Uttarakhand Assembly Polls 2017: How Credible Congress Internal Polls?

BC Khandhari


Uttarakhand Congress Chief Kishore Upadhyay in a press conference last month announced that a pre-poll internal survey conducted by the Congress in the state has predicted that the party will win 33 out of 70 seats in the state. Apart from projecting the party's victory in 33 seats, the survey also puts it in a formidable position in another 10 seats. Interpreting it as an indication that the Congress may win 43 seats, Upadhyay said if it happens it will be the biggest electoral victory for a party in the history of the state.

2012 ASSEMBLY RESULTS
Party
Vote Share             (%)
Seat Share           (No)
INC
33.79
32
BJP
33.12
31
BSP
12.19
1
UKD
1.93
1
Others
12.34
3

The Congress did not divulge their vote share projections or those of others. During the 2012 Assembly polls, the Congress registered around 34% and secured 32 seats.  Accordingly to reach 33-43 seats, it is a reasonably safe assumption that the Congress needs a vote share ranging at least between 34-40%.

So are these projections of their internal polls realistic or just spin? Can this be validated by the clutch of published polls? So let’s look at them.

POLL
Seat Share (No)

Vote Share (No)
BJP
INC
Others
BJP
INC
Others
Undecided
UTTARAKHAND POST
29
36
5
N/a
N/a
N/a

LOKNITI
39           (35-43)
26              (22-30)
0
40
33
23

AXIS
40
28
2
45
33
22

CVOTER
22      (20-24)
36           (34-38)
8             (6-10)
N/a
N/a
N/a

VDP ASSOCIATES*
40
24
6
39
34
13
7%
HAMSA
38            (37-39)
28            (27-29)
4
N/a
N/a
N/a

Av/Range
31              (22-40)
31          (26-36)
4            (0-8)
42          (39-45)
33
18         (13-23)
-
 
When analyzing published polls, it is useful to keep in mind the following:

-        VDP Associates were the only polling agency that revealed that at least 7% of their respondents remain “undecided”. This does not mean that other polls did not confront the problem of ‘undecided’. It only means that either these were not revealed or had been statistically treated to assume their polling preferences to one or other political parties. One of the major reasons pollsters experience discrepancies between their poll predictions and subsequent election outcomes the world over is the proportion of undecided respondents in their poll samples. We witnessed this phenomenon in the recent US Presidential elections too. This category of “undecided” could often be a game changer and usually includes within it sub-categories not willing to declare their voter intentions. Mostly socially vulnerable groups like Dalits belong to the latter category which are one of the reasons the vote share of parties like BSP are often grossly under-estimated by polls. Lower in the caste hierarchy, the lower the probability of being fully articulate of one’s political preferences or dislike

-         Most election forecasters in India agree that the trickiest part of forecasting an election is conversion of vote shares to seats. In India's first-past-the-post elections, it is a nightmare to convert vote shares to seat shares. In the UK, which follows the same electoral system, pollsters are bold on vote share forecasts but are usually discreet about predicting numbers of seats. But last year, UK pollsters threw such discretion to the wind to mostly predict a Labour victory, only to end up red faced to find David's Cameron's Conservatives chalk out a much bigger second term victory. Yashwant Desmukh of CVoter argues that in India it is not so much the case of pollsters but TRP hungry TV News channels that hype seat shares. We see this problem of vote-seat conversion within the clutch of published polls for Uttarakhand too. VDP Associates projecting BJP 40 seats with a 5% lead, while Axis projecting the same number of seats for the BJP despite a whopping 12% lead.  

-        In any elections, a handful of polls could be dismissed as statistical outliers. In the case of the clutch of published polls for Uttarakhand, the Axis poll apparently stands out as a sore thumb. It is not so much that outlier polls cannot turn out right. They could in occasions, like Chankaya’s 2014 Lok Sabha poll projections. But the probability of them doing so is usually statistically very low. But the Axis poll projections additionally look extremely suspect for another reason. Despite a whopping 12% vote share lead for BJP, the poll projected just a 12 seat lead for BJP instead of a Tsunami Wave wiping out the seat share of Congress all together. This in a state like Uttarakhand where historically, small vote differences resulted in huge seat margins. 

By and large, polls in India tend to over-estimate the BJP and under-estimate opposition parties including the Congress. A classic example is Bihar when even their exit polls missed the mark by a mile. Exit polls internationally are expected to conform to standards of + 1% Margin of Error (MoE) as compared to pre-polls where a MoE of 3-4% are considered acceptable as a standard. Even in 2012 Lok Sabha as the graph below illustrates, pollsters once again demonstrated this tendency. 
 

