by afew
Mon Mar 12th, 2007 at 06:04:42 AM EST
There's a fair amount of talk, on French Internet and ET too, about how the best thing would be (again) to recognize Ségolène Royal is a dud candidate, and support François Bayrou as the last white hope of the left against Nicolas Sarkozy. I won't go too far into that aspect, except to link you to this comment by desmoulins. But another side of the question is what to make of the polls six weeks out.
 Canard Enchaîné
I added up the scores of the candidates... And I got more than 100% votes! What does that mean? Bayrou went up again!
|
Note by Jerome: updated with full list of earlier posts on the French elections at the end of the story.
There are more and more polls (Ipsos has moved into daily mode), and media commentary is increasingly based on them. The polls certainly show Bayrou rising (equal to Royal at 23%, according to IFOP), and the media certainly don't stop commenting on Bayrou rising. By attacking them head-on with accusations of bias, Bayrou not only succeeded in positioning himself as the plucky underdog fighting against the foregone conclusion of a Sarkozy/Royal run-off, he also made a lot of journalists and media bosses wary of under-handling him, to the point where they over-compensate. This is at least part of the way he became the "third candidate".
Another part is the endless media search for a good story to tell. Last year it was the fairy-tale of beautiful Ségolène coming from the outside to take on the establishment; then, in January, how her coach turned out to be a pumpkin after all as she went into free fall in the polls. She wasn't actually in free fall at all, and anyway the story stopped abruptly. I expected the media to make hay with Eric Besson's attacks, but they let the whole thing drop. At the time Sarkozy was concerned that Royal not be killed off too soon, which would leave him in a difficult face-off with Bayrou. Don't take it I'm saying the media are Sarkozy's puppet... But he does enjoy more media influence than probably any presidential candidate before him. If he now feels that Bayrou's run has gone far enough, and that Royal needs to be kept in play, it wouldn't be at all surprising to find the media cooling off on the story of the rise and rise of François Bayrou...
This guy afew's in denial, you say. His candidate is polling badly, so he accuses the media! The polls are there and what they say is clear, get used to it!
Well, apart from the fact that even journalists who should know better speak of "forecasts" when polls are no more than a "snapshot" of opinion at one moment, and that the error margin of 2% to 4% is never mentioned (would the narrative be as dramatic if those two faults were corrected?), what the polls say is only apparently clear. Polling is becoming a more difficult business. According to the Canard Enchaîné (07/03):
...d'autres réalités, qui multiplie les risques d'erreur: taille des échantillons en baisse, refus de répondre croissant des sondés, sous-représentation des classes populaires et, parfois, bidonnages des enquêteurs de base. | | ...(there are) other realities, that increase the risk of error: reduced sample size, growing refusal to reply by those polled, under-representation of the lower classes, and, sometimes, fixing by investigators. |
In fact, voters at the moment in France are still unsure of their choice. A subsidiary question by IFOP (Are you certain to vote for the candidate you chose in this poll, or might you still change your mind?) gives a 52%/48% split, in other words about one in two voters have not entirely made up their minds.
Yet the percentages given for abstention/don't know are surprising: both low and varying wildly from one pollster to another. First of all, two categories, those who declare they will abstain, and those who say they don't know or will not reply, are lumped into one (Ségolène Royal said the other day that the polling institutes should be forced to present clearly the number of don't knows). Next, they are low. Abstention in the first round in French presidentials is not high, but generally in the twenties: 28% in 2002, 22% in 1995, for example. Yet only one pollster, CSA, offers numbers that might fit with that (the abstentions, without counting the don't knows); the others are all way beneath. To see the different rates offered by the six polling institutes, see this nifty presentation from Le Monde - the abstention/don't knows are profiled in white behind the candidates' curves. Roughly speaking, these are the percentages on offer:
CSA : 20%-30% (24% 7/03) |
| IFOP : 3%-11% (4% 11/03) |
IPSOS : 10%-16% (11% 5/03) |
| TNS Sofres : 9%-23% (15% 5/03) |
BVA : 10%-20% (10% 8/03) |
| LH2 : 10%-20% (15% 5/03) |
OK, we can say this is not important: all that matters is what percentage of those who express an intention, go to this candidate or that. But what it unfortunately suggests is that respondents may be being pressured into giving a reply when they don't really want to; also, possibly, that shame prevents some abstainers from stating their intention openly. Add to that a certainty, that French polls always underestimate the real electorate of Jean-Marie Le Pen (shame, or perhaps malicious pleasure in camouflaging one's fuck-them-all vote, seems to lead people to keep their choice hidden when polled), and we can conclude that the polls already have a large proportion of respondents making up an answer. Does it matter? Isn't their choice of make-believe vote revealing? Perhaps. But what if it only revealed that Bayrou is a refuge (or even apparently fuck-you) candidate at the moment? What if it simply revealed the outline of the story the media is telling? Bayrou the third man, the middle way?
