India Assembly elections 2018- Karnataka, MP, Rajasthan, Chhattisgarh, Telangana

[jaichind]

One lesson of the UP by-elections is that the best way to defeat BJP in UP in 2019 would be a BJP-AD-SBSP vs SP-BSP-RLD vs INC battle.  

The lesson of 2017 was that a SP-INC alliance failed to capture Upper Caste votes despite the fact that INC is the second choice of most Upper Caste voters plus Akhilesh Yadav's "New SP" modern image.  In the end the Upper Caste vote still saw the "New SP" and still just sees an OBC party and even the pro-INC Upper Caste voter went with BJP.

This by-election in UP showed that if SP and BSP are willing to compromise to work out an alliance then the Yadav-Muslim-Dalit (plus Nishad in Gorakhpur) vote base can come together despite historical rivalry.   Had INC joined this alliance it would merely drive INC Upper Caste voters over to BJP as well as to trigger an Upper Caste and non-Yadav OBC counter-consolidation in response.  

A situation where it is BJP-AD-SBSP vs SP-BSP-RLD vs INC battle would avoid this and INC running Upper Caste candidate across the board would eat into the BJP Upper Caste vote.  In many ways that was what the INC was trying to do in 2014 but failed due to the Modi wave.  In retrospect the SP and INC got their strategies reversed.  In 2014 the Modi wave was so strong that BJP was going to win all the vote they could have received there was no way to beat it back so a SP-INC alliance to save a few seats would have made sense in retrospect.  In 2017 a lot of votes were up in the air so SP and INC running separately could have clawed back from Upper Caste votes from BJP.

All this is clear in retrospect but hard to figure out at the time.  How SP BSP and INC choose to fight BJP in 2019 would make a difference of 30 or even more seats.

[jaichind]

Exact UP LS by-election results

Phulpur
SP            47.1%
BJP           39.0%
SP rebel     6.6%
INC            2.7%

Gorakhpur
SP            49.3%
BJP           47.0%
INC            2.0%

My "fair value" vote share based on 2017 assembly segments and the assumption that BSP RLD NISHAD votes would go SP

Phulpur
BJP           43.4%
SP            46.2%
INC            5.8%

Gorakhpur
BJP           43.2%
SP            47.8%
INC            5.5%

Overall it is clear that BSP managed mostly  transfer its vote to SP.

BJP under-performed in Phulpur part of which are explained by the INC Upper Caste candidate.  The SP rebel seems to have capture some anti-SP Muslim votes which would have gone INC as well as anti-SP Dalits which would have gone BJP.  Of course the SP rebel got some SP votes as well.

BJP mostly performed par in Gorakhpur.  The INC candidate was a Muslim which means that the INC Upper caste vote mostly gone BJP.  On the other hand there was also INC Muslim tactical voting for SP despite the INC candidate being Muslim to focus on defeating BJP which worked.

The overall narrative of INC impact on election is clear.  Where the INC nominate an Upper Caste candidate it tends to hurt BJP, where it nominates a Muslim candidate it tends to hurt SP.

[jaichind]

I'd say that the Phulpur loss isn't particularly surprising, but Gorakhpur is a bit of an embarrassment. O/c insisting that it's much more than a mere embarrassment is probably reaching a little bit...

Bit of trivia for you all: Phulpur was Nehru's constituency from 1952 until his death.

Certainly from the LS election history point of view Phulpur has historically been a marginal seat while BJP has been winning  Gorakhpur since the 1980s.  Going into the by-election I would say SP was more worried about Phulpur due to the SP rebel running in the fray.  Also as I pointed out, the assembly segments results of Gorakhpur and Phulpur from 2017 seems to indicate that both had a small BJP lean relative to UP but no more.  BJP's strong record for Gorakhpur LS election had more to do with two superstar BJP MPs (Yogi and his mentor before him) that punched above BJP's weight.

As for Phulpur, after Nehru passed away, his sister held the seat for INC for a couple of terms.  Then in the 1970s INC's VP Singh who would then from JD in 1988 and then becoming PM in 1989 also was the INC MP there. 

[jaichind]

Official results for Bihar by-elections

LS Araria by-election result
RJD      50.0% (backed by INC-HAM-CPI)
BJP      43.9%  (backed by JD(U)-LJP-RLSP)
JAP        2.1%
RJP        1.8%  (BSP splinter)


Back in 2014 it was
RJD      41.8% (backed by INC)
BJP      26.8%  (backed by LJP-RLSP)
JD(U)   22.7%  (backed by CPI)
BSP       1.8%

In 2015 assembly segments RJD rebe JAPl got 4.1%.  So if we assume 2014 JD(U) performance of 22.7% was made up of CPI 1% and what became HAM 3% a "fair value" 2018 result should be

RJD   41.7%
BJP   45.5%
JAP    4.1%

And if we swing 3% from BJP to RJD for sympathy factor (son of old RJD MP who passed away is running for RJP) it should have been

RJD   44.7%
BJP   42.5%
JAP    4.1%

So RJD "should" have won by 2% but won by 6% instead.  So around 2% of the ruling coalition vote was leaked to RJD.  Looking at the assembly election results this amount should have been much higher but given the fact that this district has high Muslim minority density and the RJD candidate is Muslim the leakage to RJD was more limited.




Assembly Bhabua  by-election result
BJP              48.1% (backed by JD(U)-LJP-RLSP)
INC              37.0% (backed by RJD-HAM-CPI)
BJP rebel       2.8%


In 2014 LS assembly segment it was
BJP        46.4% (backed by LJP-RLSP)
INC        30.8% (backed by RJD)
JD(U)     14.6% (backed by CPI)
BSP         3.8%

So if we assume 2014 JD(U) performance of 14.6% was made up of CPI 1% and what became HAM 2% a "fair value" 2018 result should be

BJP        58.0%
INC        33.8%

And if we swing 3% from INC to BJP for sympathy factor (wife of old BJP MLA who passed away is running for BJP) it should have been

BJP        61.0%
INC        30.8%

So BJP "should" have won by 30% but won by 14% (taking BJP rebel into account).  So 8% of the ruling coalition vote has leaked to INC.




