Lok Sabha elections 2019: Will BJP be able to win back key heartland states from Congress?
In 80% of the cases (18 of 23 GE races), the party that won the state election went on to win the state in the following general elections. Only in five cases did the winner differ and no particular state is affected by it. It happened twice in Uttar Pradesh and once in Rajasthan, Madhya Pradesh and Karnataka.Updated: Apr 25, 2019 12:35 IST
While the outcome of this election remains unpredictable, it is becoming quite clear that the fate of the Bharatiya Janata Party will be determined in a few large states, where it is challenged either by the Congress or regional parties. In the Hindi belt, the states of Uttar Pradesh, Madhya Pradesh and Rajasthan count for 134 seats, of which the BJP currently holds 118.
In order to maintain its position, the BJP must gain seats in other large states such as West Bengal (42 seats) and Maharashtra (48). Karnataka, with its 28 seats, is likely to be split more or less evenly between the BJP and the INC-JD(S) alliance, thus offering little scope for expansion.
In Madhya Pradesh, Rajasthan and Chhattisgarh, the Congress is feeling confident that it can repeat its good performance of last December, when it defeated the BJP in assembly elections. Conventional wisdom has it that a party that wins a state election within a year of a general election repeats the performance in the following election, usually with even higher vote shares.
In this article, I look at 23 pairs of elections that have taken places since 1998 in six major states where the assembly election closely preceded the general election: Madhya Pradesh, Rajasthan, Uttar Pradesh, Chhattisgarh, Gujarat and Karnataka).
In 80% of the cases (18 of 23 GE races), the party that won the state election went on to win the state in the following general elections. Only in five cases did the winner differ and no particular state is affected by it. It happened twice in Uttar Pradesh and once in Rajasthan, Madhya Pradesh and Karnataka.
What is interesting is the party variation. The BJP is the winner in 15 of those 18 cases of symmetric victory. It hasn’t lost a single state that it previously won in an assembly election, among those six states. The Congress, on the other hand, has lost in a general election after having won in the preceding assembly election on three occasions: Madhya Pradesh in 1998, Rajasthan in 1998 and Karnataka in 2013. The two other cases happened in Uttar Pradesh and concerned the SP and the BSP. In other terms, there is a greater spill-over effect when the BJP wins the preceding state.
Besides, the data suggests that the BJP tends to increase its vote share in the following general election, boosted by its previous recent performances.
Over the past four election cycles, the BJP in Madhya Pradesh has gained on average 7% of vote share in the following general election.
What might explain this variation between the Congress and the BJP? Generally speaking, there are three major factors that affect the spill-over effect. In a 2016 paper, Gareth Nellis finds that the spill-over effect is stronger when there are no visible rifts within parties (he looks at the reverse relation, from general to assembly elections).
Parties that appear cohesive and that have a firmer grasp on their organization tend to benefit more from a spill-over effect than parties that suffer from apparent factionalism.
Across states, the BJP appears to be more cohesive than the Congress, which has visible factions and tends to take its internal disputes to the public.
States with a bipartisan party system seem to see greater spill-over effect. The absence of triangular or quadrangular fights help parties to consolidate their gains. And similarly, states with regular alternation of majorities also have strong spill-over effect, as the swing effect that benefits the challenger usually continues for a short while after the state election has taken place.
By contrast, party factionalism, or internal divisions, tends to reduce the spill-over effect. The presence of a large number of close contests (defined as less than 5% of victory margin) can also reduce spill-over, as small number variations can have large effects in terms of seat distribution.
States where the main competing parties have similar vote share, and states with multipartisan party systems, also have lesser spill-over effect.
In this election, the Congress is clearly located in the second configuration.
Its organisation at state level remains weak and marked by divisions, between the Kamal Nath, Jyotiraditya Scindia and Digvijaya Singh’s factions. The Congress in Rajasthan also has a visible rift between the current chief minister’s faction and Sachin Pilot’s followers. Many in the party are also bystanders.
More importantly, the vote shares of the Congress and the BJP in these two states were almost identical. Both parties won many seats with razor thin margins; 32% of the seats in Madhya Pradesh were won with a margin inferior to 5% (73 out of 220), against 35% of the seats (69/199) in Rajasthan. In both cases, the BJP won more of these close races than the Congress (37 in Madhya Pradesh and 28 in Rajasthan, against 32 and 22 for the Congress).
In the last five state elections, the Congress and the BJP have been neck-to-neck in terms of vote share in three states.
Other factors play a role too. Who is the chief minister, for instance, or the length of a government’s incumbency. National factors also intervene more in a general election.
The Prime Minister for instance, is far more involved in general elections than he is in state elections. Who the candidates are also matters and complicates this spill-over logic. Does this mean that the Congress is unlikely to win in the three states it won last December?
No. Doesn’t having won those states help boost the Congress campaign and prospects in those states and beyond? It certainly does. In fact, the Congress could very well win those three states again but given the configurations of those preceding victories and on account of past experience, it is not obvious that the spill-over effect will be strong. All depends from what parties do now and on their ability to mobilize voters during the last stretch of the campaign.