The media buzz influences the polls that feed the media buzz, and round and round we go? There's some evidence that the French are sensitive to that idea, or at least to the notion that the polls have a great deal of influence on political life, in an IFOP poll from last week about how polls are considered. The question:
Would you say that the polls published during electoral periods, for example a presidential campaign, have a very strong, fairly strong, fairly weak, or very weak influence on...?

(translation mine)
The last responses could be amusing, but there's a picture there of people who see themselves as wise to the game: the polls have the power to skew public debate and other citizens' votes, but I can see through all that and keep my own counsel.
There's also some evidence for Bayrou being a convenient refuge for poll respondents at the moment. IPSOS tracks a subsidiary question on the firmness of the respondent's intention to vote for the candidate they named.
Is your choice definite or could it change? Reply: definite:
No surprise to see the firmness of the hard-core Le Pen vote, but note that Sarkozy and Royal are beating the average (see above, 52%/48% split on mind made up) by as much as Bayrou is trailing it; three out of five naming Bayrou in the poll are not sure of their choice.
A great deal can change in the six weeks that remain to first-round polling day. Six weeks out in 2002 and in 1995, the opinion polls drew a picture of the race that was nothing like the final result. Le Pen was under-polled, and a "third man" (Chevènement, Balladur) was looming larger than he performed in the end. There are fairly solid historical constituencies in French politics, and polling bubbles that expand for a supposedly "new" one-man run tend to deflate as the real campaign gets under way. Bayrou has everything to prove before consolidating the speedy progress he's making in the polls. He and the current UDF don't command a large constituency, and, in my opinion,won't consolidate to the point where Bayrou will reach Round Two.
Why I don't discount Ségolène Royal is – without exaggerating the phenomenon – that her participative approach gives her a solider basis than the polls suggest. Her rallies and meetings seem to be genuinely chock-a-block. When she features in a major TV show, she gets record audience ratings. As for the polls themselves, she appears to be fairly stable in the mid-twenties, the "natural" constituency for a PS candidate not competed with by small fellow-parties like the MRC (Chevènement) and the MRG (Left Radicals, Christiane Taubira was their candidate in 2002). (Jospin + Chevènement + Taubira = 24% in 2002).
That's not to say there's no danger, of course there's danger. Rehearse again the media narrative since the year began: Ségolène is falling, falling, Bayrou is rising, rising. (Sarko too is trending down? Who notices?) To beat the effect of the MSM that are not favourable to her, Royal has a big fight on her hands. I'm not sure she always fully realizes to what extent. She can fight from the roots, that's a very good idea. I think she would also gain by appearing more "real", less glam and smiley, less stilted, in her visuals (TV, photo). The two (roots and a simpler image) happen to go together. It's not easy. An ambitious woman who is competing at the top with men needs armour-plating like a fifteenth-century knight, and it tends to restrict movement. If Royal can work her way out of that trap and go over as more simple and genuine, (something Sarko has been trying to do using voice-and-manner coaching), I think she has an excellent chance against a divided right.
Earlier diaries on French elections:
On European Tribune:
Ségolène Royal - a rising star in French politics? by whataboutbob
Sègo vs Sarko: presidential poll update by whataboutbob
Finally it is clear: 'Blairite' = critical of the moderate left by Jerome a Paris
Big Ségolène Royal interview in Le Monde by Jerome a Paris
Sarkozy feeling increasingly nervous, turns to Blair for advice (and other thoughts) by Ben P
Royal strides ahead? by afew
Jospin Pulls Out (again) by afew
Tale of Three Candidates (Poll and Update) by afew
French Socialists: live blogging by Alex in Toulouse
PS Primaries - Voting has begun by afew
Ségolène Royal chosen as Socialist candidate by Jerome a Paris
Ségolène Royal, the Middle East and French politics by Jerome a Paris
French candidates go wild on wrong problem by Migeru
Not Monsieur Hulot by afew
Press Review: An American Neo-con with a French Passport by redstar
Sarkozy loves Anglo-Saxon model, calls Blair "one of us" by Jerome a Paris
José Bové Wades Into the Fray by afew
Trendspotting: French Presidential Race. by ManFromMiddletown
11 in a row for Sarko by afew
Royal and the Roots by afew
Roots vs MSM, synthesis? by afew
Ségolène's big day - 'With me, never again will politics take place without you' by Jerome a Paris
Bayrou's Europe speech by oldfrog
Oh no the commies might be back! by Jerome a Paris
Sarkozy's unrealistic pledges by whataboutbob
The Republic of Far Far Away by afew
French election + Swedish media by someone
French elections? (Here we go...) by afew
France Islam and Fantaisy by afew
A good Frenchman by Jerome a Paris
Bayrou Rising by afew
On DailyKos
Ségolène Royal chosen as Socialist candidate for French presidential election
French elections (II) : Royal in the Middle East
French elections (III) - Royal is -gasp- a lefty