2019 India April–May LS general elections and assembly elections of 2019

[jaichind]

Looks like ECI have canceled elections in the TN district of Vellore.  It seems that a government raid revealed a warehouse controlled by the DMK candidate of $1.6 million dollars worth of cash which is to be used for vote buying.  DMK is crying foul that they are being framed by authorities which are focused on helping AIADMK-BJP.     

The race in Vellore is AIADMK splinter PNK for NDA vs DMK for UPA.  My current model has this race neck-to-neck with a tiny lead for PNK.  So I can see why vote buying would make sense here as it could make a difference.

https://www.ndtv.com/india-news/lok-sabha-polls-in-tamil-nadus-vellore-cancelled-after-huge-cash-haul-2024089

[jaichind]

On the subject of TN when all of TN votes on Thursday as part of Phase II voting there will be 22 by-elections for TN assembly which could determine if the AIADMK government would fall or not.

The 2016 TN assembly election produced:

AIADMK   136

DMK          89
INC            8
MUL           1

AMMK won the by-election for Jayalalitha's seat and the pro-TTV Dinakaran faction of AIADMK defected to AMMK and were disqualified by the AIADMK speaker.  So now we have

AIADMK    114

AMMK          1

DMK           88
INC              8
MUL             1
 
with 22 by-elections needed.  If AIADMK does not win at least 4-5 out of the 22 by-elections it will lose its majority and most likely early TN assembly elections are needed.  In many ways this elections is more important that the TN LS elections.   NDTV reports that the going rate for vote buying for the LS elections is around $50 per vote but the going rate for assembly by-elections are around $300 per vote.   These 22 by-elections will be a 3 way battles between AIADMK vs DMK vs AMMK with what I have to assume as the DMK with the edge given the old AIADMK base is split down the middle between AIADMK and AMMK.

[jaichind]

With APP-INC talks for Delhi alliance in progress there are more media analysis that claims that AAP-INC alliance will help in Delhi but not Haryana based on math.  I still think given the history of hostility between INC and AAP in Delhi that any alliance will be hard to work on the ground while in Haryana that is not the case and an opposition alliance can take advantage of anti-incumbency there against the BJP state government.     

https://www.hindustantimes.com/lok-sabha-elections/aap-congress-coalition-likely-to-fare-well-in-delhi-than-haryana-suggests-data/story-LR6EmKdb6YQxST2rTUvYuN.html

[jaichind]

This news18 analysis of the community backgrounds of the SP, BSP, and INC candidates in UP pretty much came to the conclusion I did a while ago, that INC and SP have a de facto understanding to help each other cut into BJP votes while it is all out war between INC and BSP.

https://www.news18.com/news/politics/eye-on-post-poll-scenarios-congress-and-sp-pick-candidates-that-minimise-mutual-damage-2104317.html

Once I looked over the INC candidate list (so far) for UP I can sort of map out what the INC strategy in UP might be.  Out of the 80 UP seats the INC have solid candidates only in around 10-15 seats with another 10 that might gain some significant vote share.  The rest are either withdrawals for SP-BSP-RLD VIP candidates or just dummy candidates.  What the INC is mostly doing is to run strong Muslim candidates in heavy Muslim areas that BSP is running to cut into the BSP vote while mostly leaving SP alone.  Even when INC runs strong candidates against SP the community of the INC candidate makes it clear that INC is trying to cut into both BJP and SP candidates while a strong INC candidate running in BSP seats seems to have community makeup to cut into only BSP candidates.    Even when INC runs dummy candidates you can tell they are trying to cut into the BJP vote when a SP candidate is in the fray. 

So the INC strategy in UP seems clear: help the SP but hurt BSP.  The main reason for this is SP is local to UP but the BSP Dalit base in other Hindi heartland states are part of the vote share the INC needs to win back to be competitive with the BJP on the long run so trying to destroy BSP is part of the mid-term INC strategy.

[jaichind]

Now it is the INC's turn to threaten voters.  INC Chhattisgarh minister tells voters that if they vote for anyone else other than INC the EVM will give them an electric shock.

https://scroll.in/latest/920390/voters-will-suffer-electric-shock-if-they-press-second-third-button-on-evm-chhattisgarh-minister

[jaichind]

Voting for phase 2 almost done.  Phase 2 includes all of TN.  Money power seems to be a lot stronger in TN than previous elections.   In 2014 around $2.3 in cash was sized by authorities in TN that was related to vote buying.  In 2019 that number have surged to over $29 million.

[jaichind]

Prelim ECI figures on turnout.  Most likely will be adjusted upward by a few percentages.   If so turnout will mostly mimic 2014 with a small drop across the board.

[jaichind]

Example of cash seized by ECI.

