This is a page of the general elections of 2009 in India. I had some ideas lurking in my mind for a while and I finally have got down to penning it down. The idea is really simple. As everyone knows, in India, we follow a standard plurality voting rule i.e. the winner is the candidate with highest vote share.

Being from Pune, I am a keen follower of politics in Maharashtra and like most others take a lot of interest in national politics as well. In 2009 (if I remember correctly), there were also assembly elections in Maharashtra in the latter part of 2009. As expected, Cong-NCP combination did well through the state and came back to power. However, there was an interesting aspect to that election. Once again, I don’t have the exact stats here, but essentially here’s what happened. Raj Thackeray’s MNS was solely responsible for BJP-SS’ loss of about 30 seats. Take MNS out of equation and BJP-SS were cruising back to power.

Of course, those who follow politics closely won’t be surprised one bit by this. It was a strategic masterpiece by Cong-NCP and in the end, to the outsider; it felt as if people gave a thumping victory to the incumbents. Alas! That really was not the mandate. People certainly didn’t seem to want it that way. Verdict of the people was clearly in favour of the BJP-SS-MNS, had such a coalition existed before. The reason why I am stressing on the point of the mandate not being in favour of Cong-NCP is because, as we all know, MNS is, in no ways, different from SS. Right from its hooliganism to its appeal to emotion and a charismatic leader, it is exactly the same. So, it is perfectly reasonable to assume that had MNS not been in the picture that vote would have gone to SS.

That incident and the Wikipedia  page I opened the post with has gotten me interested into doing an exercise.

What’s the underlying  feature of the incident I talked above? It’s that the plurality rule is not respecting the proportional representation that the system is designed to yield. And this is happening mainly because of the multi-party system we have. As Surjit Bhalla writes in this fantastic piece, Congress’ vote share was roughly the same in 1998 and in 2009 but in 98 it won 114 seats while in 2009 it won 206 seats! Clearly, something is amiss with either of the outcomes. The missed part is that the small parties playing spoilsport!

Okay, most probably whatever I have said so far is not new to anyone. But what’s the solution then? You can’t just get rid of regional parties. But I do feel that there’s a way in which we can limit them. That’s by changing the electoral system! Since this is a blog, we can do some simple exercises at least to see if we are on sound footing here.

The alternative I have in mind is the two-candidate runoff system that’s followed mainly in France. Here’s how it works.

In the first round standard majority election is held.

If a candidate secures more than 40-45% votes then he’s declared a winner.

If not, then 2 candidates with highest vote share are selected.

In the second round, AN ELECTION IS HELD (I know it’s a big deal from practical perspective), between those 2 candidates and the winner is determined as the person with a larger vote share.

This system clearly has its own problems from practical perspective. But observe what a runoff does. There are only two things that can happen-

Suppose candidate A is the winner of the plurality (Indian style) voting system.

Then A will always enter the second round of runoff because he’s the guy with highest vote share. So, he could possibly win the election in the second round or lose it. If he wins the runoff hasn’t altered anything. But if he loses, that means that the majority prefers some other candidate, say B, over A. So, clearly, the majority is better off. And that’s the intention of our voting system! To elect a candidate that the majority prefers.

I am absolutely not suggesting that the majority preference is the only golden rule we should adhere to in all cases, all the times. But given that our electoral system is founded on the principle of this very sensible principle, I think we should strive to achieve at least that.

It is evident to me that in case of Maharashtra 2009 elections, a runoff would change the outcome in favour of BJP-SS. At a broader level, I would be interested in the following exercise –

Place parties ideologically on a line.

Then in each constituency, arrange them according to their vote share.

Eliminate all but top 2 parties and assign the vote share of those parties to one of the closes party on ideological basis. For example, if you remove MNS then its votes would go to BJP-SS. If you remove the communist party somewhere then its vote share would go to Cong.

In case you eliminate a party in the middle. That is, say you think that CPI (M) is more left than TMC who’s more left than Cong and then you eliminate TMC. Then, split its votes to Cong and CPI (M) in some fashion! (If I knew how to split this I would have done it already but I don’t have a robust method yet. I would go with an even split to begin with.)

And then determine the winner with these  added vote shares.

For example, suppose we decided that on the ideological spectrum the parties are located as follows –

Congress < BJP < SS < MNS. That is MNS is the extreme right. (For the sake of example only)

Now, in some constituency the vote share is as follows –
34% 31% 20% 13%
then, in the second round of runoff, we’ll have
Cong vs. BJP and vote share of 34% vs. 64%. As we can see, the outcome of the runoff system has changed.

The art really is in positioning the parties correctly and then is the labour involved in actually doing the exercise. I have all the data with me. A little help would be awesome in doing this!

And why do I think that it would work? A sceptic would just come and say that this is only an outcome of voters being myopic. If they’re forward looking then this problem would be solved. I don’t doubt this one bit. But then, we know what the majority wants and we also have considerable reasons to believe that the majority isn’t strategic or forward looking. Then, isn’t it better that we don’t put the onus on them to achieve an outcome that they prefer?

I do feel that we will find a remarkably high number of occurrences where the runoff changes the outcome. And should that happen, that would be evidence that our plurality is doing the job that’s not quite in the interest of the plurality. In this country, at the moment, there are a million things wrong. Maybe we could try fixing some to them? Hope isn’t a bad thing to have after all.

The following two tabs change content below.

Aditya Kuvalekar

A miserable engineer turned into a PhD student in economics. He enjoys decision theory, matching, game theory. Other interests include maths and cricket.

Latest posts by Aditya Kuvalekar (see all)