Assembly Jahanabad by-election result
RJD              55.5%  (backed by INC-HAM-CPI)
JD(U)           29.9%  (backed by BJP-LJP-RLSP)
CPI(ML)(L)     6.2%


Back in 2014  LS assembly segment it was
RLSP          41.0% (backed by BJP-LJP)
RJD            36.9% (backed by INC)
JD(U)         11.0%  (backed by CPI)
CPI(ML)(L)   3.5%
BSP             1.3%

So if we assume 2014 JD(U) performance of 11.0% was made up of CPI 1% and what became HAM 2% a "fair value" 2018 result should be

RJD            39.9%
JD(U)         49.0%
CPI(ML)(L)   3.5%

And if we swing 3% from JD(U) to RJD for sympathy factor (son of old RJD MLA who passed away is running for RJD) it should have been

RJD            42.9%
JD(U)         46.0%
CPI(ML)(L)   3.5%

So JD(U) "should" have won by 3% but lost by 25%.  So 14% of the ruling coalition vote has leaked to RJD.


It seems the conclusion is that in 2019 JD(U)-BJP-LJP-RLSP will have a fight on its hands with RJD-INC-HAM-CPI and will not be a cakewalk with the level of vote leakage from the ruling block to RJD-INC.

[jaichind]

TDP has quit BJP led ruling NDA.

Its hand was forced.  Its main AP rival YSRCP has been turning the state against BJP for not granting AP special status and then tying TDP to BJP.   TDP tried to fight back by pulling out of NDA government but staying for now in NDA alliance.  TDP wants it both ways and have the BJP vote base ~4-5% but not have the large anti-BJP vote consolidate around YSRCP.   YSRCP counted by moving a vote of no confidence against the NDA government  (with INC support) and daring TDP to abstain.

TDP has decided that the narrative in 2019 in AP will be fairly anti-BJP and choose to cut its losses and leave the NDA alliance and also introduced its own motion of no confidence against the NDA.

In a surprise AIADMK which is fairly pro-BJP these days also came out in support for the vote of no confidence.  Again here, the AIADMK fear is that AIADMK rebel TTV Dhinakaran's new party AMMK will capture the anti-BJP AIADMK vote if AIADMK is seen as pro-BJP when the sentiment in TN is moving against BJP.

Also in a surprise NDA member SHS announced that it will also vote for the vote of no confidence.  SHS has already announced that in 2019 it will quit NDA and run separately from BJP but for now are still in the NDA government in the Center and Maharashtra for now.  What is bizarre about this decision is SHS has MPs in the cabinet which it is now going to vote against.  So SHS ministers will be voting against itself.

In Kerala BJP ally BDJS which is a Ezhava outfit also decided to quit its alliance with BJP.  In 2016 BJP-BDJS won an unprecedented near 15% of the vote in Kerala.  If this alliance breaks up then BJP's surge in Kerala would be halted and most likely reversed. 

[jaichind]

India equity markets fell 2% overall over 2 days on by-election results plus TDP pulling out of NDA.

[jaichind]

Now that the Bihar by-elections have shown that RJD-INC would be competitive with JD(U)-BJP in 2019 there will be a significant seat sharing problem for the JD(U)-BJP alliance.  In many ways game theory would indicate it is inevitable that RLSP will leave JD(U)-BJP to join RJD-INC alliance soon.

In 2014 seats sharing in Bihar was

BJP+
BJP      30  
LJP        7
RLSP     3

RJD+
RJD     27
INC     12
NCP      1

JD(U)+
JD(U)  38
CPI       2

Now that JD(U) joined up with NDA alliance there will be a major seat sharing problem since you have to fit JD(U) BJP LJP RLSP into 40 seats.  Back in 2009 the JD(U)-BJP seat sharing  was

JD(U)   25
BJP      15

So JD(U)'s initial position is: JD(U) formed an alliance with BJP but not LJP and RSLP so seats sharing should be like 2009 and BJP accommodate LJP and RLSP out of it's share which is

JD(U)   25
BJP      10
LJP        4
RLSP     1

Which given the fact that BJP has 22 MPs, LJP have 6 MP, and RLSP have 3 MPs is clearly not possible.  JD(U) will then say, lets keep the 25:15 formula but then try to accommodate LJP and RLSP which means it will be something like

JD(U)   22
BJP      13
LJP        4
RLSP     1

Still it would result in too many incumbent BJP LJP RLSP MPs just giving up their seats which is unacceptable and lead to mass rebellion.

Then after pressure JD(U) will accept that JD(U) and BJP are at par in Bihar so there should be a equal sharing of seats between JD(U) and BJP which means it would be something like

JD(U)   17
BJP      16
LJP        5
RLSP     2

Still it leaves mostly too many of the 22 incumbent BJP MPs without a job.  Under such pressure most likely JD(U)-BJP will try to squeeze LJP and RLSP to get to

JD(U)   17
BJP      18
LJP        4
RLSP     1

At this stage RLSP which has not burned their bridges with RJD-INC would then bolt to join RJD-INC.  LJP has burned its bridges with RJD-INC somewhat back in 2014 will only leave in an extreme situation.  Without RLDP then it becomes something like

JD(U)   17
BJP      18
LJP        5

or

JD(U)   17
BJP      19
LJP        4

Which I could see it barely making it.  BJP will need to find 3 incumbent MPs to give up their seats and LJP will need to find 2  incumbent MPs to give up their seats and JD(U) will have to accept a slightly lower seat share than BJP.  Note that since it is clear that the 2019 election would be competitive then it would increase the pressure for an aggressive negotiation posture in the NDA for seats since a seat allocated most likely means a best a 60% or even 50% chance of winning a seat unlike 2014 when it turned out it was a 75% chance.




On the INC-RJD side.  HAM and most likely NCP will come aboard.  Most likely the seat share will be

RJD     26
INC     11
HAM     2
NCP      1

HAM has no MPs so once it is clear that HAM cannot go back to JD(U)-BJP they will have to accept.  If RLSP were to come over, most likely they will demand 4 seats (they have 3 MPs) but will most likely accept 3 given they would have been offered 2 or even 1 with JD(U)-BJP.  So it will become

RJD     24
INC     10
RLSP    3
HAM     2
NCP      1

Which would mostly be acceptable to all parties as RLSP and HAM would get a worst deal with JD(U)-BJP.   Also since RJD only has 4 MPs and INC only has 2 MPs this distribution gives both a good chance to expand their numbers.