Despite all these drawbacks, what do the vote shares of these published polls collectively tell? This blog had not been able to get full data on vote shares. But from those available, the following could be concluded:

-         The prediction for Uttarakhand is a mixed bag, though more polls tend to bet on the BJP winning the state, and smartly too.

-         The Congress retains its 2012 Assembly vote share of around 34%, even in the case of outlier Axis poll. In fact, despite the Modi Wave of 2014 that also encompassed Uttarakhand, 34% vote base of the Congress had not breeched. This is suggests an absence of serious anti-incumbency viz non slippage of votes. 

-          In contrast the BJP viz-a-viz Lok Sabha 2014 has suffered a whopping vote erosion projected between 10-15% depending on which poll viz. the party is experiencing massive negative momentum.

-        If the Axis poll, is ignored being a statistical outlier, the poll of polls suggest BJP’s vote share around 39% and the Congress 34% - just a difference of 5%. Though opinion polls in India claim a + 3% margin of error (MoE), their track record most often reveals their MoE is at least + 5%. 

-          Given that polls tend to over-estimate BJP and under-estimate the Congress, and if this factored, in the best of scenarios, the Congress could be even marginally ahead. Even if otherwise, the Congress main challenge would be to attract at least a 3% vote swing from the BJP that could enable to catapult them to sneak past the finishing line. Is that possible? Looks like if Prashant Kishore is advising their campaign and he would not have taken up the assignment if he felt Congress did not have a ghost of a chance like he did turning Assam assignment down. Kishore was behind Modi’s brilliant campaign in 2014 Lok Sabha polls and Nitesh Kumar led alliance win in Bihar. 

-       Further, apart from 2014 Lok Sabha polls, the contests in this state have always been razor thin close. For example TV channels had predicted a BJP victory in 2012 on the day of counting, based on leads; however, Congress emerged as the ultimate winner.  


Party
Vote Shares (%)
Vote Swing (%)
2012 (Actual)
2014 (Actual)
Dec 2016
Jan 2017
From
2012
From
2014
From
Dec to Jan
BJP
34
55
40
39
+5
-16
-1
Congress
34
34
34
36
+2
+2
+2
Others
32
11
26
25
-7
+14
-1


This is the latest CSDS poll (31st January 2017) based on mid January field data before the campaign had actually started. What we find is the BJP continuing to lose vote share (momentum is negative) while the Congress is gaining momentum and demonstrating pro incumbency (+2% viz-a-vis both 2012 and 2014). It is gaining both from the BJP as well as others.What is most heartening news for the Congress is the Muslim-Dalit consolidation behind it which should further accelerate. The bad news for BJP is that it is losing the Dalit votes it attracted in 2014. If this trend continues, the state can allude a win for the party.

Accordingly, prima facie the Congress claim of 33-44 seats based on their internal survey could well be realistic and most likely that the state could witness a cliff hanger battle. This could be the key significant battle of this 5 state assembly poll season. If on the other hand, the published polls of media are in the right direction, the BJP should be able to coast away to most probably a landslide victory. The X Factor will be turnout. A lower turnout (base 2014) should favour the Congress.

Below given in the tabular form are the strengths and weakness of both the BJP and Congress. The battle for Uttarakhand looks a bit fuzzy but primarily revolves how well or badly Mayawati of the BSP performs. 

