Two French election traditions are likely to be broken in Sunday's presidential runoff election
A few weeks ago, I wrote on the silliness of historical "rules" in US elections. While history provides us a good guide, "it is also tradition that times must and always do change, my friend". These wise words, from Prince Akeem in Coming to America, cross oceans.
No incumbent French president has lost re-election since Valéry Giscard d'Estaing in 1981. No Socialist French presidential nominee has won election since François Mitterrand in 1988. Both of those "traditions" are likely to be history come Sunday.
President Nicolas Sarkozy of the centre-right Union for a Popular Movement party will almost certainly be defeated in Sunday's French Presidential runoff by centre-left Socialist challenger François Hollande.
The reason I feel confident in my prediction is because of the polling data. As I did prior to the first round, I've put in place a Loess regression. This type of regression allows us to track changes in voter opinion over time and provides a crisp visualization of the data that does "jump" to reflect outliers.
This graph charts changes in support for the two candidates in an initially hypothetical and now real runoff since Hollande wrapped up the Socialist nomination in October 2011.
What we see is that Hollande started out with a huge lead and has held onto the advantage in every single poll taken since October. Sarkozy has closed the gap significantly. He gained incrementally from October to just before the first run round, saw a slight leveling off, and and then has risen by about 2-2.5% to around 47-47.5% since the first round. The problem for Sarkozy is that still leaves about 5 percentage points behind Hollande.
Sarkozy's gains are reminiscent to his advances in the 2007 election. In 2007, Sarkozy increased his support in runoff polls from about 52% just before the first round to 54% in polls taken just before the runoff. He ended up right in the middle at 53.1%. I wouldn't be entirely shocked if Sarkozy's 2012 second round support actually ends up in-between where he is polling now and where he was polling just before the first round.
Some might want to point to the first round polling overperformance of far-right candidate Marine Le Pen as possible precursor to second round polling errors. I would be quite cautious in extrapolating errors.
Le Pen's polling error wasn't as great as some have claimed. She ended up at 17.9%, not the 20% that some exit polls claimed she would. That's only 2% better than an average of the final polls pegged her. This error is actually less than first round polling errors for her father, Jean Marie, in 2002 and 2007.
More to the point, polls for the second round were quite accurate in both 2002 and 2007. Polling errors, as we saw in the US presidential primary elections, tend to be greater when there are a greater number of competitive candidates. With only 2 candidates in the runoff, it would be out of character for a large polling error to occur.
The biggest issue for Sarkozy is that despite overtures to Le Pen supporters he simply hasn't picked up enough of her first round support. In order to achieve victory in the second round, I tweeted that Sarkozy needed to pick up pretty much all of Le Pen voters and half of centrist François Bayrou. Sarkozy is getting that Bayrou split, but he has only picked up a little greater than 50% of Le Pen's voters. That's simply not good enough.
It's possible that the 20% of undecided Le Pen supporters could come home to Sarkozy, but why would they now if they haven't yet? Le Pen has refused to endorse him, and Bayrou has actually backed Hollande in the second round.
Sarkozy needs an election miracle, which likely won't be granted. Because of the weak French economy, the French people believe, like Prince Akeem, leadership "must change".
François Hollande is a heavy favorite to win on Sunday and be the next French president.