So gaming out negotiations on both sides seem to indicate the rational choice of RLSP would be to leave JD(U)-BJP and join up with RJD-INC.

[jaichind]

https://in.news.yahoo.com/ram-temple-social-media-shifting-075019489.html

Points out that 2019 India LS elections will cost around 25,000 crore (10 million) INR.  That is about $4 billion with straight conversion and around $14 billion in PPP terms.   In contrast the cost of 2016 US election was around $6.5 billion (Prze Hosue and Senate races put together which is comparable to LS although we are not adding in cost of Upper House RS elections.)

It seems spending on Indian elections in PPP terms is more than half of US election spending on a per capita basis (India's population is around 4 times that of USA.)

[jaichind]

In Karnataka, the INC government is making a move to recognize Lingayatism is a separate religion.   

https://www.indiatoday.in/india/story/who-are-lingayats-and-why-karnataka-chose-to-recommend-religious-minority-status-for-them-1193551-2018-03-20

Due to BJP leader and former CM Yeddyurappa who is a Lingayat, Lingayat  has the last few cycles voted BJP.  INC is trying to break this by pitting BJP Hindu nationalism to some in the Lingayat  community that does want to be recognized as a separate religion.

I am not sure such a move so close to the election would work and most likely INC would lose more non-Lingayat votes than any Lingayat votes it would get.  If anything this helps JD(S) the most as some non-Lingayat INC votes might go to JD(S) but some BJP Lingayat votes goes to INC.  The politics of identify always work likes Jujutsu: the first open attempt at polarization/consolidation does not get as much result as the counter-polarization/consolidation in response.

[jaichind]

2018 RS elections took place. Mostly went as expected.  Two critical battles were in UP and Jharkhand where two grand alliances faced off with BJP.  In UP BJP can win 8 out of 10 seats for but a grand alliance of SP BSP INC RLD can prevent a 9th BJP victory.  In Jharkhand BJP can win 1 out of 2 seats but a grand alliance of INC JMM JVM can prevent a 2nd BJP victory.   In UP the grand alliance effort failed mostly due to some BSP MLAs who are in jail being disqualified and a BSP MLA defecting last minute so it ended up being 9 BJP 1 SP instead of 8 BJP 1 SP 1 BSP.  In Jharkhand the grand alliance won and split the 2 RS seats with BJP so it ended up being 1 BJP 1 INC and not 2 BJP.

The BJP victory in UP is somewhat Pyrrhic since the BSP candidate only lost due to issues within BSP.  The grand alliance mostly held where INC SP RLD all backed the BSP candidate.  It is a good confidence building measure which makes SP-BSP alliance more likely in 2019 LS election.

[jaichind]

C-Fore Karnataka survey has INC ahead with a positive swing from 2013

           Seats        Vote Share
INC       126              46%
BJP         70              31%
JD(S)      27              16%


Back in 2013 if we add up BJP vote share with BJP splinters KJP and BRSCP (both of which have since merged back into BJP) plus BJP rebels and compare it to INC (with rebels) and JD(S) (with rebels) vote share we get

INC+      38.42%
BJP+      34.42%
JD(S)      20.78%

So this poll seems to indicate that the INC will gain a strong vote share swing from 2013 with the BJP getting more seats than in 2013 only because now the BJP vote is not split between BJP KJP and BRSCP.

I mostly find this unlikely.  I suspect JD(S) is underestimated and INC overestimated.   here Whet I get out of this poll is that INC is most likely ahead of BJP but not clear at it that it will have a majority.  Most likely election result based on this poll is INC largest party but without majority.  If so the most likely government formation out of that is an unstable BJP-JD(S) government and the prospect of an early midterm election.

[jaichind]

The churn of political alliances in Bihar is working out the way I mostly have expected post the Bihar by-elections.  I felt that the most likely outcome was that eventually RLSP will leave BJP-JD(U) and join up with RJD-INC. 

As if on queue RLSP came out saying that "all options are on the table" for 2019 alliances.  Then there were a bunch of meetings between JD(U) Nitish Kumar and LJP leader Ram Vilas Paswan, followed by RLSP leader Upendra Kushwaha, and then JAP leader Pappu Yadav.  What is key here is that Upendra Kushwaha used to be a key Nitish Kumar sidekick until he broke with Nitish Kumar in 2013 to form RLSP.  Nitish Kumar is also a long time enemy of RJD rebel and now JAP leader Pappu Yadav.  That these meetings are taking place indicates that these parties plan to gang up on BJP to get their fair share of seats for 2019.  As I pointed by the Bihar by-elections showed that the BJP led bloc strike rate in 2019 LS elections will not be 75% but at best 60% or even 50%.  So that makes competition for seat allocation even more desperate.  The BJP position is that BJP will contest 22 out of 40 seats since BJP has 22 MPs in Bihar.  That clearly is not acceptable to the BJP allies so it seems a JD(U)-LJP-RLSP bloc will form within NDA to push down the BJP seat count.

There is even talk of a massive anti-BJP anti-RJD alliance where it could be JD(U)-INC-LJP-RLSP-HAM-JAP vs BJP vs RJD.  Most likely this is a threat that the JD(U)-LJP-RLSP bloc will be using to extract more seats allocation from BJP.  Of course the main weakness in this wide alliance is Nitish Kumar and JD(U).  JD(U) has been in power in Bihar (with different partners) since 2005 and the NDA by-election setbacks are as much as about the receding of the Modi wave as about exhaustion of patience for Nitish Kumar/JD(U).  If so then even if a  JD(U)-INC-LJP-RLSP-HAM-JAP alliance is formed then most likely it will be driven the third place by BJP and RJD.   I still think the most likely outcome ofa future JD(U)-LJP-RLSP vs BJP brinkmanship is the BJP will prevail and then to make the math work RLDP will be sacrificed followed by RLSP joining up with RJD-INC.

[jaichind]

In WB the pro-Gorkhaland GJM has quit its alliance with BJP indicating that BJP has done nothing to advance the creation of Gorkhaland despite pre-election promises.  Most likely that in the 2019 the BJP will lose the Darjeeling LS seat which it held in both the 2009 and 2014 LS elections based GJM support.