BJP
What’s favourable going on?
What’s unfavourable going on?
1. BJP is in the right side of historical trends. Like Kerala, Uttarakhand, public throws out the incumbent government every five years. And this has happened since inception of the state with Congress and the BJP taking alternate turns in ruling the state.
1. No CM candidate
2. Out of the 5 states going to poll, only Goa and Uttarakhand offer the BJP a reasonable chance of winning. With chances in Uttarakhand much higher than Goa, the central leadership is pulling all stops to win the state.
2. BJP has no leader of stature that can match Congress Harish Rawat in the state.
3. With the exception of 2014 Lok Sabha, the state is highly polarised between support for the BJP and Congress respectively in somewhat equal strength
3. Faces a double whammy – no perceptible anti-incumbency at the state level against the Congress while experiencing huge anti-incumbency at the central level. The BJP is finding difficulty to hang on to whatever little lead it has over the Congress. Not only has the Modi Wave fast dissipating, but Modi himself is now a shade of himself as a vote catcher 2014.
4.  Though crowds seem to have deserted Modi in other states, he still drew a somewhat sizeable decent crowd at his recent Parivartan Rally in Dehradun. This could indicate that unlike other states going to poll, BJP’s core support base  is more or less intact and demonstrate no evidence of total collapse as compared to other states
4. The BJP is deeply faction riven and this is embodied by the fact that it has 4 CM hopefuls. Rebellion and no cooperation from cadres could prove very costly to the party, particularly in seats where the contest is very fierce.
5. BJP is betting on replicating the successful Assam formula in Uttarakhand. Finding it unable to enthuse the electorate by themselves, they try to weaken the opposition and by giving tickets to defectors who are assumed to have high connect with voters, augmenting BJP voter strength, to boost their strike rate. 
5. The party is unable to come out with an innovative message of change in their campaign. The Assam strategy from media reports doesn’t seem working as it seems have widely demotivated and alienated the loyal party workers, upset at defectors being given both prominence within the party and given tickets. The ex Congress CM, Bahuguna has reportedly emerged one of the centre of rival powers, increasing factionalism
6.  The party strategy is to win two of three regions of the state, targeting Garhwal and at least one more, transcending traditional Thakur-Brahmin support base to woo the Dalit vote, failing which, hoping they go to Mayawati rather than boost Congress vote share. The BJP should do well in Garhwal region. Standing no chance in Kumaun Hills, the battle for Uttarakhand comes down to the Tera/Maidani region.
6 But the party is finding immense difficulty in transcending its traditional Thakur-Brahmin base. Dalits are leaving the party en masse is one of the reasons why BJP is finding themselves in fast decline. To stem this erosion, the party opted to induct leadership of this community by engineering defection mainly from the Congress. The backup plan also seems failing as Mayawati appears just a shadow of herself in the state as compared to 2012.
7.  The party also finds itself unable to hold on to a section of Muslim votes it attracted in 2014. Fearing that the Muslims will consolidate around the political force best placed to beat the BJP, their hope centres around Mayawati mopping these votes. But again Mayawati, if she is not able to hold on to Dalit votes, is highly unlikely to attract Muslim consolidation so it is advantage Congress
 
CM Preference (CSDS-ABP News survey: Jan 31st 2017)
Harish Rawat
31
BC Khandari
14
Other BJP leaders
18
Other Congress Leaders
2
Leaders from other parties
1
Not divulge preference
34

CONGRESS
What’s favourable going on?
What’s unfavourable going on?
1. Harish Rawat has emerged the tallest leader not only within the Congress but the entire state of Uttarakhand. Most polls rate him as the most preferred CM candidate among those standing for elections – as much as with a 30% lead
1. Harish Rawat is embroiled with a corruption scam allegation, which taken a small toll on his personal image in the state
2. Despite some corruption allegations, the Congress Administration had delivered good governance by the state’s historical experience. This has prevented the Congress from anti-incumbency eroding its vote base. Rawat has introduced and delivered many populist schemes that not only help him retain public support but could hold the potential of increasing it.
2. The Congress party lost most of its top leadership. “Congress is practically finished in Uttarakhand. It is only Harish Rawat as an individual who would be fighting the BJP in the ensuing election”; says a former CM and BJP leader from the state.
3.  With most dissidents and potential turncoats already exported to the BJP, the Congress goes to poll as a much more unified party, with lesser dissent. The additional flip side was that Congress managed to get rid of most of its anti-incumbency against sitting MLAs and Ministers by dumping them on to the BJP and by inducting fresher faces as candidates.
3. The BJP did not engineer defections at random but specifically targeted the Dalit & Backward leadership of the Congress. These include a minister Yashpal Arya. Mr Arya, a prominent Dalit leader, the Chairman of the State Council for Backward Classes, Santosh Kashyap, and State OBC Cell chairman, Shekhar Kashyap. In July 2016, Modi, inducted BJP's Almora Lok Sabha MP Ajay Tamta, a prominent Dalit face in Uttarakhand, as part of the BJP's attempt to reach out to Dalits and OBCs.
4. Rawat is the strongman from Kumaun Hills and expected to clean sweep the region. The Congress strategy is to give a fight for Garhwal and win Terai/Maidan where Dalits (19%) and Muslims (15%) hold the key to winning. With Mayawati weakened and the contagion effect in the state of the SP-Congress Tsunami Wave in Uttar Pradesh, the Congress hopes to consolidate Muslim votes and grab a significant chunk of Dalit votes
4. Nine out of 10 Congress MLAs who rebelled and quit the party are also from Garhwal and its adjoining plains. Consequently, Congress is not expected to get the upper edge in Garhwal. It suits them to concentrate on Terai to win the state.

Like us on Facebook

Follow us on Twitter