March 9, 1998

A hung Lok Sabha--and a new voting scheme to prevent it

Kaushik Basu

Amajor concern for all Indians today is the repeat appearance of a hung Lok Sabha. If, after an election, no party manages to establish a majority in the Lower House of Parliament, the standard recourse has been to call a mid-term election. However, with around 600 million voters, elections are expensive. Further, there is no guarantee that the new Lok Sabha will yield a majority. In this essay, I suggest a new voting system which, without sacrificing the basic principles of democracy, will almost always guarantee that some party will come out with a majority.

A few criteria are worth bearing in mind. First, the new system must be simple; so that even the illiterate voter can understand it. Second, it must not be so different from the present system that it gets rejected out of hand as too alien to Indian democracy. Indeed, I believe India would be better served by a presidential system and we could use the system of run-offs that many countries use to ensure that the elected president has majority support. But this would involve too major a change for it to be immediately implementable. Finally, the new system must not be too expensive.

On every ballot paper, voters will be asked to mark their most preferred candidate and, if they wish, their second most preferred candidate. To start with, only the information concerning the first-preference votes will be used to find out which party has got how many seats in the Lok Sabha. If some party gets a majority of seats, that party forms the government; and there will be no difference with the present system. The difference will occur if, on the basis of the first-preference count, no party emerges with a majority. Then the following procedure will have to be followed. Check which party has got the fewest seats -- let us call this the Marginal Party. In all constituencies from where a member of the Marginal Party was elected, reopen the ballot papers and wherever it is found that the first-preference vote was given to the Marginal Party, ignore that and treat the second-preference vote of the voter -- wherever a second preference is expressed -- as his or her choice. With this correction find out who the winner from each such constituency is.

While I say that we shall have to reopen the ballot papers, that is simply a manner of speaking, since all the information (concerning first- and second-preference votes) in the ballot papers should have been collected during the initial count. Hence, this second count will take very little effort. Indeed, the information from the ballots from all over India can be entered into a computer and a programme can be written so that, should we need to recount votes from some constituencies because the first-preference votes gave us a hung House, it can be done in minutes.

If with this recount a majority party emerges, it forms the government and that is the end of the matter. If not, we look at the party which now has the fewest seats in the Lok Sabha and repeat the process described above. Barring some very special events, this system of voting will invariably result in a government which commands a majority in the Lok Sabha. It will not involve repeat elections, will be marginally more expensive than our current elections -- voters will have to exercise their forearms twice and the counting will have to take account of second preferences -- and it will give us the political stability we all seek.

My expectation is once a system like this is instituted, the very fact of its being there will make it much more likely that the first count will yield a majority. The reason is that under this system, it will be in the interest of small parties to enter into formal alliances with other parties before the general elections. They will realise that if they get too few seats in the first count, they may end up with no seats in the final count.

With the rise of regional groups, it is likely that over the next decades India will see a diverse set of parties in power in different states. What the new voting system will ensure is that these parties will, at the time of the parliamentary elections, seek out like-minded parties to form large, formal coalitions.

This is just an outline. The mechanism will have to be fine-tuned before it can be put into effect. Also, we may want to debate some variants of the above system before settling on one. For this, we will not have to work from scratch because a large amount of literature on voting theory is already available.

In the 19th century, Charles Lutwidge Dodgson, better known by his pseudonym of Lewis Carroll, had investigated the properties of alternative voting systems. Even before that, Condorcet, the French philosopher and mathematician, had uncovered a paradox of the majority voting system. The culmination of this long line of work occurred in the 1950s when economist Kenneth Arrow proved one of the most dramatic theorems in the social sciences, the "Arrow impossibility theorem".

It shows that all voting systems must suffer from fairly major defects. So the choice of a voting system is basically a trade-off in defects. Arrow's theorem assures us the new system cannot be a panacea for all evils. But with some thought and debate, we should be able to do better than we have thus far.

The author is C. Marks professor of economics, Cornell University.

Write to us | Subscriptions

© Living Media India Ltd