[jaichind]

Turnout in Srinagar in Kashmir will most likely be around 17% which is higher than the 7% in the 2017 by-election but still low.  The level of militancy which have been dropping in the 1995-2010 period is rising again and this type of turnout is reflective of the anger as well as the relative safety given threats from radical separatist groups. 

[jaichind]

I have been reading up on defections this LS election and found the one in Ahmednagar in Maharashtra quite interesting and  at the same time gives us some clues on how things are going. 

The situation in Ahmednagar is Sujay Vikhe Patil, the son of the INC leader of opposition in the state assembly Radhakrishna Vikhe Patil,  defected to BJP and was given the BJP ticket to run.  The BJP dropped the incumbent Dilip Kumar Gandhi who won here in 1999 2009 and 2014 whose son will be running as a BJP rebel.    Radhakrishna Vikhe Patil the father of the INC defector seems to insist that he will stay in INC but was seen campaigning for his son who is now running on the BJP ticket.

https://www.ndtv.com/india-news/lok-sabha-elections-2019-congress-red-faced-maharashtra-leader-campaigns-for-son-who-joined-bjp-2023895

The background of this is interesting.  INC leader of the opposition in the state assembly Radhakrishna Vikhe Patil is the son of Balasaheb Vikhe Patil who is a long time INC leader who passed away in 2016.  Balasaheb Vikhe Patil was a key rival to Sharad Pawar (who went on to form NCP in 1999) inside the INC.  When Sharad Pawar was giving the power to run the INC campaign in Maharashtra in 1998,  Balasaheb Vikhe Patil defected to SHS who fielded him in Ahmednagar and defeated the INC candidate despite a INC sweep in the rest of Maharashtra.  In 1999 Sharad Pawar split from INC and formed NCP.     Balasaheb Vikhe Patil stayed in SHS given the national wave seems to be on the BJP side in 1999.  The Ahmednagar seat was given by SHS to BJP so Balasaheb Vikhe Patil ran for SHS in another seat and won.  But soon after that Balasaheb Vikhe Patil defected back to INC running for in this nearby seat and winning in 2004.  As for  Ahmednagar Dilip Kumar Gandhi ran and won in 1999 with the INC and NCP candidate splitting the anti-BJP vote.   In 2004 INC-NCP formed an alliance and the seat was given to NCP and BJP dropped Dilip Kumar Gandhi which resulted in the victory of NCP in 2004.  In 2009 Dilip Kumar Gandhi came back to run for BJP and defeated the NCP candidate due to the presence of an INC rebel.  In 2014 Dilip Kumar Gandhi won in a landslide over NCP on the Modi wave.

Now in 2019 Balasaheb Vikhe Patil  has passed away and his son Radhakrishna Vikhe Patil is the INC leader of the opposition in the state assembly.  Radhakrishna Vikhe Patil wanted Ahmednagar to be given back to INC by NCP who the Patil clan view as a rival given the Patil-Pawar rivalry going back decades so his own son Sujay Vikhe Patil can run here.  INC-NCP talks led to NCP retaining the seat so    Sujay Vikhe Patil defected to BJP who gave him the ticket to run for BJP and dropping Dilipkumar Mansukhlal Gandhi along the way. Dilipkumar Mansukhlal Gandhi's son jumped in as a BJP rebel.

The Patil family is called a "sugar baron" since they control all the sugar mills in the area and all the sugar co-ops and workers are dependent on them economically so for the BJP the Patil clan is a nice catch.  The question is, why would the BJP spend the political capital to bring on board the Patil clan when in 2014 the result was

BJP    57.0%
NCP   37.3%

The BJP dropped an incumbent who won by 20% last election to import an rebel from INC and along the way provoked an internal BJP rebellion.  It logically does not make sense.  The logical expatiation would be a combination of BJP's desire to get the Patil clan on-board which would be useful later in the year in state assembly elections and the fact that BJP-SHS, at least in this area, is nowhere as strong as 2014 as the Modi wave recedes so this is a risk worth taking.  If so then at least in central Maharashtra the BJP-SHS might be in more trouble than recent polling would suggest. 

[jaichind]

In UP, a Dalit BSP supporter voted for BJP by mistake and chopped his finger off in anger and then put this news on social media.

https://www.indiatoday.in/elections/lok-sabha-2019/story/dalit-youth-bsp-supporter-chops-finger-vote-bjp-elections-mistake-1505330-2019-04-18

[jaichind]

My initial reactions to the turnout figures are

1) TN turnout seems to have stayed at historic highs which in the current context is pretty negative for AIADMK-BJP combine.
2) UP an Bihar turnout might fall somewhat which I guess is receding of the Modi wave the clearly took place in 2014.  What is interesting is that in Muslim areas in Bihar turnout if anything is higher than 2014.  Of course this narrative might be reversed if more up-to-date figures show the turnout falloff ends up being very small.
3) Karnataka turnout seems to have dropped off significantly.    Either this is a falloff in the Modi wave or it is INC and JD(S) voters not voting due to anger of the INC-JD(S) alliance which does not seem to be working on the ground.  Most likely it is both so I think these two effects might be canceling each other out.

