Second round forecast

There are no last minute polls today. As the official campaign ended at midnight last night, the last slew of polls was yesterday afternoon, all showing a tightening of the race. The average of the final polls of the eight major instituts de sondage have Hollande winning with 52.75%. This is the narrowest overall spread since the campaign began, which may or may not indicate a dynamic in Sarkozy’s favor. I am reminded of the last-minute tightening in the 1992 US race between Bush Sr and Clinton, but which the latter ended up winning handily. The problem for pollsters in France is trying to predict the transfer of votes from eliminated first round candidates—here Le Pen, Mélenchon, and Bayrou—in view of the small sample sizes and thus large margins of errors. Mélenchon voters will of course transfer massively to Hollande and with almost none to Sarkozy, though if over 20% abstain that will not be good for Hollande. In order for Sarkozy to pull of a miraculous victory he will need 80-90% of all first round MLP voters, at least 60% of Bayrou’s, over 20% abstention of JLM’s, and a clear majority of first round abstainers who decide to vote this time. A perfect storm. Sarkozy also has another thin silver lining in the cloud, which was the lower-than-poll-suggested stock of first round votes for left candidates: 43.8% and not the 45-46% that some polls were indicating. Mélenchon’s weaker than expected score deprived Hollande of a reservoir he was no doubt counting on.

So it’s not totally hopeless for Sarkozy, though I just don’t see him pulling it off. Everything in this campaign has been going Hollande’s way: he has made no mistakes or committed any gaffes, his personal approval ratings remain high—over 50%—whereas Sarkozy’s have not significantly moved from the disastrously low high 30s, at no point has any poll had Sarkozy in the second round lead or even close to it, Hollande’s party has been rock solidly behind him and he has had no real problems from other left candidates, the other major candidates (MLP, JLM, FB) want Sarkozy defeated and even Bayrou will be voting Hollande, Jacques Chirac will be voting for him and major UMP personalities (notably J-F Copé) discreetly want Sarkozy to lose, the UMP has been psychologically readying itself for Sarkozy’s defeat, Hollande clearly won the debate with Sarkozy and demonstrated his credibility to all but the most blindly partisan Sarkozy supporters, the French electorate is ready for a change, alternance is in the air… In short, there is no reason whatever for Hollande to unexpectedly lose this one. I thus predict

Hollande: 52.5%
Sarkozy: 47.5%
participation rate: 83%

It would be nice if Hollande received a higher score but given the reality of the French electorate—the majority of which is on the right—I would be surprised. A certain number of Sarkozy voters tomorrow will be disappointed conservatives who give him the thumbs down but nonetheless prefer him to the left in power. But it won’t be enough to push him over 50%.

FWIW, my 2007 second round prediction was

Sarkozy: 51.5%
Royal: 48.5%
participation rate: 82%

I was off by 1.6%. Even if I’m off by that much tomorrow, it won’t alter the result. On verra bien.

UPDATE: Sunday morning: A report late last night on RTBF: “Selon un institut de sondages bien connu sur la place de Paris, François Hollande l’emporterait toujours par 53% contre 47 % au président sortant Nicolas Sarkozy.”