[jaichind]

If the INC can become the largest party in Karnataka in the May elections then it might signal the shift of the assembly election super-cycle against the BJP.  In many ways its is destined to come.  The Nov 2013 assembly elections of  MP, Rajasthan, Chhattisgarh, and Delhi was the start of the BJP victory super-cycle taking advantage of the fact that the ruling party of said state was a non-BJP party or when the ruling party at the center was the UPA.

If you look at the various large Indian states and what the winning party by year you can derive what the winning party in the next part of the cycle of which most went in favor of BJP in the 2013-2017 period but now will begin to turn against BJP.  Of course I assume that in 2019 LS elections BJP led alliance will come back to power (likely with a reduced majority) so the anti-incumbency on the center will weigh down the BJP.


Assam
INC tends to have the edge.  2016 INC defeat has to do with the fact that it has been in power for 3 terms.  The 2016 INC defeat had the INC vote share mostly intact and most things equal INC will win in 2021.

2021   INC
2016   BJP-AGP
2011   INC
2006   INC
2001   INC
1996   AGP
1991   INC
1985   AGP


Bihar
By 2015 it was clear the voters of Bihar was getting tired of JD(U) after 2 terms just like by 2005 it was getting tired of RJD and put in JD(U)-BJP.  The JD(U)-RJD grand alliance managed to beat the BJP in 2015.  In 2020 the BJP will be tied to 3 terms of incumbency of JD(U) and most likely RJD-INC will win.

2020   RJD-INC
2015   JD(U)-RJD
2010   JD(U)-BJP
2005   JD(U)-BJP
2000   RJD  
1995   JD
1990   JD
1985   INC


Chhattisgarh
BJP is now in its 3rd term and there is a 50/50 shot it will lose in 2018 given the fact that the BJP-INC gap have pretty much closed to nil by 2013.  Even if it narrowly wins in 2018 it will lose for sure to INC in 2023.

2018   INC
2013   BJP
2008   BJP
2003   BJP
1998   INC
1993   INC


Delhi
Here the cycle will work in favor of BJP.  As AAP stumbles with real power BJP should come storming back in 2020.

2020   BJP
2015   AAP
2013   No winner (BJP largest party narrowly over AAP,  AAP-INC forms government)
2008   INC
2003   INC
1998   INC
1993   BJP


Gujarat
BJP narrowly won in 2017.  After winning 7 terms in a row (it was going to get defeated in 2002 until the Gujarat riots of 2002 helped Modi to win) the INC is very likely to win in 2022.

2022   INC
2017   BJP
2012   BJP
2007   BJP
2002   BJP
1998   BJP
1995   BJP
1990   no winner (JD narrowly beat out BJP for largest party)
1985   INC

Haryana
3 cornered state between BJP INC and INLD.  BJP won in 2014 due to 2 term INC incumbency.  A reversion of the mean will have INC come back in 2019.

2019   INC
2014   BJP
2009   INC
2005   INC
2000   INLD-BJP
1996   HVP-BJP
1991   INC
1987   LKD
1982   INC


HP
Here INC and BJP alternates winning elections.  So in 2022 it will be the INC's turn.

2022   INC
2017   BJP
2012   INC
2007   BJP
2002   INC
1998   BJP (tie with INC)
1993   INC
1990   BJP
1985   INC
1982   INC


Jharkhand
BJP has been strong here but support tends to be split between BJP INC JMM and JVM.  Most likely in 2019 anti-incumbency and a INC-JMM-JVM grand alliance will defeat BJP.

2019   INC-JMM-JVM
2014   BJP-AJSU
2009   INC-JVM (largest bloc)
2005   BJP
2000   BJP


AP
BJP only does well when allied with TDP.  YSRCP has taken over from INC due to the 2013 Telengana
 split.  Now TDP has broken ties with BJP it will be TDP or YSRCP in 2019 with BJP out in the cold.  YSRCP most likely will win in 2019.

2019   YSRCP
2014   TDP-BJP
2009   INC
2004   INC
1999   TDP-BJP
1995   TDP  
1990   INC
1985   TDP
1983   TDP
1978   INC


Karnataka
No ruling party has won re-election since 1985.  But if the CM manages to finish a 5 year term the ruling party can prevent an opposition majority which was the case in 2004 and most likely the case in 2018.

2018   No winner (INC most likely largest party but with BJP-JD(S) forms government)
2013   INC
2008   BJP
2004   No winner (BJP largest party but INC-JD(S) forms government)
1999   INC
1994   JD
1989   INC
1985   JNP


MP
BJP has won 3 terms in a row when going into 2018.  Even if it wins in 2018 which is a better than 50/50 shot its margin would be reduced and is is very likely in 2023 INC will win.

2023   INC
2018   BJP
2013   BJP
2008   BJP
2003   BJP
1998   INC
1993   INC
1990   BJP
1985   INC


Maharashtra
Area of traditional INC strength but BJP has been rising in strength here.  BJP becoming the senior partner has worked to drive away SHS.  If 2019 SHS runs separately from BJP then INC-NCP are likely to win.

2019   INC-NCP
2014   BJP-SHS
2009   INC-NCP
2004   INC-NCP
1999   INC-NCP
1995   SHS-BJP
1990   INC
1985   INC


Orissa
State went from a BJD vs INC state to a BJD vs INC vs BJP state after a period of BJD-BJP alliance.  Now it is a battler for second place (BJP or INC.)  In 2019 BJD is likely to win but establish the BJP as the alternative to BJD.  In 2024 BJP will most likely win.

2024   BJP
2019   BJD
2014   BJD
2009   BJD
2004   BJD-BJP
2000   BJD-BJP
1995   INC
1990   JD
1985   INC


Punjab
Here the cycle will move in favor of SAD-BJP.  Usually the incumbent does not win re-election and the 2012 SAD-BJP win was an exception which only set up a massive victory for INC in 2017. AAP wave has mostly receded.  In 2022 SAD-BJP are likely to come back.  

2022   SAD-BJP
2017   INC
2012   SAD-BJP
2007   SAD-BJP
2002   INC
1997   SAD-BJP
1992   INC
1985   SAD  


Rajasthan
Last few election cycles BJP and INC alternates in power with large wild swings.  In 2018 it will be INC's turn to win.

