A voting system that could change UP's face

By Siddharth Varadarajan
Published March 11th 2004 in The India Times
NEW DELHI : The Election Commission may not have taken a decision on providing a ‘none of the above’ option to voters but it has highlighted a key anomaly with the Indian (or rather Westminster ) system of voting:

In multi-cornered first-past-the-post contests, it is quite normal for candidates to be elected on vote shares that are much less than 50% and sometimes as low as 26%.

In the 1999 general elections, for example, 42 out of Uttar Pradesh’s 85 seats were won on vote shares of less than 35%. Only four MPs were elected with a vote share greater than 50%. Candidate Atal Bihari Vajpayee did better than most, but he too clocked in at 48%. That means 52% of those who voted in Lucknow that year didn’t even want him as MP, let alone PM. 

In 1859, the English lawyer and publicist Thomas Hare conceived of a way around this problem of ‘under-representation’. In A Treatise on the Election of Representatives, he proposed single transferable votes  (STVs) in which a voter ranked a given panel of candidates in order of preference. If a candidate got more than 50% first preferences, he would be  elected. But if no one polled more than 50% votes of the candidate who polled least would be transferred to the remaining candidates in line with preferences indicated by voters on their ballot papers. 

This would continue until at least one candidate crossed the 50% mark. 

Even though John Stuart Mill strongly backed Hare’s idea, the English establishment was distrustful of STV. Today, only Malta and Ireland run national elections on transferable votes. 

The advantages of STV are obvious: it allows voters to back candidates or parties they support — even if they are considered ‘marginal’— without fear of wasting their vote. In India , where many vote on the basis of caste, STV forces them to think beyond the first iteration of identity. STV also allows different political parties to take on the ruling party singly, without the danger of the latter winning because of opposition disunity. 

In UP, for example, there would be no need for pre-poll alliances between the Congress and BSP or SP. Instead, voters would be free to mark Congress as their first preference and BSP, SP or even BJP as their second, or third, or fourth preferences. In 1999, Murli Manohar Joshi won Allahabad with 33.8% votes. Reoti Raman Singh of the SP polled 22.9% followed by Reeta Bahuguna of the Congress with 20.5% and the BSP with 16.8%. With STV, BSP would be eliminated from the race and its votes reallocated to the other three based on second preferences. Assuming the bulk of BSP voters put down Congress as their second choice, the Congress score would go up to around 35% eliminating SP and spreading its votes between BJP and Congress. 

Given what we know of the support base of the SP, the Congress would pick up the bulk of second preferences and Murli Manohar Joshi would likely have lost. Other prominent BJP leaders who would probably be squeezed out by STV would be Vinay Katiyar, who won Faizabad with 29.4% and Swami Chinmayanand, who won Jaunpur with 30%. 

However, all parties with marginal wins would stand to lose. The SP won Machhilishahr in 1999 with 30.8%, followed by BJP with 27.6%, BSP with 21%, Apna Dal with 10% and Congress with 7.7%. As losing candidates get eliminated, it is possible that the BJP might also pick up second preference votes and forge ahead of SP. In any case, the outcomes would be very different — and far more representative of actual voter choices than they currently are.