Karnataka Elections: Making Sense of the ‘Tamasha’ of Exit Polls

Exit polls are generally more accurate than opinion polls in predicting the mandate, writes Amitabh Tiwari.

The exit poll results for Karnataka are out. The tamasha of exit polls continues and different polls have predicted different winners, most of them projecting a hung Assembly.

Axis and Chanakya are the two agencies that have the best track record, and they predict a split verdict. Axis predicts Congress will get between 106-118 seats while Chanakya predicts a Bharatiya Janata Party (BJP) win with 120 seats.

So, it is between Axis and Chanakya on the 15th. The ‘poll of polls’ shows the BJP ahead at 103, followed by Congress at 85 and the Janata Dal (Secular) at 32 seats.

Poll of Polls

Seat Tally:

Heaps of data and numbers have been thrown at us by about 9-10 agencies. In every election season, the names of new polling agencies crop up, and it’s difficult to decide which one to go by.

Exit polls are generally more accurate than opinion polls in predicting the direction of the mandate. By this logic, BJP seems to be better placed.

In this piece, I try to decipher the numbers behind the seat projections. I have selected three polls for this analysis; Axis, Chanakya and C-Voter. I have included C-Voter because they got a very close and waveless election of Tamil Nadu right where both Chanakya and Axis got it wrong.

Seat & Vote Share Projections

• Axis projects both BJP and Congress to gain 3 percent vote share each, at the expense of JDS (-3 percent) and others.

• Chanakya predicts BJP to gain 7 percent and Congress 3 percent vote share. JDS to lose 2 percent.

• C-Voter projects BJP to gain 9 percent and Congress 3 percent, JDS to lose 3 percent.

All the three polls are unanimous about JDS performance. It is losing significantly compared to 2013 (14-16 seats). As per these polls its dream of playing the kingmaker is over. If this is true, then a section of voters fearing a hung verdict have shifted to Congress (Axis) and BJP (Chanakya/C-Voter).

Axis and Chanakya predict inverted vote shares for BJP and Congress. At 39 percent Axis predicts 112 seats for Congress and at same vote share Chanakya predicts 120 seats for BJP.

This ties in with the general belief that the Congress vote is more spread out while the BJP’s is more concentrated – and BJP has a better vote share to seat share conversion ratio.

In my opinion the, C-Voter numbers don’t add up.

First of all, they have predicted only 3 percent for independents and others, which is very low, historically it has been in 8-10 percent range. Secondly, at 41 percent, BJP would easily touch around 125 seats, with a vote share to seat share conversion ratio of 3. Recall, in 2008, BJP won 110 seats with only 34 percent vote share.

Caste-Wise Voting

Before we proceed, let’s do a quick back-of-the-envelope check of whether the caste-wise voting figures of Chanakya and Axis tie in with the overall vote share. Recall, Chanakya had got this horribly wrong in Bihar. A basic check of BJP caste-wise support numbers gives me 38 percent for Chanakya and 34.5 percent for Axis, which reconciles with the numbers they have projected.

BJP Vote Share:

What both polls show is that the Lingayat issue has boomeranged and there is no benefit to Congress out of this. In fact, it has led to reverse consolidation and BJP has gained 3 percent to 4 percent votes from the community compared to 2013.

Both Axis and Chanakya also confirm that the BJP has gained significantly among non-Kuruba OBCs (20 percent+). BJP has been through innovative social engineering, bagging OBC votes across states – this is similar to UP where party bagged the non-Yadav votes. Siddaramaiah belongs to the Kuruba community.

INC Vote Share

Both polls show that Congress is losing votes among the dominant Lingayat and Vokkaliga community. The separate religion issue and Siddaramaiah AHINDA consolidation strategy seems to have resulted this.

The party is gaining significantly among Muslims (+30 percent) vis-à-vis 2013 as has been the pattern elsewhere in the country, this is also one of the primary reasons of JDS decline as per these surveys.

