- NL Sena
Percentages are a tool and one should know when to use them.
On May 23, 2019, a few hours after the counting had started, it was more or less clear that Narendra Modi and the Bharatiya Janata Party (BJP), were going to win the elections to the 17th Lok Sabha.
I was part of a team on a TV channel analysing the results as they trickled in. The stock market greeted the news of Modi winning with great enthusiasm and the Bombay Stock Exchange (BSE) Sensex—India’s most popular stock market index—was up 800-900 points, from the previous day’s close. As soon as this happened, we discussed the tendency of the media to report stock market rises in absolute terms and not in percentages.
A day before, on May 22, the BSE Sensex had closed at 39,110.21 points. An 800-900 point rally over this meant a rise of around 2-2.3 per cent. A 2-2.3 per cent rise in the BSE Sensex over a short period of time is pretty good. But it is nowhere as sexy as telling the viewer or the reader that the Sensex was up by 800-900 points.
Later in the day, we were talking about the dismal performance of the Congress party. In a lighter vein, I suggested, we should say that the party has improved its performance by 13-14 per cent, given that it was leading on 50 seats at that point of time, against the 44 seats it had won last time. This would make the performance look much better than it actually was.
Still later, and this was off air from what I remember, we talked about the dismal performance of the Congress in Uttar Pradesh, where the number of seats the party won had come down to 1. But it had won two seats in 2014. Hence, this implied a fall of 50 per cent in the number of seats the party won in Uttar Pradesh.
To cut a long story short, the lesson here is that percentages are a tool and one should know when to use them. Many times, this important distinction isn’t made. Let’s look at a few more such things from the results of the 17th Lok Sabha.
1) An estimate made by The Hindu newspaper suggests that the BJP won around 37.4 per cent of the votes polled in the 2019 Lok Sabha elections. At the same time, it won around 303 seats of the 542 seats which went to poll. This works out to around 56 per cent. Basically, a 37.4 per cent vote share led to the party winning 56 per cent of the seats contested. In fact, in 2014, the BJP had won 31 per cent of the votes polled and 52 per cent of the seats.
On the flip side, at 19.5 per cent of the votes polled, the Congress won around half the votes of what the BJP won. It won 52 out of the 542 seats which went to poll. This works out to 9.5 per cent of the seats in the Lok Sabha. Basically, the vote share of the Congress was around half of that of the BJP, but the number of seats that it won came to around 17 per cent of the total seats won by the BJP (52 expressed as a percentage of 303). This is how the first past the post system tends to work, with the party winning the election getting more bang for the buck.
2) Taking the first point forward, there are some WhatsApp forwards going around and this is what they have to say. The voter turnout was 67.11 per cent in the Lok Sabha elections. This basically meant that on an average for every 10,000 people who could vote, only 6,711 did. The remaining didn’t.
The BJP won 37.4 per cent of the votes that were cast. Let’s say the size of the electorate was 10,000 in total (ie 10,000 people could cast their vote, though eventually, a lower number of 6,711 voters did). The BJP won 37.4 per cent of the 6,711 votes that were cast. This means that it won 2,510 votes (37.4 per cent of the 6,711 votes) in total.
Hence, the BJP won 2,510 votes from the total electorate. This means that only 25 per cent of those who could (2,510 votes expressed as a percentage of 10,000), cast their votes in favour of the BJP. “This means that 75 per cent of the people are still against Modi.”
The line in italics is the WhatsApp forward going around. Let’s try and understand why this is wrong. There are multiple reasons for it. First and foremost, while the vote share of the BJP is 37.4 per cent, the vote share of the National Democratic Alliance is at 45 per cent. Second, just because people did not vote, does not mean that they are against the BJP and Modi.
The third point is the most important of the lot. Using the same logic, the Congress won 19.5 per cent of the votes polled. This means that the Congress won around 1,309 votes of the 6,711 votes polled, from an electorate of 10,000 people. Hence, a little over 13 per cent of those who could, cast their votes in favour of the Congress. Going by the same logic, this means that 87 per cent of the people are against Rahul Gandhi. Hence, Modi is more popular.
The broader point here is that Indian democracy runs on a first past the post system, where multiple candidates fight elections and the one winning the most votes wins. And Modi and the BJP did just that. In this scenario, very rarely do governments win more than 50 per cent of the votes on the whole. Even in the 1984 Lok Sabha election, when there was a huge sympathy vote for the Congress after Indira Gandhi was assassinated, the party got a less than 50 per cent of the votes. That’s how the arithmetic works out.
Hence, this does not mean that a major part of the population is against the leader who is elected, primarily because we don’t know how those who did not vote, would have voted, if they had voted.
3) Another interesting point that is being made is that this Lok Sabha has the highest number of women MPs ever—78 women have been elected to the Lok Sabha. While this sounds quite high on its own, it forms only 14.4 per cent of the strength of the Lok Sabha. This is extremely low when it comes to representation of women in different Parliaments around the world. As PRS Legislative Research points out: “Though the percentage of women MPs has increased over the years, it is still lower in comparison to some countries. These include Rwanda (61 per cent), South Africa (43 per cent), UK (32 per cent), USA (24 per cent), Bangladesh (21 per cent).”
At the same time, Prannoy Roy and Dorab Sopariwala point out in their book The Verdict that women over 18 years and above are around 97.2 per cent of the male population. Given this, basic fact, we need many more women in the Lok Sabha, though a 14.4 per cent representation is the highest ever and an improvement over the past.
4) Another parameter using which we can check the participation of women in the democratic process, is by looking at what proportion of the women’s electorate actually voted. While the proportion isn’t available as of now, Deputy Election Commissioner Umesh Sinha did tell the Times of India: “The gap between turnout of male and female voters has narrowed down significantly from 9 per cent in 2009 to only 0.4 per cent in this election.”
On top of this one needs to keep in mind the fact that men form a greater part of the electorate. Of the total electorate of around 89.6 crore, the number of men was at 46.47 crore, and the number of women was at 43.13 crore. The third gender came in at 39,683 individuals.
The point is that even though more and more women are voting, the number of men who vote continues to be more. Also, women’s turnout has increased post-2009. From 1977 to 2009, the turnout of the women voters remained higher than fifty per cent but lower than 60 per cent. The interesting thing is that it reached a then peak of 58.6 per cent in 1984 and was only at 55.8 per cent in 2009. In 2014, the turnout among women voters was 65.5 per cent. Of the nearly 39.70 crore women who could have voted, around 26.02 crore did. A rough back of the envelope calculation suggests that this time around, the turnout would be slightly better than 65.5 per cent, recorded previously.
Another point that needs to be made here is that in states like Bihar where the proportionate turnout of women is higher than that of men, may not always be a good thing. Many men leave Bihar in search of better work opportunities and it is not always possible for them to go back and cast their vote during the election time.
To conclude, in case of any data, election or otherwise, it is important to see the data from a holistic perspective and not just by using percentages or otherwise.