the AIUDF, and the most influential Muslim leader in Assam. Muslim consolidation in favour of INC might hurt AIUDF in rest of Assam, but in Dhubri I can't see Ajmal losing.

Also, what do you make of pre-election data for Assam? This is from CNN-IBN in April.

I simply can't understand how a massive lead for INC in Upper Assam has become a BJP clean sweep. Well, at least this is what I am gathering from recent reports. Another thing is that INC leads BJP strongly in rural areas. Upper Assam is mostly rural. If anything, the BJP should be winning its seats in areas like Guwahati.

As for Bengal, I will be surprised if BJP wins even a single seat. But, then again, there is a first time for everything. If 2014 is a wave election, I suppose the BJP might get lucky and win a seat or two.

You make a very good point about Dhubri.  But I think unlike 2009 the BJP is in fray this time and might change the dynamics of the race to shift votes from AUDF to INC to make sure BJP does not win (of course the BJP has no chance in Dhubri anyway).  Of course the counter trend is even more likely since AUDF is most likely to be seen to be stronger of the two.

As for the CNN-IBN poll, it has AUDF losing a third of its vote from 2009 to INC (I believe this and suspect the bleed might be more then that), but INC not losing any vote share to BJP nor AGP from 2009.  With BJP and AGP running seperately as opposed to 2009 this poll really mean a complete sweep for INC with the result being: 12 INC, 1 BPF, 1 AUDF.  Just like I think the various polls overestimate BJP on other parts of India I think this poll overestimate INC.  But yes, if this poll is true then INC will sweep in Upper Assam as well.  But IF INC did lose ground from 2009 aside from the surge it will get from AUDF vote bloc going to INC in Lower Assam, then that loss is more likely in Upper Assam where AUDF was weak anyway and tactical voting by AUDF supporters will not help INC.  

Speaking of polls it is also interesting that NDTV last pre-election poll had INC at 9 and BJP at 5 in Assam. It seems NDTV also made the the assumption I did about Dhubri although now you are point this out I am re-thinking this one.  On top of this this prediction makes no sense as in Kokrajhar neither INC nor BJP are running (although BJP is supporting an independent and INC is support BPF)  So the INC+BJP seat count cannot be 14.  Unless NDTV counts the BJP supported independent as BJP feel that he will win ? I think BPF will win Kokrajhar.

For WB, it will not surprise me at all if BJP, after all the huffing and puffing, end of with zero. I just figure they might get lucky and win one.  Of course the bigger effect of the BJP surge in WB is that is upsetting the expected TMC landslide.    If BJP stayed weak, then other some INC and Left Front bastions, I expected TMC to sweep the rest.  But with BJP running much stronger it is taking way from TMC votes which is forcing TMC to run an anti-Modi campaign to consolidate the Muslim vote.  As it is the BJP surge is helping Left Front but if the BJP surge is even stronger than expected then it might actually pull the anti-TMC vote which might otherwise go to Left Front and win a seat or two.

[jaichind]

Speaking of  Kokrajhar.  There was a bunch of post-election violence in  Kokrajhar where it seems some Bodo extremest went on a killing rampage of Muslims in this region.  I am not so sure but I suspect it has to do with the fact that these Muslims voted for the TMC candidate or the BPF candidate and not "the Bodo candidate" such as the ex-BPF candidate running as an independent who is the Bodo nationalist candidate and backed by the psudo-extremist ABSU.  BPF nominated a more inclusive candidate and this might have created tensions in the Bodoland movement as BPF is suppose to be THE BODO party.   

[jaichind]

Since the Indian stock market surge has been based on an assumption of a Modi government, a lot of investment banks are getting into the act in terms of election prediction.

Nomura Holdings sees the medium case as 220 seats for NDA, the 25 percentile case for NDA is 200 seats .

IIFL Holdings expects 225 seats for NDA which matches by current prediction exactly.

Credit Suisse expects 220-230 seats for NDA and says their survey of institutional investors shows average expectations of 233 seats for NDA.

