High stakes for the Congress: If the Congress wins in Karnataka, it can rightfully claim that the 2014 national election, and UP 2017, were two troughs from where it is manfully bouncing back.
That Karnataka is an important election is a foregone conclusion. How important, and how asymmetrically different the election implications are for the two major contenders, BJP and Congress (INC), is discussed below. High stakes for the Congress: If the Congress wins in Karnataka, it can rightfully claim that the 2014 national election, and UP 2017, were two troughs from where it is manfully bouncing back. It will be able to argue, somewhat convincingly, that the worst is over—and that Rahul Gandhi is emerging as a viable leader for the Congress, and a viable alternative for the Indian electorate in 2019. If the Congress gets the maximum number of seats, but doesn’t form the government (i.e., BJP and JDS form a coalition), the Congress victory would not be as sweet, but it can live to fight another day—especially in Rajasthan and Madhya Pradesh, and perhaps even in Chhattisgarh. But it would have been weakened, especially relative to its recent “peak” after the Gorakhpur victory, and the by-polls in Rajasthan.
But what will likely happen if the Congress is in second place in terms of seats—and BJP is close to a majority on its own (our most likely forecast)? One of the statistical “facts” about this three-party fight is that if the JDS performs as expected by virtually all the opinion polls (around 30 seats, but 10 less than what it obtained in 2013), then the only way that a party can win outright is if it wins between 50-90 % more seats than the second-placed party. The strength of the Karnataka assembly is 224 seats, and independents are expected to get around six seats. A majority is 113 seats, and if JDS obtains 30, then what is left for the second-placed party is (224- 6-113- 30) or 75 seats. This implies a ratio of seats (#1/#2) as 113/75 or 1.5.
Now, let us tweak the seat-shares just a little bit. Assume the winning party gets 120 seats, and JDS gets 35; the second party will obtain 64 seats; the above ratio jumps from 1.5 to 1.9. The point of this numerical exercise is to emphasise the non-linearity of expected outcomes. And it is a rare election that is linear, close, and fully expected!
If Congress is in second place—as our calculations suggest it will be—and if the BJP wins 120 seats, and Congress less than 70 seats, what happens to Congress’s (mis)fortunes going forward? There has been a concerted expectation, especially in Congress quarters, that the tide has turned, and that Rahul Gandhi is a transformed and dynamic leader. A bad showing in Karnataka would mean that the Congress will go into a rethink of expectations, and perhaps even the leadership of the Gandhi family.
It is easy to build up expectations and rally the faithful (and those at the margin) when you have won Gorakhpur and by-polls. It is quite another task to make believe that with only two states under your belt (one, Punjab, with 13 seats and a Union territory, Puducherry, with 1 seat) that you are a contender (with apologies to Marlon Brando in On the Waterfront, “I coulda been a contender”).
One additional point—in our travels in Karnataka with the travelling band of limousine liberals (with apologies to gypsies), we found reasonable support for JDS, and Congress, especially in the south (around Mysore). But, and this is a relevant but, several JDS and Congress supporters said they would be voting for the local leadership in the assembly election, but would be voting for BJP, sorry Modi, in the national election. If one goes past the national English dailies, it does appear that Modi is very popular; the reputed Pew survey shows that Modi has retained his high popularity ratings in the mid 80s! So even in the small chance that Congress wins a majority, it will only be a small dent in Modi’s appeal.
Math behind the forecast Curiously, one very relevant aspect of the Karnataka election has not been emphasised by either the pollsters or the analysts. What we all know is that both BJP and the JDS won 40 seats apiece in the 2013 election, and the Congress won 122 seats. But how many of us know that if the 2013 election is adjusted for BJP “allies”, then the election was indeed very close—83 seats for the BJP to Congress’s 89, and JDS’s 31. In other words, the JDS lost only 9 seats when BJP fought as one unit, and the INC lost 33 seats. Who are these mysterious allies that we have made the adjustment for? Well, none other than former BJP chief minister BS Yeddyurappa who had formed his own party in 2012 (after being thrown out of the BJP) and fought the 2013 election as leader of the KJP (Karnataka Janata Paksha). In addition, Sriramulu, the leader of the small BHRC party, came out in support of the BJP in 2014.
If the votes of these allies are added to the 2013 BJP-vote (this assumes that a KJP & BHRC voter will vote for the BJP rather than INC or JDS—a reasonable assumption), then the 2013 election appears as BJP 32.4% of vote, and INC slightly ahead at 36.6%. In the 2014 Lok Sabha election, the BJP increased its share of the vote by 10 percentage points (the Modi effect) and increased its seat tally to 132 assembly constituencies. [Data are available at the assembly constituency level for the national election]. This additional 10 percentage points added 49 seats to the BJP tally.
A well-documented result around the world is that the political party that wins the national election, loses in subsequent local elections. A typical loss percentage for India is around 4-6 ppt. Which implies that with a 5% loss in vote-share from the 2014 LS election, the BJP can lose 25 seats, i.e., obtain 107 seats in 2018. A 5% loss in BJP vote-share over 2014 means that the BJP will obtain 38% of the vote—not very far from the 35% vote share envisaged by the average opinion poll. This is a back of the envelope calculation. More refined models (and hopefully more accurate!) converge on BJP obtaining seats in the range 107-125, or, on an average, 116 seats.
The respective seat shares of INC cannot be estimated in as straightforward a fashion. But JDS seat share around 25-30 looks likely. Which implies that the Congress obtains around 74 seats—the fourth lowest (and equal to) what Congress obtained in the first assembly election in 1952. One of my favourite forecaster sayings is the following: Forecast often, and always remind people when you are right. If I am near-correct, I will come back to remind; if I am drastically wrong, I will work towards improving the models, and do so in hiding!
Senior India analyst at Observatory Group, contributing editor, The Financial Express, and part-time member of PMEAC
Views are personal Twitter: @surjitbhalla