[jaichind]

The AAP-INC alliance talks collapsed again.  It seems once again it is over AAP's desire to link it to an alliance in Haryana.  The back an forth seems to be:

INC: Lets go for an alliance in Delhi where it is AAP 4 seats INC 3 seats
AAP:  Either it is AAP 5 seats INC 2 seats in Delhi OR AAP 4 seats INC 3 seats in Delhi IF there is also an INC-AAP alliance in Haryana
INC: AAP 5 seats INC 2 seats in Delhi is not acceptable.  What sort of alliance do you propose in Haryana
AAP: In Haryana we already formed an alliance with JJP where it is JJP 7 seats AAP 3 seats.  Just for you we are willing to accept INC 6 seats JJG 3 seats AAP 1 seat.
INC: No, in Haryana the best we can do is INC 7 seat JJP 2 seats AAP 1 seat
AAP: There is no way we can get JJP to accept that
INC: OK, so I guess the alliance talks are coming to an end.


Still with 3-4 more days before candidates are finalized I suspect there will be one more round of brinkmanship before this alliance is truly dead.

[jaichind]

One thing that is interesting about the on again off again AAP-INC alliance talks are that the BJP now is constantly delaying coming out with its own candidate list.  BJP ideally want to know if the AAP-INC alliance is on before figuring out which candidates to run in Delhi.  All 7 Delhi incumbents are from BJP and for them this is torture since they cannot really get going to spend the money and resources to mobilize for the election.  I am not if that is a feature or a bug of the AAP-INC on again and off again alliance talks but in Delhi candidate selection has become frozen for the BJP.  To some extend this sort of proves that the BJP bravado that BJP will win all 7 Delhi seats no matter what is not really true although I personally think the BJP will have the edge in Delhi one way or another.

[jaichind]

A first since 1995.  Mayawati and Mulayam Singh Yadav, both former UP CM and bitter enemies for decades, now finally share the same stage.  SP-BSP formed an alliance in 1993 to stop BJP and ruled UP in 1995 when they had a massive falling out.  Now they are together again to stop BJP.

[jaichind]

In a dramatic turn a couple days ago the BJP have nominated Pragya Singh Thakur for Bhopal in MP against former INC CM Digvijaya Singh.  Pragya Singh Thakur was a student leader and a what can best described as a Hindu nun who is the main accused in a 2008 terrorist attack in a bombing of a mosque.  Her trail in in progress and she is out on bail. Since she is not convicted, yet, she is eligible to contest.


It seems the BJP is hoping this will provoke a large emotional response from Muslims to help BJP to counter-polarize Hindu votes.  This seems to indicate voting in the first couple of rounds is not going as well as the BJP have hoped and stronger medicine is needed to consolidate the Hindu vote.

[jaichind]

Latest covter survey shows that the salience of economic issues continues to rise while national security continues to decline after the peak reached during the terror attack plus the response.


Similar trend for approval of Modi/BJP


Breakdown by state shows a familiar pattern where national security has greater salience in states the BJP is expected to do well

[jaichind]

NDTV analysis of Karnataka

Points out that in the 14 seats that have voted in Karnataka turnout have fallen


Impact of INC-JD(S) alliance is fairly minor since JD(S) is weak in Northern Karnataka and have overlapping strength with INC in Southern Karnataka



Which happens to match poll of polls


Community breakdown in Karnataka


Gives info on how communities vote



Impact of swings fairly large in Karnataka


BJP stronger with men


BJP stronger with youth


BJP stronger in urban areas

[jaichind]

The alliances in J&K are fairly complicated.  Due to the trauma of the end of the BJP-PDP alliance ancient rivals JKN and PDP are now ready to cooperate in limited ways now. 

In Kashmir it is BJP vs INC where JKN and PDP are backing INC.
In Jammu it is PDP vs JKN where INC are tactically backing JKN.  BJP is in the fray but as a bit player.
In Ladakh it is INC vs BJP vs ISK where JKN and PDP are both backing ISK in order to consolidate the Muslim vote while INC and BJP fight over the Buddhist vote.
 