The following brokers indicated what the Indian stock market reaction is going to be based on the number of seats captured by NDA

Lok Sabha Seats           Sell off below   Rally above
Brokers
HDFC  Securities                   220           250
Kotak Securities                   250            --
Motilal Oswal                        230           255
KR Choksey Securities          260           290
IIFL Holdings                         200           230
Geojit BNP Paribas                230           250
SMC Capital                          200           240
Prabhudas  Lilladher            240           273
Magnum Broking                   220           260
Maximus Securities               240           275
Dalton Capital                      240           260
Augment Financial Serv        240           280
KRIS Ltd                               215           280
Equinomics Research           200           250
Sumedha Fiscal                    240           300
CNI Research                      255           280
Pratibhuti Viniyog                230           270
Alpha Ideas                        230           280
Kitna Paise                          230           260

AVERAGE                             230           266

I guess one can then derive that the average broker reading of market expectation should be the average of the sell below and rally which would be (230+266)/2 = 248 which pretty much matches the average pre-poll surveys for the number of seats for NDA.

So the various investment communities seems think the result will be anywhere between 220 to 250 for NDA.  If exit polls comes out in the lower end of that range then my opinion this that NDA will get less then that as it shows a shift in momentum from before the election in which the exit polls might not fully capture. 

[jaichind]

Exit polls should be coming in 2 days right?

5/12 9am EST

[jaichind]

I have been at a conference all day today but was able to look at some of the exit poll results now.  It seems that most of them converge around 270-280 for NDA.  If so it seems that the momentum is on the side of NDA.  Some of the state levels projections seem very problematic to me but I think many of them cancel each other out across all these different exit polls. 

A summery would be



Times Now state level results make the least sense to me and Today's Chanakya results also make no sense.  In terms of vote share Today's Chanakya is similar to the others but just has a very high NDA seat conversion ratio which makes no sense to me.  Throwing these two out we have NDA at around 270-280 or so for NDA which is higher than the 250 or so pre-election survey. 

My prediction of 225 for NDA was based on exit polls averaging around 240-250 for NDA but with exit polls converging at a higher number it seems 1) the pre-election polls has been confirmed and 2) the NDA has momentum on its side.  It would not surprise me if the NDA will come in at higher than 280.  The only word of caution is I suspect these exit polls still does not  capture anti-Modi tactical voting in in which case NDA will come in more around 250.  Either way we are looking at a Modi government.

[jaichind]

I will try to recalibrate my predictions based on these exit polls and re-post them later this week.

[jaichind]

INC internal assessment seems to be that INC might capture as much as 130 seats but that might be on the high side.  I guess they feel it is more likely it would be around 110 if 130 is considered high.  If so that would be consistent with NDA around 240-260 or so.  Sensing defeat the INC is trying to deflect blame on Rahul Gandhi by saying that the party leadership will have to take collective responsibility for the results. 

[jaichind]

Of course the wide variation at the state level between the different exit polls does lead to some doubt on how accurate they are.  For example, if I cherry picked the most pro-UPA result for each state amoung the different exit polls and reconstructed the most pro-UPA exit poll we get

                        NDA         UPA
Maharastra       29           19      Times Now
AP                     12             7      C Voter
WB                     1              5      AC-Nielsen
Bihar                 20           15      AC-Nielsen
TN                      0              1      Times Now
MP                    18             9       Times Now
Karnataka          8            17       Cicero
Gujarat            22              4       Times Now
Rajasthan        11            14       Times Now
Orissa               2               5       Times Now
Assam               1             10       CNN-IBN
Jharkhand         7               6       Times Now
Punjab              5               7       AC-Nielsen
Chattisgrah      7               4       Times Now
Harayana          3               7      Times Now
Delhi                 5               2       Times Now
JK                     1               5       Times Now
Uttarkhand       1               4       Times Now
Goa                  1                1       Times Now
HP                    1               3        Times Now
UP                   46              8        AC-Nielsen
Kerela              0              17       Times Now
NE                    3               5        Times Now
Federal            0               6       CNN-IBN
--------------------------------------------
                    204            181

Now, the exit poll results which are the most pro-UPA for MP and Rajasthan are from Times Now and are highly suspicious, especially Rajasthan.  Times Now claim that in both states AAP will take 10% of the vote and cut into the anti-INC pro-BJP vote share.  This is very dubious since both states are not very urban and for AAP to win 10% of the vote they will have to sweep all urban areas which is very unlikely.  For Rajasthan, Times Now did say that if they took a pro-BJP view of vote share within the margin of error it would be NDA 17 UPA 8 instead of NDA 11 UPA 14.  Even if we did that  change our cherry picking model of exit polls would be NDA 210 UPA 175.   So the margin of error for these exit polls seems to be quite large.  So while it is unlikely NDA would not cross 250 seats this Friday there exist significant statistical change even by these exit polls that NDA could end up with 230 or even 220.

