NO DOUBTS

Mathematically speaking, the BJP in 2017 is a worthier winner than the SP in 2012

Big win.
Big win.
Image: Reuters/Pawan Kumar

The Bharatiya Janata Party’s (BJP) victory in the 2017 Uttar Pradesh (UP) assembly elections has attracted two types of responses. For the party’s followers, this was like rainfall on parched earth. The narrative is that UP has voted for development and everyone likes the BJP now. Winning around 80% of the seats means 80% of UP’s electorate wanted the party in power.

The BJP’s rivals are, however, using the results to prompt a rethink on the first-past-the-post system, where the winning candidate is simply the person who wins the most votes in a particular constituency. It might mean that a candidate with two votes can win if the rest of the candidates have one each. Although the BJP won only 39.7% of the votes in the state, it has 77.4% of the seats. Stranger even, the Bahujan Samaj Party (BSP) has only managed 4.7% of the seats with a vote share of 23%.

Viewed differently, approximately 18% of those who voted for the BSP are not represented in the assembly at the aggregate level, while the BJP is representing approximately 38% of those who did not vote for it. The concentration of power with the BJP, therefore, doesn’t reflect people’s true preferences at the aggregate level.

In this piece, though, we dig deeper into the numbers to gauge the quality of the BJP’s victory—whether it’s a lucky winner or a strong winner.

A fairly basic and reliable indicator of the quality of an electoral win is the margin. As a rule, if the party wins a majority of the constituencies, it wins the elections. However, if the margin of win is thin in these places, it means the party just got lucky and could be ousted the next time with a little swing. As a corollary, if the runner-up has won its constituencies (however many) by a huge margin, then we assume it has a stronghold in those parts.

The other variable that indicates the strength of the victory is the vote share. If the value is large, then the party’s victory is construed to be of a high quality.

Besides, the BJP’s victory should not be analysed in isolation. The 2012 elections pose a perfect benchmark, given that the Samajwadi Party (SP) also had a huge victory. Full data is not available yet but with whatever little is available, these two values—margin and vote share—can be estimated at the constituency level.

The SP was the winner in the 2012 assembly elections, securing 224 out of the 403 seats, i.e. 55.5% of all constituencies. In 2017, the BJP has secured 312 seats (for neat results, the coalition is not being factored in), which is 77.4%.

The result of the evaluation is remarkable.

In 2012, the SP had an average margin of 9% in the 224 constituencies it won. Its average vote share was 36.25%. This means that wherever it won, it defeated the runner up by an average of 9% of the total votes. In these seats, on an average, it got 36.25% of the total votes. For the whole state, the SP’s vote share stood at 29.9%.

Contrast this with the BJP’s performance in 2017. The party’s average margin in the 312 constituencies it won is 15.04% and the average vote share 44.1% (39.7% for the whole state).

The BJP had a significantly higher margin and vote share in 2017 than the SP in 2012. Thus, the BJP in 2017 is unequivocally a worthier winner than the SP in 2012.

However, these average values could be deceptive if the distributions are highly skewed, i.e. the presence of extreme values on the higher side can lead to a higher overall average. Thus, if the winner gets a 40% margin in many of the constituencies, the average might go up. So, if the margins are spread enormously in the winning constituencies, it doesn’t tell us much. It could well be the case that in many places, the BJP’s margins in 2017 were low, but in some very high, pulling the average a little higher than the SP’s in 2012.

Mathematically speaking, only if the margins and vote share in the winning constituencies are normally distributed will the BJP’s high values be comparable to the SP’s.

The evaluation of the results found the median values close to normal distribution. The median value shows the point until where half the elements are covered. The median winning margin in the SP’s 224 seats of 2012 is 7.96% and median vote share 35.42%. Remember, the average values are very close these figures—9% margin and 36.25% vote share. For the BJP in 2017, the 312 winning seats yielded a median margin of 13.97% (average of 15.04%) and median vote share of 43.64% (average 44.1%).

This means the average values are indeed the real representation of the election results, which show the superiority of the BJP’s victory.

The clarity here is surprisingly telling. The BJP’s victory is not a regret case of first-past-the-post. Its mandate carries preferences, more strongly than that carried by the SP’s in 2012. Wherever the BJP had won, its victory was powerful, certain, and sizeable.

If one observes these values for other parties, it’s a similar story. While in 2012 almost all the parties had a similar average vote share in their respective winning constituencies (34.2% plus/minus 2%), the range (the highest vote share minus the lowest vote share) in 2017 is high (the BJP’s 44.1% to the BSP’s 36.03%). Similar is the pattern of average margin of victory in whichever seat the respective party has won.

Looking closely, we also see the poorest quality of victory has been that of the BSP, both in 2012 and 2017. It has the lowest average margin of victory and lowest vote share in the seats it won. When analysts lament the BSP’s low seat share and high vote share, they must also factor in how weak the foundations of the BSP’s victories have been anyway.

The authors teach Economics at OP Jindal Global University. We welcome your comments at ideas.india@qz.com