2018   INC
2013   BJP
2008   INC
2003   BJP
1998   INC
1993   BJP
1990   BJP-JD
1985   INC


UP
No incumbent party has won re-election since 1985.  2022 will be no different.  Assuming SP-BSP alliance holds BJP will be defeated in 2022.

2022   SP-BSP
2017   BJP
2012   SP
2007   BSP
2002   SP (largest party)
1996   BJP (largest party)
1993   SP-BSP (largest bloc)
1991   BJP
1989   JD
1985   INC


Uttarakhand
INC and BJP alternatives in power.  2022 will be the INC's turn to win.

2022   INC
2017   BJP
2012   INC
2007   BJP
2002   INC
1996   BJP


Telengana
Like Orissa we have a dominate regional party, TRS, with INC and BJP fighting for second.  More likely or not INC will emerge as the main alternative to TRS.  2019 TRS will win due to split opposition but in 2024 INC will win.

2024   INC
2019   TRS
2014   TRS


TN
AIADMK and DMK alternates in power with the party that has the stronger alliance partner (usually INC) winning.  AIADMK won re-election in 2016, a first since 1984.  But with Jalaylalitha gone and AIADMK split it is clear that in 2021 DMK-INC will come back.  AIADMK is more pro-BJP than DMK so this next shift will be part of the anti-BJP cycle.

2021   DMK-INC
2016   AIADMK
2011   AIADMK-DMDK
2006   DMK-INC
2001   AIADMK-INC
1996   DMK-TMC
1991   AIADMK-INC
1989   DMK
1984   AIADMK-INC
1980   AIADMK-INC
1977   AIADMK

[jaichind]

In WB the pro-Gorkhaland GJM has quit its alliance with BJP indicating that BJP has done nothing to advance the creation of Gorkhaland despite pre-election promises.  Most likely that in the 2019 the BJP will lose the Darjeeling LS seat which it held in both the 2009 and 2014 LS elections based GJM support.

Who are they backing now? Going solo?

Unclear.  This move by GJM seems to have provoked a civil war within the GJM between pro-BJP and anti-BJP factions so perhaps GJM might come back.  Of course the underlying conflict is over BJP's pre-2014 support of a separate Gorkhaland which was the basis of GJM-BJP alliance in Darjeeling of WB.  Now that BJP is in charge at the center GJM and the BJP MP of Darjeeling have been demanding the Modi government act on the BJP's promise.  Modi and BJP had to dance around this since if they forced the creation of Gorkhaland then the BJP can kiss being  relevant political force in WB for a couple of election cycles. 

[jaichind]

With by-election setbacks and TDP leaving NDA the various regional opposition parties smell blood for 2019 LS elections.  There seems to be 3 streams of attempts to form an anti-BJP front in 2019.  Those that want to form an regional party alliance with INC to take on BJP, those that want to form an non-INC but anti-BJP front, and those that want to form an anti-INC anti-BJP front.  It really depends on how the regional parties relationship with INC in their home state.

NCP and RJD are allies with INC in Maharashtra and Bihar which in turn mean they are for a federal front allied with INC.  AITC which are rivals with INC (which might ally with Left Front to take on AITC) does not see INC as a long term threat but does see BJP as a long term threat are for a non-INC but anti-BJP front.  TRS TDP and BJD which have an anti-INC past or sees INC as rivals in Telangana, AP and Orissa want an ant-INC anti-BJP front.

The position of JD(S) in Karnataka interesting.  On paper they should be like AITC where they are for non-INC anti-BJP front since JD(S) was formed as an anti-BJP party in 1999 to protest JD(U) desire to form an alliance with BJP.  In reality everyone knows that all things equal JD(S) is much more likely to form an alliance with BJP post Karnataka election.  The reason is because in Karnataka JD(S) is relevant only in South Karnataka where its Vokkaliga base is large.  In fact Karnataka politics last few cycles are mostly about INC vs JS(S) in the 60 some seats in South Karnataka and INC vs BJP in the 160 some seats in North and Coastal Karnataka as the JD(S) and JD(US) base outside of South Karnataka has drifted toward BJP.  This is the main reason why INC tends to under-perform in terms of seats relative to vote share last few election cycles.  So in many ways BJP and JD(S) are natural allies since their respective bases are not in conflict with each other.  AITC has been trying to get JD(S) to join an non-INC anti-BJP front to no avail.  The reason why is obvious.

[jaichind]

jaichind, what do you anticipate the main issues in the 2019 LS elections to be?

Most likely a combo of various communal issues plus developmental issues (employment, inflation etc etc.) I think what is key is not the balance of those issues but if the election will be a "Presidential" election (like 1971, 1977, 1984, and 2014) or one which is a confederation of state elections (2004 is a very good example of this.)  If the former then Modi will come back to power with NDA getting a majority and BJP most likely barely missing a majority by itself.  If the later then Modi might be in trouble and BJP might be reduced to around 200 seats and NDA will struggle to get a majority and the BJP might have to dump Modi to win enough allies to form a government.

The opposition main advantage is that in many states the alliance math and/or the election cycle is moving against BJP.  The main BJP advantage is the Modi brand and that the opposition are not able to project a PM candidate of equal stature.

[jaichind]

A good part of Northern India was shut down today due to mass Dalit demonstrations and rioting in high Dalit concentrated areas (UP MP Punjab Haryana Rajasthan).  This was to protest a recent Supreme Court ruling that watered down the  Scheduled Caste and Scheduled Tribe (Prevention of Atrocities) Act that put a greater burden of proof on victims of anti-Dalit atrocities.  







It seems a good part of the demonstrations is directed at the BJP despite the fact that it was a SC ruling and nothing to do with the Modi regime.  This is ominous as a good part of the pro-BJP swing in 2014 from 2009 was from the BSP Dalit vote bloc.  If this continues to be an issue then this swing will be reversed in 2019.

[jaichind]

http://www.timesnownews.com/india/article/karnataka-congress-leader-plays-swing-exercise-quits-party-in-morning-joins-bjp-comes-back-by-evening/215347

Funny story about how Karnataka Panemangaluru block INC secretary Sundara Devinagara was welcomed to the BJP at a function in the morning by U Rajesh Naik, who is to contest from Bantwal constituency and then rejoined INC at another programme held at Mani a few hours later, after apparently being cajoled by INC party leaders.   So he changed parties twice within a 8 hour period!!