[jaichind]

Seems like INC is not the only one trying to lower expectations.  It is rumored that RSS, the parent body of BJP, leaked out an internal assessment of the election and projected 220 for NDA.

[jaichind]

One of the more comical exit polls is News24 Todays Chanakya UP

It has

Vote Share
BJP+          34%
BSP           22%
SP             23%
INC+         12%
AAP            5%

Seat Share
BJP+          70
BSP             3
SP               4
INC             3

In theory if the relative strengths of these parties were uniform across UP then these vote share could translate into these seats shares.  But in fact the relative strength of SP BSP INC are quite scattered and non-uniform across UP in a significant way.  The only way these vote shares translate into these seat shares if there are pro-BJP minority tactical voting where anti-BJP votes to avoid the strongest challenger to BJP to ensure the victory for BJP.  This proposition of course is absurd based on intention and ability of the anti-BJP bloc.

[jaichind]

So there won't be a gradual release of the results, just one huge announcement on Friday?

Nope.  Friday morning they will count votes and results released as the count progresses.

[jaichind]

While it is not clear the scale of the NDA victory, it is clear that UPA is defeated which has been clear for a couple of years now.  INC last day or two seems to be gearing up to lay blame on PM M Singh for defeat.  INC leaders are talking about how Rahul Gandhi was not part of the cabinet which in turn infer that those in the cabinet should take the blame for defeat.

[jaichind]

While it is not clear the scale of the NDA victory, it is clear that UPA is defeated which has been clear for a couple of years now.  INC last day or two seems to be gearing up to lay blame on PM M Singh for defeat.  INC leaders are talking about how Rahul Gandhi was not part of the cabinet which in turn infer that those in the cabinet should take the blame for defeat.

If there is anyone who should get blame in the INC, it should be Rahul and to a lesser extent his mother.

Sonia Gandhi came out and said that M Singh cannot be blamed for defeat if INC were defeated.  She pushed for collective responsibility.  For me defeat was a function of of the UPA-IIt track record as well as an election campaign that lacked clarity. Rahul wanted to run a grassroots campaign and Sonia wanted to run a tried and tested traditional INC campaign.  INC ended doing neither.

[jaichind]

Correction.  The RSS exit polls has been officially released.   It has BJP at 226 and NDA at 259.  This seems to be somewhat lower than the medium exit poll which has NDA at 270-280.

Chart below has RSS projection for BJP seats by state



RSS also expect TDP at 8 seats, SS at 11, SAD at 5.

[jaichind]

INC internal survey suggest a quixotic 193 seats for UPA and 161 for INC.  To be fair back in 2009 the same survey had INC at 180 and UPA at 235 and it ended up being INC 206 UPA 261.

UPA breakdown would be


Uttar Pradesh     15
Bihar                   15
Madhya Pradesh 11
Andhra Pradesh  12
Gujarat               10
Maharashtra       24
Karnataka          14
Kerala                 12
Punjab                10
Rajasthan             8
Delhi                     2
Tamil Nadu           1
Assam                10
North East           7
Uttrakhand          3
HP                        3
J&K                      4
Goa                      1
Haryana`             4
Jharkhand            4
Chattisgarh          4
West Bengal        6
Orrisa                  7
Union Territories  6
Total                193

What is interesting about this survey is that INC seems to have given up on Jharkhand only projecting 4 and accepting a disappointing result in Karnataka at 14. I find it hard to believe that UPA can win 193 seats if they only win 14 in Karnataka.  The rest of the numbers does seem to be around the UPA cap in each state if everything comes up trumps for them.  

[jaichind]

NDTV exit poll has NDA 279 UPA 103.  NDTV seems pretty consistent with CNN-IBM at 276, Cicero 272, C-Voter 289 and Nielson 281.  Times Now at 249 and Today's Chanakya at 340 seems to be the outliers.  I am a bit skeptical of NDTV's results for TN where they have AIADMK at 48%, DMK+ at 25%, DMDK-BJP+ at 14% and INC at 3%.  The gap between AIADMK and DMK+ seems to large to me.  The BJP+ vote share of 40% in UP also seems high but is not too out of line with other exit polls.