A fallacy informs much of the analysis of the performance of the Samajwadi Party. Everyone assumes that since the SP had won the last election and has lost this one, it is voters’ rage against the SP’s misrule, a vote against ‘goonda raj’, a comment on the company Mulayam Singh keeps etc. It is treated as yet another instance of anti-incumbency, a word that apparently needs no further explanation.

The trouble, of course, is that the SP did not ‘win’ the last election, nor did it lose popular support in this election. In the last election the SP won 143 seats on the basis of 25.4 per cent votes. This time its seats tally dropped to merely 97, while its vote share remained about the same as last time. In fact, it rose by a fraction to reach 25.5 per cent. This is not the routine anti-incumbency verdict we are so used to explaining away. There is a puzzle here.

In purely psephological terms, the puzzle — loss of seats without loss of votes — can be resolved very simply. In the first-past-the-post system, the number of seats a party wins does not depend only on its own votes; it depends on how others have performed. The system rewards the one who finishes first and punishes everyone else. Last time the SP’s vote share appeared big compared to others, for it was placed 2.3 percentage points higher than its nearest rival. This time the same vote share appears small when compared to the BSP, which is placed five percentage points above it. The SP has not fallen behind; it has been overtaken.

... contd.