The Bhartiya Janata Party-led National Democratic Alliance won 31 out of 40 seats in the 2014 Lok Sabha elections in Bihar – a 77% rate of win.

The victory led many to believe that it was a clean sweep. However, an analysis of the underlying figures shows that even the 2014 parliamentary elections were closer than most people realise.

When you look at assembly segment leads within the larger parliamentary constituency (below table data from Vozag), you end up with NDA at 172 seats, and the Janata combine at 64. This implies that if the BJP wins at the same rate, they would have 70% of Bihar’s 243 assembly seats.  This is a lower rate of win at the assembly level than the parliamentary level.

When you dig even deeper and look at the vote-shares, the numbers appear more closer.

While in 2014, the Congress, RJD and JD(U) fought separately, they pulled a total of 45% share of the vote, which together is 7% more than that of the NDA. However, as we know in politics, one plus one doesn’t necessarily equal two. For example, a Kurmi (the caste of Nitish Kumar) may not vote for a RJD/Lalu candidate; likewise, a Yadav voter may not vote for a JD(U) candidate.

Additionally, the BJP, with its alliance with Manjhi, may get more votes of the Mahadalit community thereby reducing JD(U)’s vote-share.  Also, the Owaisi factor may split Muslim votes in Seemanchal, which may reduce the vote share of the Janata Mahagathbandhan.  So, after accounting for these three factors, if we split the 7% vote differential and give a few points to the opposite camp, we end up with a very close vote-share percentage.

Now comes the inevitable hedging of our predictions! As we explained last year in evaluating exit poll projections, an election maybe extremely close from a vote-share perspective, but you could get completely one-sided results when it is translated to seats. Sometimes, one or two percent vote-share difference could lead to a difference of 10 to 20 seats. So, if the vote-share of two main blocks is not close, please do call us out on it; but if one party wins with a big majority despite the close vote-share, we can still claim “I told you so”.