Lok Sabha polls 2019: 3 key imports so far, and the median voter theorem

Now that the general elections are here and the second round of voting is underway, three things are becoming clear. One, that this is an issueless election. Two, that there is no Modi wave. And three that voters probably want Modi back as prime minister but not the BJP in full strength so that some of the excesses of the current government can be mitigated.

 

In such a situation, it is hard to guess what the likely outcomes might be.

 

That said, there is a way of doing some rational analysis. This was first devised by Harold Hotelling, an American economist, in 1929. Although he was writing about the nature of economic competition he observed that political manifestos of different parties tend to converge. Contrary to popular belief, this happens to be the BJP and the Congress manifestoes also.

 

Many other economists expanded on Hotelling’s theme and eventually the Median Voter Theorem emerged. What it says is of crucial importance in this general election, namely, that it is not the polarised voter but the non-polarised voter who will determine the outcome.

 

This, in turn, means that despite their best efforts, neither the BJP with its nationalism and religion platform, nor the Congress with its income support and secularism are going to cut much ice. This is perhaps the main reason why this election is turning out to be so hard to predict in terms of the range of seats that the two parties will win and why there is a strong chance of a third set of parties winning perhaps half the seats.

 

The Theorem

 

The median voter theorem is based on the idea that when manifestos look very similar the voter does not really care who he votes for as long as what he wants is on the menu. This in turn means that there is a very large number of indifferent voters who are the median voters. The term median is used in the statistical sense of equidistance.

 

A key assumption of the theorem is that voters decidedly prefer one alternative to the other, such as say Modi to anyone else as prime minister. This is described in the theorem as the voter’s preference being ‘single peaked’.

 

Another assumption is that voting is on a unidimensional plane, that is, non-policy factors (say caste in India) do not matter and that only the policy choices placed before the voter matter. This tends to weaken the applicability of the theorem.

 

There are five other assumptions of varying realism. But none is so completely unrealistic that it knocks out the basic conclusion of the theorem that it is the median voter who matters the most.

 

This is because the theorem predicts that all political parties will focus on short-term promises to sway the median voter. Bribes to voters are an extreme example of this but there are many manifestations of this which we can see in India.

 

How many seats the BJP will win has to be seen in the context of this theorem. The failure of the BJP to target the median voter — nationalism instead of the Ram temple at Ayodhya — could turn out to be its biggest mistake.

 

We will know in about 45 days just how good the theorem is.