Snakes and ladders of the turn-out game

The French election will be decided by those who do not vote. This is true, if you like, of all elections; but it is especially true in the French system, a two-round hybrid of proportionalism and first-past-the- postism.

The rules for the election tomorrow and tomorrow week, and the recalcitrant mood of the electorate, make the level of turn-out a crucial and unpredictable factor. Turn-out in French parliamentary elections is always lower than presidential elections and has been on a falling curve for years. Last time, in 1993, it was just below 70 per cent. This time, it is forecast that it could fall as low as 65 per cent.

Every percentage point higher or lower scrambles the already mind-bending arithmetic of the election. Why?

There are 6,360 candidates (a record) running in 577 constituencies, including those in the outposts in the Caribbean and the Pacific, Atlantic and Indian oceans which are forever France. If any candidate scores more than 50 per cent in the first round, the election is over in that constituency. Otherwise, the first two candidates in each electoral district go automatically into the next round. They can be joined by a third candidate (and theoretically a fourth) if he or she scores more than 12.5 per cent of the voters registered in that constituency (that is, not 12.5 per cent of the votes actually cast that day).

This is why turn-out is so crucial. If participation falls to, say, 65 per cent in any constituency, the qualifying score for a third candidate to enter the decisive round on 1 June becomes a forbidding 19 per cent. The lower the turnout, the fewer third-place candidates will qualify.

This is vital largely because the third-place candidates are often from the far-right National Front (FN). The opinion poll scores of the main left and centre-right alliances are both hovering around 40 per cent. But all polling organisations forecast that the centre-right should gain a comfortable majority in the second round.

This is because the voters for the knocked-out FN candidates tend to switch to the centre-right or not vote again at all. The more FN candidates who survive to fight another day, the fewer votes, and constituencies, will swing to the centre-right on Sunday week. At least 40, and as many as 100, constituencies will be decided in this way. In theory, the lower the turn-out, the fewer three-cornered contests in the second round, and the better the chances of centre-right parties to stay in power.

This is the theory; in reality, the variables are mind-boggling. The calculations of the polling organisations seem precise but depend on a series of uncheckable hunches and assumptions. Even if turn-out is low, the FN vote may hold up. A 16 per cent score for the FN nationwide would force 120 or 130 triangular contests. Each 1 per cent more puts around 50 extra FN candidates into the 1 June run-off. If the FN score goes as high as 17 or 18 per cent, the result of the election will depend on scores of three-way marginals which will be decided by a handful of votes each.

The election could, in other words, become a lottery.