|The results of the first round of the French presidential elections conformed entirely to expectations. There were no surprises, only minor adjustments. We would therefore fully expect François Hollande to enter the Elysée in a fortnight’s time.|
These seem like shocking claims given the coverage of the results. According to the media, and a number of independent commentators, there were surprises:
1. Nicolas Sarkozy was the first incumbent president not to win the first round of a Presidential election.
2. Marine Le Pen won a record number of votes, easily beating her father’s performance in 2002.
3. Jean-Luc Mélenchon performed much worse than expected.
4. Turnout was one of the highest in recent memory, despite concerns in the latter days of the campaign that abstention would be significant.
For surprises (1) and (4), we need to be careful of the small n trap. Only four other incumbents have ever stood for re-election. 1 in 5 is not a headline proportion. Given the parlous state of Sarkozy’s polls only weeks ago, that he managed to come second with only a shortfall of less than 1.5 per cent on Hollande is impressive, and more likely to merit the label ‘surprise’. That he is in second place as a highly unpopular and divisive incumbent is no surprise. Similarly the level of turnout is high but within a small number of percentage points of the majority of presidential races (2002 excluded), and is actually almost exactly the average turnout for all Fifth Republic first-round ballots. An inkling of low turnout only manifested itself in the latter days of the campaign, and has proved to be false. As with many ‘truths’ manifested in polling, the unsurprising counter-factual has proved to be the case.
Surprise (2) is perhaps the most contentious. Our reservations would first come from the fact that our own long-range model predicted 17.4 per cent. Having been announced at 20 per cent at the beginning of the evening – a calamitous failure of polling estimates which we deal with elsewhere – her score fell to 17.9 per cent. A 2.1 per cent over-estimate perhaps seems forgivable, pollsters will no doubt say. However, for a candidature as controversial as Marine Le Pen’s, to place it above the psychological bar of 20 per cent is a watershed moment. A number of other polling institutes provided much more accurate estimates for Le Pen as the polling booths closed. CSA, for instance estimated 18.2%, Harris 18.5%. It is embarrassing that the company chosen by French state media to announce the outcome turned out to have the most wayward forecast.
That a surge in support for the FN candidate has manifested in this election is undeniable, and comes in striking contrast with the outcome of the 2007 election where the party was left almost moribund. The raw numbers are also many digits superior to Le Pen 2002 – 6.4 million voters chose Le Pen fille compared with 4.8 million ten years ago – but so were turnout and the actual pool of registered voters. Whilst there is no getting away from her having performed remarkably well, it would have been absurd to expect that the ‘additional’ votes above and beyond 2002 would not have boosted the raw number. Looking at scores based on registered voters helps nuance the current assessment of a ‘massive wave’ of enthusiasm behind the new FN passionaria: she polled 13.9 per cent, compared with an average of about 11.5 per cent for her father across the 1988 (11.5), 1995 (11.4) and 2002 (11.7) elections. The actual percentage difference was of the order of 2 per cent - very similar to the combined Extreme Right score in 2002 of 13.3 per cent of the registered voters. Beside the personality effect, some of Marine Le Pen’s bonus could well be due to her capacity to bring in former MNR voters from the cold.
Finally, Jean-Luc Mélenchon, largely a product of media hype. His final score of 11.1 per cent was indeed disappointing for his many supporters, but hardly a surprise. In an election manifesting a bolstering of the extremes in polls and broader coverage, and resembling the 2002 race in its competitive dynamics, a performance slightly below that of Laguiller, Besancenot, Hue and Gluckstein appeared a reasonable estimate: together with the small Trotskyite candidates, the radical left has polled 12.8 per cent of the votes this year compared with 13.8 per cent in both the 1995 and 2002 elections. The surprise would have been the Front de gauche candidate realising anything similar to Le Pen’s performance. That this seemed reasonable to expect derived both from a media fascination with a compelling tribun riding a wave of populist support – but not popular support; and by far the most professional and adept social media campaign of any of the candidates. In the end, the style outweighed the substance.
What of François Hollande and Nicolas Sarkozy? After a frenetic but relatively short campaign, Nicolas Sarkozy pulled back much of the space between him and the Socialist, but not quite sufficient to win. Hollande, for his part, capitalised on an excessively long campaign, and held firm to win the first round with a stable, sober programme and candidacy. He enters the second round as clear favourite, but with a challenging fortnight ahead of him. It would be difficult to unearth any manner of surprise here.
The second round now lines up in predictable fashion. Neither Le Pen nor Bayrou have given any indication of a preference between the two candidates. Bayrou will give one in due course. Given her strong –and to date consistent anti-system– position, Le Pen seems much more unlikely to recommend either Hollande or Sarkozy when she delivers the traditional FN speech on 1st May. With the legislatives on the horizon, the temptation for Bayrou to try to ensure some safe parliamentary seats may be compelling after no less than a decade of political isolation. The FN, which has enjoyed a seat desert for many years, seems to be betting instead on the implosion of the mainstream right following Sarkozy’s defeat. Locally, UMP candidates under the threat of the FN might indeed decide to enter strategic talks with the far right, in what could be a replication of the tactical alliances that took place in the 1998 regionals.
On the left, Mélenchon has not called for his voters to move to Hollande, but he has called on them to defeat Sarkozy. In a two-candidate race, options are therefore limited. Joly, for her 2.3 per cent, has indicated similarly. Poutou and Arthaud voters are also likely to move en masse (if their combined electorates of 1.71 per cent constitute une masse) to Hollande.
Two key variables remain, then. Firstly, does Sarkozy court Le Pen or Bayrou voters? Le Pen has twice as many to court, but they are less predictable. Bayrou’s voters would require concessions to the centre by the incumbent, but if so instructed, would be likely to follow their leader’s instructions. It will be impossible to court both: social liberals and authoritarian Right are not electorates with anything in common, even a dislike of a Leftist candidate like Hollande. In one sense, however, the choice is moot. Whichever way Sarkozy turns, the maths simply do not add up. The only possibility would be for Sarkozy to win back some additional support in the first few days of second-round campaigning, trounce Hollande in the debate(s) still to come, and then rely upon essentially stochastic elements to secure a hairline victory. That is not plannable, only serendipity (at least, for the candidate concerned). In the end, Hollande must win.
Perhaps the biggest shock, however, came from the village of Donzy, in the centre of France. Fêted as "the village which predicts elections", having placed presidential candidates in the national order since 1981, Donzy's run of luck finally came to an end, with the village placing Sarkozy well ahead of Hollande. It could be that, in two weeks' time, as the removal vans leave empty from the Elysée forecourt, we may well realise that Donzy was right ahead of time, for once – it was the rest of France that was wrong.
But perhaps not.