[jaichind]

Phase III done.  Turnout numbers by state are

Chhattisgarh: 68.62%
Karnataka: 65.29%
Gujarat: 61.71%
Goa: 71.34%
Kerala: 76.35%
West Bengal: 79.36%
Jammu and Kashmir: 12.86%
Assam: 79.16%
Bihar: 59.97%
Maharashtra: 57.53%
Tripura: 78.67%
Odisha: 58.18%
Uttar Pradesh: 58.91%
Dadra and Nagar Haveli: 71.43%
Daman and Diu: 65.34%

Turnout roughly the same as 2014.  Gujarat turnout down a couple of percentages points.  Most likely negative news, even if slightly, for BJP.  Odisha turnout down significantly.  Not clear if this is due to BJD base not turning out of a decline of the Modi wave. 

[jaichind]

Updated projection from me now that we are half way done with voting.  Mostly the same as before with some minor adjustments.

My basic approach is to view the election as a confederation of state elections.  The main driver being state level anti-incumbency.   If a state just held an assembly election then I tend to think polls underestimate the incumbent party,  if the state held assembly elections 1-2 years ago then I think polls tend to be par.  If the state had assembly elections 4-5 years ago then I tend to think polll overestimate the incumbent party. 

Based on that principle I created separate models every state to generate swings and also take into account key changes in candidates and alliance dynamics.   I did not take into account rebels yet so the vote share calculations are mostly overestimating the dominate parties.

J&K: I decided to break up my projection by each region (Kashmir, Ladakh, Jammu)

Kashmir:
Very low turnout so who knows what will take place.  But generally slight edge for JKN

2019
NDA:       0 BJP          3.7%
UPA:        0 INC         2.2% (tactically alliance with JKN)
Others     2 JKN        41.2% (tactical alliance with INC)
               1 PDP       34.3%
               0 JPC       12.5% (Separatist)

2014
NDA:      0 (0 BJP)                   1.4%
UPA:       0 (0 JKN)                34.9%
Others:   3 PDP                     46.9%
              0 JPC                       6.3%


Ladakh:
4 way battle with INC winning as it gains Buddhist vote from BJP

NDA:       0 BJP          23.6%
UPA:        1 INC         33.2% 
Others     0 ISK          28.1% (backed by JKN and PDP) 
               1 INC rebel  13.1%

2014
NDA:      1 (1 BJP)                 26.6%
UPA:       0 (0 INC)                22.6%
Others:   0 ISK                      24,2%
              0 INC rebel              26.6%


Jammu:
Hindu consolidation should see BJP through

2019
NDA:       2 BJP        46.3%
UPA:        0 INC       41.3% (backed by JKN and PDP)
Others     0 BJP rebel  5.0%
 
2014
NDA:      2 (2 BJP)                 48.6%
UPA:       0 (0 INC)                34.6%
              0 PDP                      8.7%



HP:
Mostly replicates 2017 assembly elections, high number of Upper Caste voters gives BJP a sweep.

2019
NDA: 4 BJP     53.7%
UPA: 0 INC     43.4%

2014
NDA: 4 BJP      53.9%
UPA: 0 INC      41.1%


Haryana:
INLD split with splinter JPP helps both BJP and INC.  BJP splinter LSP does not make that much of an impact other than in the seat of the LSP leader.

2019
NDA:      7 BJP              39.7%
UPA:      3 INC              28.4%
Others:  0  BSP-LSP        7.3%
             0 INLD              6.3%
             0 JJP-AAP        15.4%

2014:
NDA:     7 (7 BJP 0 HJC)        40.9%
UPA:     1 INC                       23.0%
Others: 2 INLD                     24.4%
            0 BSP                        4.6%
            0 AAP                        4.3%


Punjab:
AAP has lost a lot of ground since 2017 assembly elections.  They could have gone to SAD to hurt INC.  But most of that support seems to gone to AAP splinter PEP which split the anti-INC vote to help INC.

2019
NDA:     4 (4 SAD, 0 BJP)     34.1%
UPA:     9 INC                     38.5%
Others: 0 BSP-PEP-LIF         12.0%
            0 AAP                     11.8%

2014:
NDA:     6 (4 SAD, 2 BJP)  35.2%
UPA:     3 INC                  33.2%
Others: 4 AAP                  24.5%
            0 BSP                   1.9%


Delhi
AAP landslide victory of 2015 assembly elections is now a distant memory with AAP administration losing a lot of credibility.    AAP loses support  back to INC which means BJP sweeps the polls again

2019
NDA:    7 BJP     42.4%
UPA:     0 INC    25.2%
Others: 0 AAP    27.6%

2014:
NDA:    7 BJP   46.6%
UPA:     0 INC  15.2%
Others: 0 AAP  33.1%


Uttarakhand
Results will mostly mimic 2017 assembly election with BJP losing some ground due to anti-incumbency which is partly mitigated by Modi's popularity in the Hindu heartland 

2019
NDA:     4 BJP           47.7%
UPA:     1 INC           43.5%
Others  0 BSP             5.5%

2014:
NDA:    5 BJP    55.9%
UPA:     0 INC    34.4%
Others: 0 BSP     4.8%


UP:
SP-BSP-RLD Grand Alliance consolidates backward vote to beat back BJP.  INC resurgence in some seats and SP splinter PSP cuts into Grand Alliance votes and prevents even greater BJP losses.