Party is losing support among the non-Kuruba OBCs, BJP’s allegations that Siddaramiah only worked for the benefit of his Kuruba community seems to have worked.

Both the polls differ in their SC/ST support of the two parties. SCs and STs have traditionally voted for the Congress, 44 percent and 39 percent respectively in 2013, and have influence over 51 reserved seats. Axis claims that their support is intact for Congress, benefitted by a general apathy of Dalits towards the BJP due to cow vigilantes, SC-ST Atrocities act fiasco, alleged torture of Dalits in BJP-ruled states etc.

Dalits are divided broadly into Holeyas, the right-wing faction, and Madigas, the left-wing.

According to reports, there is discontent brewing in the left faction, over what they see as cornering of various welfare programmes by the right wing and they are veering towards the BJP. Additionally, the party through the projection of Sriramulu as Deputy CM has been trying to woo the ST voters. Chanakya numbers confirm this theory.

If there is one thing which needs to be watched out for on 15 May, it is which way the SC/ST seats are going. That will decide the course of the election in 2018. Almost 3 percent of the vote share difference between BJP and Congress in the 2 polls is explained by the SC/ST vote share pattern. 

Regional Variations

Karnataka is a very heterogeneous state and a nightmare for pollsters. There are six regions, each having its own dynamics and they have historically displayed different voting patterns. Regional vote shares are not provided by Chanakya so I am using C-Voter numbers.

INC Regional Vote Share:

Both polls suggest that Congress is holding onto its Old Mysuru support base where it is locked in a direct contest with Deve Gowda’s JDS.

Same holds true for JDS, it is expected to garner 35 percent support (average of the two polls). It is seen losing support in Bengaluru, in line with the national phenomenon of urban voters flocking to BJP. Both polls suggest it will gain 5-10 percent in Hyderabad Karnataka which is party’s stronghold, Congress gaining at the expense of JDS.

It is also almost holding onto its vote shares in Coastal and Central Karnataka. In Bombay Karnataka, which is a BJP stronghold, Axis suggests Congress will maintain its support base, while C-Voter suggests a 5 percent increase.

This is intriguing; if Congress is gaining in both Hyderabad Karnataka and Bombay Karnataka and holding Old Mysuru fort as C-Voter suggests, then how come BJP is winning the state? Some issues with C-Voter data here.

BJP Regional Vote Share

BJP is seen improving its support among all regions in both the polls. In Old Mysuru, though BJP is gaining, Congress and JDS are holding onto their support base. BJP is gaining in Bengaluru (urban voters), in Bombay Karnataka due to the Lingayat issue and Yeddyurappa’s popularity.

The difference between Axis and C-Voter is mainly in Hyderabad Karnataka (-8 percent) and Bengaluru region (-6 percent).

C-Voter is predicting a higher vote share than Axis. As pointed out above, it is showing BJP gaining in a Congress stronghold and Congress gaining in BJP’s, which might be due to some data error. Coastal Karnataka usually swings each election, and this time is going to BJP.

Region-wise C-Voter data is inconsistent. In two regions, Hyderabad Karnataka and Bombay Karnataka, it is showing a neck-and-neck fight and in Old Mysuru, it is showing Congress ahead.

In both Bombay Karnataka and Hyderabad Karnataka, it is showing Congress gaining as well. These regions account for 154 seats, 69 percent of total assembly strength – despite this, it is showing a BJP victory. So we will have to wait for 15 May.

To conclude, it’s essentially between Chanakya and Axis on the 15 May. Enjoy your popcorn and movies today, sleep well, the trailers don’t reveal the climax of the battle of Karnataka.

(Amitabh Tiwari is a former corporate and investment banker turned political commentator, analyst and strategist. He tweets @politicalbaaba.)

(The Quint is now on WhatsApp. To receive handpicked stories on topics you care about, subscribe to our WhatsApp services. Just go to TheQuint.com/WhatsApp and hit Send.)