[jaichind]

With UP BJP government trying to deal with a resurgent opposition with the SP-BSP alliance as well as Dalit protests and clear anger by Dalit BJP MPs and MLA toward the UP BJP government, yet another blow has taken place.  A 16 year old girl tried to commit suicide in front of the UP BJP CM Yogi Adityanath's house.

It seems that a BJP MLA and minister in the UP cabinet Kuldeep Singh Sengar

who is a famous party hopper (he started in the Youth INC in the late 1990s then was elected as MLA with BSP in 2002 and then elected in 2007 and 2012 with SP and then defected to BJP and elected again in 2017 has alleged organized a gang rape of this 16 year old last year with his brother.  After the rape the girl's family worked to file charges to no avail with the police. The girl's family got repeated calls (which were recorded) from the BJP MLA Kuldeep Singh Sengar threatening them to drop charges. Then the BJP MLA  Kuldeep Singh Sengar and brother got the police to arrest the girl's father who then was tortured and then died in prison.  This triggered the girl to then to go to the house of the UP CM to unsuccessfully commit suicide.   

This story is leading to large outrage in the media especially when  Kuldeep Singh Sengar still have not been questioned by the police yes and still is a member of the UP cabinet.

[jaichind]

India Today-Karvy poll for Karnataka has hung assembly with slight INC lead
 
              Seats    vote share
INC             96           37%
BJP             82           35%
JD(S)-BSP   39           19%
Others          7             9%


which is pretty much 2013 results but with the 2013 BJP splinters (KJP BSRCP) merging back into the BJP vote.

[jaichind]

In Karnataka even though JD(S) has an alliance with BSP, it seems that with the anti-BJP SP-BSP alliance being formed in UP BSP has not broken the alliance but now seems much more passive about the election given the perception that post election JD(S) is more likely to back BJP than INC. NCP which had planned to perhaps ally with JD(S) now back backed out of that and will back INC in the Karnataka.

Given the bad blood between BJP and TDP after the split of TDP from NDA over the Special status for AP, TDP is telling to Telegu population in Karnataka to vote against BJP.

[jaichind]

More signs of political churn ahead of the 2019 LS elections.  We already had the unprecedented SP-BSP alliance which was unthinkable given their post-1995 rivalry. 

Now there are signs of INC and TDP coming together with TDP backing INC in the Karnataka assembly elections and INC in Telangana openly talking about pursuing an INC-TDP alliance which breaks the old INC-TDP rivalry that started in 1983

In Maharashtra NCP came out and said that SHS does not pose a danger to secularism which implies that NCP is open to an alliance with SHS. 

Both events have geographical reasons.  When AP was united it was very simple: it was INC vs TDP.  Then when the Telangana issue started the early 2000s the TDP branch in Telangana broke off and formed TRS.  As Telangana was formed the Seemandhra branch of INC broke off and formed YSRCP.  So in Seemandhra (now AP) it is now a battle between TDP vs YSRCP.  In Telangana it is now INC vs TRS.  So both TDP and INC have different rivals in the two regions making it much easier to form an alliance.

In Maharashtra SHS tends to be strong in the Mumbai area and weak in the rest of Maharashtra while SHS is exactly the opposite.  So these two parties natural allies.  The idea of a NCP-SHS alliance has been talked about before but it is coming up again.

[jaichind]

With about a year to go before 2019 LS elections I decided to do a bottoms up district-by-district prediction of 2019 LS results.

My assumptions are
1) Unlike 2014 this will be a confederation of various state dynamics driven by alliance math as well as anti-incumbency.  2004 was like this model.  2014 and to some extent 199 there was a Modi and Vajpayee wave that made the election more "Presidential" choice election.   If 2019 was such a choice election then BJP/Modi will do fairly well.  I am assuming that this election will be "wave-less"
2) Various anti-BJP alliance driven by the rise of the BJP gets created in several critical states (UP, Jharkhand), INC gaining some critical allies (TDP in AP and Telangana), and BJP unable to retain key allies (SHS and TDP)
3) Most of the 2014 AAP vote outside of places like Delhi Punjab and Haryana were from urban middle class voters that will mostly flow back to BJP now that AAP fad is over and AAP will mostly retreat to the Greater Delhi area.
4) State level support tend to match last assembly elections with ever greater anti-incumbency.
5) Karnataka assembly election in May 2019 gives INC an edge over BJP
6) For Nov 2018 elections, MP assembly election has BJP winning but INC closing the gap, INC coming very close to winning in Chhattisgarh and INC winning in Rajasthan.

Going state by state we have (in terms of seats and estimated voted share)

J&K
BJP-PDP will have an alliance which actually works against them as their voting bases are not compatible and lose votes to INC-NC.  The support for BJP is skewed which works against them in terms of seats.

2019  
NDA: 1 (1 BJP, 0 PDP) ~42%
UPA: 5 (3 NC, 2 INC)  ~43%

2014
NDA:       3 (3 BJP)               32.6%
UPA:       0 (0 NC, 0 INC)       34.3%
Others:   3 PDP                     20.7%


HP:
Mostly replicates 2017 HP assembly elections with a bit of anti-incumbency that is somewhat mitigated by Modi's popularity in the Hindi heartland

2019
NDA: 3 BJP   ~53%
UPA: 1 INC   ~45%

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


Haryana
3 way battle between BJP INC and INLD.  AAP also will be in the hunt.  BSP will back INLD.  BJP loses ground from 2014 assembly elections but somewhat mitigated by Modi popularity in the Hindi heartland.

2019
NDA:      5 BJP                                 ~34%
UPA:      2 INC                                 ~27%
Others:  3 INLD-BSP (3 INLD 0 BSP)  ~29%
             0 AAP                                   ~7%

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 mostly to NDA since the 2017 assembly elections.  NDA will do better in relative terms when compared to 2017 assembly elections.