2019
NDA:    33 (31 BJP 1 AD(S))                            40.4%
UPA:      2 INC-JAP-AD (2 INC)                        11.1%
Others: 45 SP-BSP-RLD(23 SP 20 BSP 2 RLD )  42.2%
             0 PSP-PECP                                        2.4%

2014
NDA:    73 (71 BJP 2 AD)   43.6%
UPA:       2 INC                   8.5%
Others:   5 SP                   22.3%
              0 BSP                 19.8%


Bihar:
Nitish Kumar's popularity pulls NDA ahead despite anti-incumbency.  CPI(ML) has tactical alliance with UPA.
 
2019
NDA:   28 (15 BJP 8 JD(U) 5 LJP)                      48.3%
UPA:    12 (5 RJD 5 INC 1 RLSP 1 HAM 0 VIP)    39.6%
Others   0 CPI(ML)-CPM-CPI                               3.2%

2014
NDA:   31 (22 BJP 6 LJP 3 RLSP)         39.5%
UPA:     7 (4 RJD 2 INC 1 NCP)           30.3%
Others: 2 JD(U)-CPI (2 JD(U) 0 CPI)   17.2%


MP
Modi wave eats into the INC advantage from the 2018 assembly election victory
 
2019
NDA:  19 BJP              48.2%
UPA:   10 INC             43.8%
Others 0 SP-BSP-GGP   4.7%

2014
NDA:  27 BJP     54.8%
UPA:    2 INC     35.4%


Rajasthan
Modi wave eats into the INC advantage from the 2018 assembly election victory

2019
NDA: 18 BJP (17 BJP 1 RLP)     50.3%
UPA:   7 INC                           44.9%

2014
NDA: 25 BJP    55.6%
UPA:   0 INC    30.7%


Chhattisgarh
Tribal votes seems to be shifting away from BJP which adds to the INC edge after the 2018 assembly election victory which will contain the Modi wave in Upper Caste areas.

2019
NDA:    3 BJP          41.7%
UPA:     8 INC         47.5%
Others: 0 BSP-JCC    5.3%

2014
NDA:  10 BJP    49.7%
UPA:    1 INC    39.1%


Jharkhand
INC-JMM-JVM  alliance will create a base too large for BJP to beat even with AJSU alliance, especially given anti-incumbency building up.  In the end RJD had a tactical alliance with UPA

2019
NDA:    3 (5 BJP 0 AJSU)                      40.4%
UPA:   11 (4 JMM 4 INC 2 JVM )            45.0%
Others: 1 RJD                                       3.3%

2014
NDA:  12 BJP                            40.7%                 
UPA:    2 (2 JMM 0 INC 0 RJD)    24.6%
Others:0 JVM                            12.3%
           0 AJSU                            3.8%


Gujarat
INC make s a slight comeback but the Modi wave is still powerful in his home state.

2019
NDA:  21 BJP        52.9%
UPA:    5 INC        41.8%

2014
NDA:  26 BJP                       60.1%
UPA:    0 (0 INC 0 NCP)        34.4%


Maharashtra
BJP got SHS to come aboard to an alliance which might not work on the ground.  INC and NCP deeply divided but does have de facto NMS support.  Dalit-Muslim BVA will split the INC-NCP vote and help give NDA a partial victory despite anti-incumbency.

2019
NDA:       33 (19 BJP 14 SHS)                             47.5%
UPA:        15 (5 INC 8 NCP 1 SWP 1 BVA 0 YSP)   42.0%
Others:     0 VBA                                                 3.9%

2014
NDA:    42 (23 BJP 18 SHS 1 SWP)         51.6%
UPA:      6 (2 INC 4 NCP)                       35.0%
Others:  0 NMS                                       1.5%


Goa
BJP government in turmoil should give INC a chance

2019
NDA:  1 BJP     49.9%
UPA:   1 INC    43.8%

2014
NDA: 2 BJP    54.1%
UPA:  0 INC   37.0%


WB:
Left Front implosion and lack of INC-Left Front tactical alliance will doom Left Front.  BJP surges as it captures the anti-incumbency vote.

2019
NDA:      12 BJP             34.3%
UPA:         3 INC              7.8%
Others:   27 AITC           39.7%
               0  Left Front    15.4%

2014
NDA:      2 BJP           17.0%
UPA        4 INC            9.7%
Others: 34 AITC         39.8%
             2 Left Front   30.1%


Orissa:
INC imploding and anti-incumbency votes going to BJP leads to a shock BJP narrow victory here. 

