Why BJP loss would be only surprise of UP poll

By Praveen Chakravarty

Electorally, 2017 is billed as the year of the Uttar Pradesh assembly elections, with commentators arguing that this is a crucial election for the Bharatiya Janata Party (BJP) in Indias most populous state. True to its billing, as elections were announced on January 4, 2017, the theatrics continue in the Samajwadi Party (SP), which rules Uttar Pradesh, over a father-son duel triggering a vertical split in the party.

The overwhelming narrative is that the Uttar Pradesh elections will be a harbinger for the general elections of 2019. The drama aside, the only result of the elections that should be surprising is a BJP loss.

Let's understand why.

In the 2014 Lok Sabha elections, the BJP won 71 parliament seats by winning 328 (81 per cent) of 403 assembly segments. This was unprecedented in the recent history of Uttar Pradesh elections. To put that in context, the last time a political party won more than 80 per cent of all constituencies in the state was in 1977, when the Janata Party won 80 per cent of the seats in a post-Emergency landslide

Not only did the BJP win 81 per cent of seats in 2014, but it did so with a massive margin. There are four major political parties in the fray in Uttar Pradesh -- the ruling SP, the Bahujan Samaj Party (BSP), the Congress and the BJP. These parties have historically contested elections in the state independently.

In a four-cornered, first-past-the-post system, the winning party needs only a 25-30 per cent vote share in each constituency. The BJP won 253 out of the 403 constituencies with greater than 40% vote share in the 2014 elections. This suggests that even if there had been opposition unity, the BJP may have still won well more than half of all seats.

Further, it won an outright majority (more than 50 per cent of votes) in 94 constituencies. There has been no political party that has won so many seats in Uttar Pradesh with such huge margins in recent history.

It was this victory in Uttar Pradesh that enabled the BJP to be the first to win an outright majority in Parliament since 1984. The enormity of the BJP victory in the state just three years ago should ideally mean only one very predictable outcome of the 2017 state elections -- a BJP majority.

Yet, there is palpable excitement over other possible outcomes.

For the BJP to get less than 200 seats and lose Uttar Pradesh in 2017, it would have to lose nearly 40 per cent of its seats from 2014. For that to happen, it would take nearly 15 per cent of BJP voters in the 2014 election to switch loyalties en-masse to another political party.

To put that in context, the BJP lost the Bihar elections of 2015 after winning it in 2014, by having no more than 4 per cent of its voters switch loyalties. The BJP would need a performance nearly four times worse than Bihar to lose Uttar Pradesh, given its magnitude of victory in 2014.

If any two of the opposition parties get together, the BJP would still have to lose more than 10 per cent of its voters to lose the election. Whichever way one looks at the data, it would take a massive shift of loyalties of 2014 BJP voters for the party to lose the 2017 election.

The standard counter to this analysis will be that the 2017 election is a state election while the 2014 election was a national election and hence voting patterns will be very different. That voters choose differently for state and national elections is one of the long-held myths about Indian elections for which there is little evidence.

The typical evidence dished out in favour of this argument is comparison of national and state elections across different time periods when many other changes have occurred to be able to attribute outcomes solely to voters choosing differently for the state and national elections.

The 2015 loss of the BJP in Delhi to the Aam Aadmi Party (AAP) coming on the heels of a BJP sweep of Delhi in the 2014 elections is another popularly cited example. This again misses the fact that it was primarily the collapse of the Congress vote share that helped propel AAP to victory as opposed to any conclusive evidence of voters choosing BJP in the national elections and AAP in the state elections. To the contrary, my research has shown that when elections are held simultaneously to state and Centre, 77 per cent of the voters choose the same party and do not differentiate between a state and a national election.

Further, in five of the last six state and national elections in Uttar Pradesh since 2002, the two regional parties have won a greater majority of the votes combined than the two national parties -- BJP and Congress -- except in 2014. The SP won the 2002 state election and the subsequent 2004 national election. The BSP won the 2007 state election and was the leading party in the 2009 national election.

Despite all this, even if one were to accept that voters may vote slightly differently for state elections vis-a-vis national, this difference alone cannot explain a BJP loss in Uttar Pradesh, if it were to occur.

The BJP's performance in the state in 2014 is so far ahead of the other parties that an outright majority in the 2017 elections should have ideally been par for the course. This is also the reason why the speculation about the Prime Minister's demonetisation initiative being motivated by the Uttar Pradesh elections is, perhaps, misplaced.

Should the BJP secure less than 200 seats in the coming election, it would only mean that a large number of people changed the decision they made in 2014 and would also, perhaps, go down as among the most significant vote-share swings in India's electoral history. Only under this scenario can the 2017 electoral outcome be a true harbinger for the 2019 national elections, not otherwise.

(In arrangement with IndiaSpend.org, a data-driven, non-profit, public interest journalism platform. Praveen Chakravarty is Senior Fellow in Political Economy at IDFC Institute & Founding Trustee, IndiaSpend. The views expressed are those of IndiaSpend. Feedback at respond@indiaspend.org)