2019
NDA:    7 (6 SAD, 1 BJP)    ~38%
UPA:     6 INC                   ~37%
Others: 0 AAP-LIF             ~18%
            0 BSP                     ~3%

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      ~44%
UPA:  0 INC     ~23%
Others: 0 AAP  ~30%

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    ~48%
UPA:     1 INC    ~42%
Others  0 BSP      ~8%

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


UP
This is the big one.  I assume some swing away from BJP due to anti-incumbency from its 2017 assembly election victory.  I also assume that SP-BSP-RLD-PECP-NISHAD alliance will be formed.  INC will have a tactical alliance with this new mega front.  INC will run in only half the seats in UP with this mega-front not running in 7 seats where INC is the clear alternative to BJP.  These alliances are so broad in terms of social breath that some votes will be lost to rebels as well as the BJP.  Net effect is BJP from (BJP AD(S) and SBSP) actually gains votes relative to 2014 but loses a bunch of seats due to the consolidation of the anti-BJP vote.

2019
NDA: 35 (32 BJP 2 AD(S) 1 SBSP)                                          ~45%
UPA:   4 INC                                                                          ~7%
Others: Grand alliance 41 (25 SP 15 BSP 1 RLD 0 NISHAD)      ~42%

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


Bihar
On paper the NDA now have a very large coalition with BJP-JD(U)-LJP-RLSP.  Most likely not all these parties can be accommodated and RLSP will go over to UPA.  UPA will counter with RLD-INC-RLSP-HAM-LJD-NCP-JMM-CPM-CPI.  HAM is the JD(U) splinter that was with NDA in the 2015 assembly elections but already went over to UPA.  LJD is the pro-RJP JD(U) faction led by Sharad Yadav.  In theory NDA has the numbers to sweep the polls.  In reality having JD(U) come over to NDA also brings with it anti-incumbency and the result will be an almost even split between NDA and UPA.

2019
NDA:  22 (15 BJP 4 JD(U) 3 LJP)                                    ~45%
UPA:   18 (12 RJD 4 INC 1 LJD 1 NCP 0 HAM 0 RLSP)      ~44%

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
Assuming 2018 assembly results are a BJP victory with a reduced margin we should expect the 2019 LS election to mimic that result with a small swing toward BJP due to Modi's popularity in the Hindi heartland.

2019
NDA:  21 BJP     ~51%
UPA:    8 INC     ~41%

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


Rajasthan
Rajasthan is very elastic and always alternates in power each election.   Here I expect INC to win a narrow victory in 2018 assembly elections and for the LS election results to mostly mimic the assembly election result with a small swing toward BJP due to Modi's popularity in the Hindi heartland.

2019
NDA: 11 BJP      ~47%
UPA:  14 INC     ~46%

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


Chhattisgarh
Chhattisgarh has been trending INC last few election cycles and I expect a narrow BJP or INC victory in the 2018 assembly elections.  Former INC CM Ajit Jogi has split off and formed CJC but should have fairly limited impact beyond hurting INC in a couple of seats.  Modi has been popular here both with the OBCs and tribals and should give the BJP an edge here.

2019
NDA:  8 BJP       ~47%
UPA:  3 INC       ~42%

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


Jharkhand
Here NDA has gained an ally since the 2014 LS elections in the form of AJSU.  But that is counted by anti-incumbency since the 2014 assembly elections plus a grand alliance of JMM-INC-JVM-RJD.  The result should be mostly a draw.

2019
NDA:  7 (7 BJP 0 AJSU)                         ~43%
UPA:   7 (3 JMM 3 INC 1 JVM 0 RJD)       ~44%

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

Gujarat
I expect the election result to mimic the 2017 assembly election with again a small swing toward Modi due to the favorite son affect and Modi's popularity in the Hindi heartland.  INC should get some seats based on an INC-BTP-NCP alliance to consolidate the anti-BJP vote.

2019
NDA:  21 BJP                                ~53%
UPA:    5 (4 INC 1 BTP 0 NCP)        ~43%

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


Maharashtra
Here the old NDA parties BJP and SHS have, if anything, expanded since 2014.  Main issue for BJP here is SHS pretty much has decided to go off on their own given their fear that BJP is trying to displace them.  Also some alliance swaps took place. SWP which was with NDA in 2014 has go over to the UPA while BVA which was with UPA in 2014 has go over to the BJP.  I also assume that UPA will get PWPI to join UPA.  The net affect of BJP and SHS splitting and INC-NCP alliance holding is that INC-NCP will gain seats even if their vote share might stagnate or decline.   SHS splinter NMS which also has turned hostile to Modi recently could end up in an alliance with SHS which I will assume for now.

2019
NDA:                 14 (13 BJP 1 BVA)                             ~32%
UPA:                 27 (11 INC 13 NCP 2 SWP 1 PWPI)      ~37%
Others:  SHS+     7 (7 SHS 0 NMS)                             ~23%

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
I expect the results to match the 2016 assembly election with some swing away from BJP due to anti-incumbency as well as infighting within NDA.

2019
NDA: 1 BJP     ~51%
UPA: 1 INC     ~42%

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


WB
Here I expect a BJP surge mostly at the expense of the Left Front and to some extent AITC.  INC and Left front will form a tactical alliance where INC will back Left Front in most seats and Left Front will back INC in a few seats with some seats where both are strong having both contesting.  The INC-Left Front tactical alliance are able to save a few seats (mostly to the benefit of INC) while the BJP surge is not enough to gain BJP a large number of seats, yet.  GJM has called off its support for BJP and switched to AITC which means BJP will lose Darjeeling to AITC.

2019
NDA:        6 BJP            ~27%
UPA:        6 INC              ~7%
Others:  29 AITC           ~37%
              1  Left Front    ~26%

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


Orissa
Here BJP is emerging as the main opponent to BJD with INC losing support fast to both BJP and BJD.  BJP will gain a lot in terms of vote share an seats but not enough to displace BJD as the number one party.

2019
NDA:      8 BJP     ~37%
UPA:      0 INC     ~15%
Others: 13 BJD    ~43%

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


Karnataka
Here I expect INC to emerge as the largest party in the 2018 assembly elections if not a majority.  One key aspect is the relative revival of JD(S) which will un-do some anti-INC tactical voting for BJP in INC-BJP marginal seats.   Due to the distribution of vote share support BJP will hold a seat advantage even with similar vote share.