2019
NDA:    13 BJP                         40.1%
UPA:      0 INC-JMM-CPI-CPM    20.2%
Others:  8 BJD                         36.7%

2014
NDA:     1 BJP     21.9%
UPA       0 INC    26.4%
Others: 20 BJD   44.8%


Karnataka
INC-JD(S) alliance will not work well on the ground letting BJP fight INC-JD(S) to a tie.

2019
NDA:        14 (14 BJP 0 BJP Ind)  47.1%
UPA:        14 (11 INC 4 JD(S))     46.2%
Otherss:     0 UPP                        3.0%

2014
NDA:   17 BJP      43.4%
UPA:     9 INC      41.2%
Others: 2 JD(S)   11.2%


AP
5 way battle and JSP splitting the anti-incumbency vote  allows TDP to avoid a total wipe-out. YSRCP most likely will back BJP post-election.  YSRCP alliance with TRS also drives some anti-Telagana votes to TDP.

2019
NDA:       0 BJP                       5.0%
UPA:        0 INC                      4.2%
Others:  20 YSRCP                 41.1%
              5 TDP                     35.2%
              0 JSP-BSP               10.2% (tactical alliance with Left)
              0 CPI-CPM-CPM(ML)   2.2% (tactical alliance with JSP-BSP)

2014
NDA:    17 (15 TDP 2 BJP)    48.0%
UPA:       0 INC                     2.9%
Others:   8 YSRCP                45.7%


Telangana
TRS to repeat the 2018 assembly election landslide.  CPI TDP are backing to INC to no avail as AIMIM has a tactical alliance with TRS.  YSRCP pretty much pulled out of the state to back TRS.

2019
NDA:      0 BJP                   13.7%
UPA:       0 INC                  31.3%
Others: 16 TRS                  45.5%
              1 AIMIM                2.7%
              0 JSP-BSP             2.9% (tactical alliance with CPI-CPM)
              0 CPI-CPM            1.0% (tactical alliance with JSP-BSP)

2014
NDA:      2 (1 BJP 1 TDP)     22.8%
UPA:      2 INC                    25.7%
Others: 11 TRS                   34.9%
             1 YSRCP                 4.5%
             1 AIMIM                 3.5%


TN
AIADMK split into splinter AMMK and anti-incumbency will defeat NDA despite the size of the alliance.  The NDA defeat could end up being worse than this.   NMN eats into the UPA vote or else it will be a massive UPA landslide.

Note Vellore election was called off but my projection will still be based on all 39 seats.

2019
NDA:      7 (3 AIADMK 2 BJP 1 PMK 1 DMDK 0 PT 0 TMC 1 PNK)         35.7%
UPA      32 (18 DMK 8 INC 2 VCK 2 CPI 2 CPM 0 MUL 1 IJK 0 KMDK )  43.9%
Others:  0 AMMK-SDPI                                                                    11.7%
             0 NMN                                                                                5.6%
             
2014
NDA:       2 (1 PMK 1 BJP 0 DMDK 0 MDMK)     18.7%
UPA:       0 INC                                                4.4%
Others:  39 AIADMK                                       44.9%
              0 DMK                                             27.2%


Kerala
Sabarimala temple issue will crush Left Front and lead to a large scale defection of Left Front vote to BJP who will finally win a seat in Kerala.  This and anti-incumbency will lead to a UPA landslide.  Rahul Gandhi running in Kerala is part of the UPA moving in for the kill.
 
2019
NDA:       1 (1 BJP 0 BDJS 0 KEC(T))              21.3%
UPA:      17 (13 INC 2 MUL 1 KEC(M) 1 RSP)  42.1%
Others:   2  Left Front                                   33.6%

2014
NDA:     0 BJP                                              10.9%
UPA:    12 (8 INC 2 MUL 1 KEC(M) 1 RSP)      42.5%
Others:  8 Left Front                                    41.6%


Assam
NDA was about to sweep the state until the new citizenship bill.  Now it is a real fight with anti-incumbency building up.  AIUDF only running in 3 seats helps INC partly but also drives Hindu votes away from INC to NDA.

