Karnataka polls may turn out to be anybody’s game

Opinion polls have started making their predictions ahead of the Karnataka elections next month. The latest one, conducted by India Today and Karvy Insights, predicts a hung assembly, with the Congress getting 90-101 seats and the Bharatiya Janata Party (BJP) 78-86 seats in the 224-member assembly. The Janata Dal (Secular), or JD(S) is forecast to win most of the remaining seats, according to the poll.

Readers might recall that there were a couple of splits in the BJP in Karnataka prior to the assembly elections in 2013, following a series of corruption allegations against former chief minister B.S. Yeddyurappa. He left the BJP to form his own party called Karnataka Janata Paksha (KJP). B. Sriramulu, a member of Yeddyurappa’s cabinet, also broke away from the BJP to form his own party (called BSR Congress). 

These splits hurt the BJP hard in the 2013 elections, reducing it to 40 seats, tied with the JD(S) in second place. The Congress won a comfortable majority, giving Karnataka its first full-term government since 1985. 

KJP and BSRC subsequently returned to the BJP fold ahead of the 2014 general elections, with Yeddyurappa and Sriramulu both winning Lok Sabha seats on BJP tickets. The party rode the Modi wave to win 17 of the 28 Lok Sabha seats from Karnataka. 

Election Metrics had remarked after the 2013 elections that “our calculations show that had the BJP, the KJP and the BSRC fought the polls as one party, the combine would have secured 87 seats while the Congress would have got only 91 seats—leading to a hung assembly”. The numbers predicted by the India Today-Karvy Insights combine are remarkably close to this! 

In other words, if the split and re-merger of the BJP are to be taken into account, not much has changed in Karnataka since 2013. Even the vote share predictions by India Today and Karvy are not very different from what the parties had obtained in 2013. 

As with any election in India, what makes forecasting tricky is the conversion of votes to seats, as Election Metrics has noted several times in the past. Given the first-past-the-post polling system in India, the distribution of votes across constituencies matters more than the absolute number of votes a party gets. 

In Karnataka itself, in the 2008 elections, the BJP had trailed the Congress in terms of vote share (the two parties got 33.9% and 34.6% of the popular vote respectively), but nearly got an absolute majority, winning 110 of the 224 seats (the Congress won 79). The BJP had been helped by the fact that its support base was concentrated in the northern and coastal parts of the state, and around Bengaluru, while the Congress had more broad-based support. 

A series of political events, however, have resulted in traditional support bases being redrawn. Ahead of Rajya Sabha elections earlier this year, 7 legislators from the JD(S) defected to the Congress—they’ve all been given tickets by the Congress to contest the elections. Then, the Congress has played the caste card by proposing minority religion status to the Lingayats, a prominent caste which has traditionally backed the BJP. The BJP, on its part, has engineered defections from the Congress to break into the Old Mysore region where it’s been traditionally weak. 

While this might drastically alter distribution of votes for each party across constituencies, we should still go ahead and hazard a prediction on what it might take for any party to get an absolute majority (112 of 224 seats). In order to do this, we will assume that the distribution of votes across constituencies will follow the same pattern as in 2013, with the BJP assumed to get all of the KJP and BSRC votes. Then, we will calculate what proportion of total votes each party needs to get the most votes in 112 seats. 

Sparing the messy arithmetic (which is by no means elegant), our calculations show that the Congress needs a total of 40% of the popular vote to hit the magic number of 112. The BJP, on the other hand, needs only a total of 37% of the popular vote to hit 112 if the distribution of the party’s votes follows the same pattern as in 2013.

Putting it another way, the Congress got 36.6% of the popular vote in 2013, implying it needs a 3.4 percentage point swing in its favour to win an absolute majority. The three parties that now form the BJP together got 32.5% of the vote in 2013, again implying a 3.5 percentage point swing for the party to win majority. 

Three-and-a-half percentage points is not a small swing, but we’ve seen such swings in other states in the recent past. The Karnataka polls could yet be anybody’s game.