2019
NDA:   14 BJP      ~38%
UPA:    10 INC     ~39%
Others:  4 JD(S)  ~20%

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


AP
Here the end of TDP-BJP alliance puts TDP in trouble relative to YSRCP.  The TDP-YSRCP battle of the special status for AP has poised the environment so much that no party will ally with BJP.  TDP also has to deal with the fact that JSP which backed TDP in 2014 will now strike out on its own.  As a result I assume that a shocking TDP-INC alliance will be formed to prevent a total TDP meltdown.  Such an alliance could not prevent the loss of seats to YSRCP but does limit the damage.  Post-election YSRCP could end up joining forces with BJP.

2019
NDA:       0 BJP                      ~4%
UPA:        9 (7 TDP 2 INC)     ~40%
Others:  16 YSRCP                ~43%
              0 JSP                     ~11%

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


Telangana
Here the fall of TDP-BJP alliance means at race between INC and TRS to win over TDP.  I assume that INC will win that race.  YSRCP has completely pulled out of the state.  INC-TDP will fight TRS to a draw.

2019
NDA:    1 BJP                       ~17%
UPA:     7 (4 INC 3 TDP)       ~39%
Others: 8 TRS                      ~36%
            1 AIMIM                    ~4%

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
TN in politics is in turmoil.  AIADMK has splintered since 2016 after the death of Jayalaliatha.  The Sasikala faction led by TTV Dhinakaran has created AMMK which will cut into AIADMK votes.   Also film Rajinikanth has formed his own new party which like DMDK a decade ago I would expect to capture around 10% of the vote.  Rajinikanth's party name is not announced yet but I assume it will be RMM which is the name of his fan club.  Rajinikanth views are closer to BJP most likely will ally with BJP (this is what BJP wants) although BJP's reputation in TN is fairly low given their repeated intervention in the AIADMK civil war which TN votes view as an infringement of TN autonomy.  To some extent the rise of AMMK also stems from the same trend.  The winner of all this is clearly DMK which has recreated its alliance with INC and will get other pro-DMK parties like VCK PT MMK to continue their alliance.  The icing on the cake is getting MDMK to join up with the DMK front.  PMK and DMDK will also contest either together or separately but that will only split the anti-DMK vote. With AIADMK splintered DMK bloc is poised to sweep the polls given the uniform swing nature of TN.

2019
NDA:      0 (0 RMM 0 BJP)                                 ~13%
UPA      33 (23 DMK 5 INC 3 MDMK 1 VCK 1 PT) ~37%
Others:   5 AIADMK                                          ~30%
             0 AMMK                                               ~9%
             1 PMK

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
Here I expect some anti-incumbency to hit the Left Front since the 2016 assembly elections but while UPA will gain from Left Front UPA will lose more to a surging BJP.  BJP has lost its alliance with BDJS but I assume it will regain it by the time of the LS election.  BJP will win its first seat in Kerala.  KEC(M) has broken has INC but I expect by the LS election for KEC(M) to come back.  

2019
NDA:       1 BJP                                          ~15%
UPA:       7 (3 INC 2 MUL 1 KEC(M) 1 RSP)   ~41%
Others:  12 Left Front                                 ~41%

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
BJP is stronger than 2014 LS election given its new allies (AGP and BPF).  On the other hand there is anti-incumbency since the 2016 assembly elections.  As a result I expect the result to be evenly matched between NDA and UPA.

2019
NDA:       6 (5 BJP 1 BPF 0 AGP)       ~43%
UPA:        5 INC                              ~36%
Others:    3 AIUDF                          ~14%

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 couple of seats

2019
NDA:     3 (3 BJP 0 AINRC)
UPA:     3 (2 INC 1 NCP)

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



11 Northeast seats
I expect a complete NDA sweep here given recent assembly elections results there as well as the fact these backward states tend to vote for the incumbent national party given its dependency on federal subsidies.

2019
NDA:    11 (6 BJP 1 NDPP 1 NPP 1 UDF(M) 1 UDP 1 SDF)
UPA:      0 INC
Others:  0 Left Front

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: 217 (194 BJP)  ~35.5% (~31.5% BJP)
UPA   182 (101 INC)  ~30.5% (~21.5% INC)

Back in 2014 it was

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

which is a major setback for BJP.  Note that BJP vote share of ~31.5% is nearly identical to 2014 but losing key allies as well as consolidation of the anti-BJP vote brings down its seat share dramatically.  The distribution of BJP vote is also not as favorable.  INC vote share of ~21.5% is only 2.0% higher than in 2014 but the right alliances gives it a lot more seats along with a more favorable distribution of votes.  

[jaichind]

It seems that in UP 4 BJP Dalit MPs are openly attacking the BJP leadership over treatment of Dalits and most likely will not be renominated.

https://news.webindia123.com/news/Articles/India/20180417/3323177.html

It is pointed out that all 4 were defectors from BSP to BJP and won on the BJP ticket in 2014 are looking to defect back to BSP now that with the SP-BSP alliance it is likely they will lose re-election.  All 4 it seems were able to pull in Dalit votes in 2014 but given the issues with Dalits and the current UP regime that seems unlikely to repeat itself.  So in order come up with a reason to defect back they came out openly to attack the BJP leadership. 

This would be a good data set to back-test my predictions since all 4 BJP Dalit MPs view their seats are likely to be not winnable in 2019.    The 4 seats in question are (along with 2014 and my 2019 model results)

Etawah
2014
BJP  46.8%
SP   28.4%
BSP  20.5%
INC   1.4%

2019 (INC withdraws from this seat)
BJP  47.9%
SP   46.6%


Nagina
2014
BJP  39.0%
SP   29.2%
BSP  26.1%
PECP  2.3%
MD    1.4% (INC ally)

2019 (INC withdraws from this seat)
SP   52.8%
BJP  40.9%


Robertsganj
2014
BJP  42.7%
SP   21.1%
BSP  15.3%
INC   9.7%
CPI    2.8%

2019 (INC runs here)
BJP   43.4%
BSP  38.7%
INC    7.8%


Bahraich
2014
BJP  46.3%
SP   36.1%
BSP  10.4%
INC   2.6%

2019 (INC withdraws from this seat)
SP   47.5%
BJP  45.9%


So under my model BJP actually would win 2 of the 4 seats (1 of them narrowly.)  In Robertsganj had INC not run a candidate under my model BJP would also narrowly win.  So these Dalit BJP MPs, if anything, is slightly more negative than my model on the prospects of BJP in UP against a SP-BSP grand alliance.