2019
NDA:       7 (6 BJP 1 BPF 0 AGP)     43.7%
UPA:        4 INC                            38.5%
Others:    3 AIUDF                         11.4%

2014
NDA:      7 BJP                    38.7%
UPA:       3 (3 INC 0 BPF)     31.5%
Others:   3 AIUDF                15.0%
              1 ex-ULFA               4.2%
              0 AGP                     3.9%


6 Federal districts
I expect here to be a small swing away from NDA to UPA with INC gaining a seat

2019
NDA:     4 (4 BJP 0 AINRC)
UPA:     2 (1 INC 1 NCP)

2014
NDA:      5 (4 BJP 1 AINRC)
UPA:     1 (0 INC 1 NCP)


11 Northeast seats
Before the new citizenship bill NDA was going to sweep.  Now several pro-BJP parties are running separately and away from BJP.  Still NDA plus pro-NDA parties will have the edge. 

2019
NDA:     6 (5 BJP 1 NDPP 1)
UPA:      1 INC
Others:  0 Left Front
             4 pro-NDA (1 SDF 1 NPP 1 UDP 1 MNF)

2014
NDA:      4 (1 BJP 1 NPF 1 NPP 1 SDF)
UPA:      5 INC
Others:  2 Left Front


The total result of all this are

NDA: 251 (210 BJP)  39.1% (31.4% BJP)
UPA   155 (101 INC)  30.8% (22.7% INC)

Back in 2014 it was

NDA:  336 (282 BJP)  38.5% (31.1% BJP)
UPA:     59 ( 44 INC)  23.1% (19.3% INC)

So the BJP vote share are mostly similar to 2014 but INC saw a moderate increase.

In 2019
BJP will come in first place in 210 seats and second place 160 seats making its "reach" 370
INC will come in first place in 101 seats and second place 162 seats making its "reach" 263

In 2014
BJP come in first place in 282 seats and second place 54  seats making its "reach" 336
INC will come in first place in 44 seats and second place 224 seats making its "reach" 268

So INC's "reach" stayed around the same (it became more relevant in TN but lost relevance in WB, Odisha and the Northesst) and BJP's "reach" increased  by 40 some seats (mostly in WB, Odisha and the Northeast)

Rest of the seats are
Pro-NDA            4 (1 SDF 1 NPP 1 UDP 1 MNF)
Pro-UPA             7 (5 TDP 2 JKN)
Federal Front     44 (20 YSRCP 16 TRS 8 BJD)
Grand Alliance   72 (27 AITC 23 SP 20 BSP 2 RLD)
Left                   5 (2 Left Front 2 CPI 1 CPM) [2 CPI 1 CPM from alliance with UPA in TN)
Anti-NDA           5 (3 AIUDF 1 AIMIM 1 PDP)


NDA should be able to form a government with pro-NDA parties and some Federal Front party support.  YSRCP will for sure back NDA as most likely will TRS.  BJD might stay away to deal with its new rival the BJP in Odisha.

[jaichind]

The main story of this election is about the regression to the mean for the BJP in its core states and its battle to try to make up for it in non-core states.

I define BJP core states are the North where Hindi is dominate plus some pro-BJP Western states.  Collectively they are: Jammu of J&K, HP, Haryana, Punjab, Delhi, Uttarakhand, UP, Bihar, MP, Rajasthan, Chhattisgarh, Jharkhand, Gujarat, Maharashtra, Goa, Chandigarh, D&N H, and D&D.  These states and and federal territories make up 319 out of 543 seats.  The rest of India or non-core states make up 224. seats.

In 2014 BJP won 248 out of 319 seats core seats.  NDA overall won 282 out of 319 seats (or 88%) to UPA's 24 seats.   In non-core states the BJP won 34 out of 224 seats with NDA winning 54 out of 224 seats (or 24%) to 35 for UPA.

So the problem for the BJP is that a 88% win ratio is going to be hard to replicate and regression to the mean would mean large losses even if NDA/BJP continues to do well in these seats.  This is why the BJP has to enter into non-core states to try to expand NDA/BJP.  This would be places like WB, Odisha, TN and the Northeast.  And the BJP has to contend with losing TDP in AP as a ally and source of possible seats.

If you take my current projection.  I project that the BJP will go to 156 out of 319 core seats with NDA winning 190 out of 319 seats (or 60%) to UPA's 84 seats.  60% strike rate is still very good but this does mean a massive loss of 92 seats for NDA. 

For non-core states I project BJP will win 54 seats with NDA winning 61 out of 224 seats (or 27%) to 71 for NDA.   A gain of 7 seats for NDA is actually quite impressive given the loss of 17 seats it won in 2014 which are all lose for 2019 given the departure of TDP for NDA.  So the NDA actually achieved a growth of 24 seats in non-cores seats.  If you add in the 4 pro-NDA parties seats in the Northeast  it pushes the NDA strike rate in non-core seats to 29% AFTER the loss of TDP in AP. 

But all of this is nowhere enough to make up for the loss of 92 seats for the NDA in core seats.

This is the explanation of why the NDA vote share in 2014 and 2019 are similar but with very different seat outcomes.  In core seats the sensitively to seat losses with vote share loss are high since they are often bipolar seats where a swing against BJP will bring large losses.  In non-core states the NDA vote share was so low in 2014 that a large vote share swing for NDA which makes up for the vote share losses in core states does not even come close to making up for seat losses.   

In other words in 2014 the NDA had an unusually good vote distribution as far as vote share to seat share conversion was concerned.  2019 is a shift to a more normal distribution of vote for the NDA and it will cost it in terms of seats.

[jaichind]

https://anthro.ai/indianelections/notes/a-socialist-tuesday/

Seems to have done some ground reporting/analysis of the 10 UP seats that voted in Phase III.  The result seem pretty negative for BJP.   These 10 seats are mostly in Yadav heavy areas and in 2014 it went 7 BJP 3 SP.  They have BJP losing 6 seats for sure and the other 4 (Aonla, Bareilly, Etah and Pilibhit) being tossups where they expect BJP to win 2 out of 4.  My model agree with their analysis in the sense that we agree on the 3 seats BJP won in 2014 that it will lose to SP (Moradabad, Rampur, Sambhal.  On the 4 tossups seats they have I have all of them going to BJP by anywhere between 3% to 10%.   Pilibhit is Varun Gandhi's seat (grandson of Indira Gandhi and son of deceased Sanjay Gandhi) and if that is a tossup then that is truly problematic for BJP.   My model has NDA at 33 seats and 40.3% vote share in UP.  If this projection is true then NDA might be looking at low 20s in terms of seats and something like 36%-37% vote share.

[jaichind]

Thinking about my concept of core-BJP states of (Jammu of J&K, HP, Haryana, Punjab, Delhi, Uttarakhand, UP, Bihar, MP, Rajasthan, Chhattisgarh, Jharkhand, Gujarat, Maharashtra, Goa, Chandigarh, D&N H, and D&D.) which make up 319 out of 543 seats versus non-core seats which make up 224 out of 543 seats.  It seems it would be good to look at the history of BJP, BJP allies and NDA seats since 1991 grouped by core and non-core seats.


                      BJP                          BJP Allies                          NDA
          Core   Non-Core  Total    Core  Non-Core   Total     Core  Non-Core   Total
1991   113          7        120       4          0            4          117      7           124
1996   154          7        161      25         0           25         179      7           186
1998   153        29        182     29        46           75         182     75           257
1999   150        32        182      44       75         119         194    107          301
2004   109        29        138      27       24           51         136     53          189
2009     91        25        116      41        3           44          132     28          160
2014   248        34        282      34       21           55         282     55          337
2019   156        54        210      34       11           45         190     65          255

2019 are my current projections

The story is pretty clear.  
1) In 1991 BJP had no allies nut was strong in its core seats but nowhere in non-core seats.  
2) In 1996 there was a BJP surge in just its core seats with the help of allies but still nowhere in non-core seats.
3) In 1998 BJP got a bunch of allies in non-core seats while repeating its 1996 results in core seats to be able to form government.  
4) In 1999 being in government allowed the BJP to rope in more allies in core states but especially in non-core states to get to a majority for NDA by itself.  
5) 2004 saw BJP and allies lose ground across the board in both core and non-core seats as the NDA is ousted from power.  
6) In 2009 with the help with allies in core states the NDA repeated its performance in 2004 for core seats but the loss of allies in non-core seats saw BJP still hold its 2004 ground there but NDA loses ground due to loss of allies.
7) In 2014 there as a massive Modi wave in core seats which allowed the BJP to win over some allies in non-core seats and improve NDA tallies there as well. A  BJP majority on its own was pretty much built by the unprecedented saturation of BJP in core seats.
8  ) In 2019 my projections have BJP falling to 1996-1999 levels in core seats which represents the historical norm for a good BJP election but not a fantastic one (like in 2014) and BJP gaining a lot of ground on its own in non-core seats with fairly little help from allies.

If my 2019 projection is correct then the main trend of 2019 is the expansion of BJP into some non-core states (WB and Odisha) without help of allies while growth for BJP in the South still stunted while in core seats it is above reversion to the mean to a good but not excellent result for BJP.  What Modi and BJP has to hope for is that in core states the 2014 set a new benchmark for the BJP and that another excellent result for BJP is achievable.  

[jaichind]

It seems INC have nominated a ex-MLA and its 2014 candidate for Varanasi which means that Priyanka Gandh will not contest.  I guess INC high command made a call that tying down Modi  in Varanasi it not worth making this election more "Presidential" by having a high profile face-off.  The key to INC has to be to have less of Modi in the media.  Not having a high profile contest I guess was their call to achieve that.    I still feel that having Rahul Gandhi run against Modi sets him up the best to become PM in 2024.