Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.

Polls A La Française

by NordicStorm Tue Apr 24th, 2007 at 04:06:01 AM EST

Those of you who read the open threads regularly might know I've been conducting a minor experiment over the past few weeks. Inspired by the poll trackers at Pollster.com, I took every poll on the French presidential election I could get my hands on and put them in an Excel file. I then plotted the average of the five latest published polls for each of the four top candidates (Ségolène Royal (PS), Nicolas Sarkozy (UMP), François Bayrou (UDF) and Jean-Marie Le Pen (FN)), the resulting graph being a trend estimate of sorts. As I'm posting this, it's the day after the first round of the French presidential election; time to take a look at how well my trend estimates held up.

From the diaries - whataboutbob

The Pollsters
There were six organisations conducting polling of the presidential election:

Of which IPSOS was the most prolific during the last few weeks of the campaign, releasing six polls a week, whereas the other pollsters published their polls less frequently, for example once a week or every five days, or entirely irregularly. During the last week, some of the other pollsters released two or three polls, some even releasing two polls on the same day!

Lies, Damned Lies And Polls
My poll averages weren't without their flaws. The first major issue was that of absentions (that is, people who weren't planning on voting for any candidate) and undecided voters. Some pollsters mashed the numbers together, while some only reported the abstentions.

The other major issue was one of frequency. The pollsters released their polls at very different frequencies, meaning that every calculated average would not only be an effect of possible changes in opinion, but also of which pollster had happened to release a poll lately and what particular biases in the numbers their respective methodologies would cause. Beginning in March, IPSOS started releasing a new poll six times a week. This skewed the average towards IPSOS's findings, as any given average point from thereon out would contain two or three of IPSOS's latest polls.

And then there's of course various issues with the polls themselves (undecideds, redressement, selection of people to call etc).

The Data
Below you'll find two graphs: the first one contains the trend lines, starting January 8 and continuing until April 20 (the last day before the election first round polls were allowed to be published); the second one contains all the poll numbers in relation to the trend estimate.
I've also kept track and am keeping track of the polling for the second round; those graphs will be posted in a another diary after the second round on May 6.

If the graph was right, the election results would have been somewhere in the ballpark of:
Sarkozy: 28.3%
Royal: 24.3%
Bayrou: 17.9%
Le Pen: 13.9%

The actual election result was (98% counted):
Sarkozy: 31.1%
Royal: 25.8%
Bayrou: 18.6%
Le Pen: 10.5%

The last average before the election was thus reasonably close on Royal's and Bayrou's eventual numbers, while Le Pen was overestimated, and Sarkozy was underestimated.
The last polls before the election had this to say:

Le Pen12.5%16.5%13.0%13.5%14.0%14.0%

No pollster had everyone pegged; the closest was BVA.

Events Affecting The Poll Numbers
Another important question is that of how much effect various events throughout the year had on the actual poll numbers. Two events in particular interested me.

Sarkozy was officially elected (though unopposed) as the UMP candidate for president on January 14, 2007. Shortly thereafter Sarkozy had his best month as far as poll numbers go, one pollster even giving him 36%. One can wonder how much of his rise can be attributed to the UMP congress (or, should I say, the media coverage of the congress), seeing as it had already been known for some time he would be candidate and the election was ultimately a formality.

Royal made a well received appearance on TF1's J'ai une question à vous poser on February 19, 2007. At that point she was already slightly rising in the polls, but it does look like she received a slight - and shortlived - bump in the polls.

As everyone underestimated Sarkozy and overestimated Le Pen, so did the poll average. My assumption was that all polls would be somewhat wrong, and that averaging would bring us closer to the truth. But if all the polls are wrong in the same way, averaging them won't remove that error. Ironically the poll average came the closest on Bayrou, who had been said to be rather difficult to poll.

Excellent! (Get out the vote!)

"Once in awhile we get shown the light, in the strangest of places, if we look at it right" - Hunter/Garcia
by whataboutbob on Mon Apr 23rd, 2007 at 03:55:22 AM EST
Thank you for effort, Nordic.  It was nice to know how it was moving, even if this result is so unpalatable.

Our knowledge has surpassed our wisdom. -Charu Saxena.
by metavision on Mon Apr 23rd, 2007 at 06:35:40 AM EST
quoting a margin of error?

I'd estimate it around 3%. So with the exception of Sarkozy and Le Pen who were on the edges of that, your averages were very close indeed.

Maybe for the next round drop CSA and possibly IFOP if they're reliably unreliable?

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Apr 23rd, 2007 at 06:50:31 AM EST
Some of the pollster say "around 3% for the big candidates."
Maybe I can cook up some sort of error-correcting formula for calculating the average based on how wrong the pollsters were in the first round. Then again, that'll probably end up being even more wrong...

"The basis of optimism is sheer terror" - Oscar Wilde
by NordicStorm (m<-at->sturmbaum.net) on Mon Apr 23rd, 2007 at 08:17:29 AM EST
[ Parent ]
MEthods used by French polling do not allow for calculation of a statistical margin of error. The margin of error is based on the ratio of the sample size to the overall universe, but that only applies if the sample is presumed to be random. The sample sizes generally used (>900) would be sufficient to generate a MOE of +_3.

However, in a multi-candidate race, MOE is less relevant because # of different possible outcomes within the "fork" (damn, I've forgotten the term we use in English for this, spread? range?) for each candidate generates multiple potential outcomes for the entire field. So, there is an entire series of possible outcomes for all the other candidates with LePen at 14 and a whole other series of possible outcomes with LePen at 11. And likewise, for each candidate whose score is more than the range (in this case, for each of the big 4).

Hence the claim you see some places of a margin of 3 for each of the major candidates. But thats not statistically accurate.

French institutes all use a "representative sample" which doesn't allow for margin of error. It is also generally hard to constitute if one doesn't have a clear sense of the make-up of the overall electorate.
But either the institutes have gotten very good at estimating a representative sample, or last night they just got lucky.

by desmoulins (gsb6@lycos.com) on Mon Apr 23rd, 2007 at 01:44:06 PM EST
[ Parent ]
Here was my prediction this Sunday:
Royal: 25%
Bayrou: 22%
Sarkozy: 21%
Le Pen: 16%

The only one I was even remotely close on was Royal. Don't ask me how I arrived at 21% for Sarkozy; it was wishful thinking in the extreme.

"The basis of optimism is sheer terror" - Oscar Wilde

by NordicStorm (m<-at->sturmbaum.net) on Mon Apr 23rd, 2007 at 08:10:54 AM EST
Thanks for your work NordicStorm!
by Laurent GUERBY on Mon Apr 23rd, 2007 at 04:59:10 PM EST
In the last elections everyone underestimated Le Pen. Did pollsters correct too much this time to find the much-fabled 'hidden' Le Pen voters? Did Sarkozy's number get affected most by this? Or was it all because how the undecideds broke?
by nanne (zwaerdenmaecker@gmail.com) on Tue Apr 24th, 2007 at 06:02:50 AM EST
Pollsters multiplied Le Pen answers by a certain amount to estimate his score, essentially to reach the same amount of people who voted for him in 2002.

But a fair amount of these voters who voted Le Pen in 2002 but didn't say so to pollsters, ended up voting for Sarkozy. Libération published the maps of the evolution of the Sarkozy vote and the Le Pen vote : Sarkozy rised exactly in the same places Le Pen went down.

Pollsters counted the proper amount of Le Pen voters, except a fair amount of them voted for the other anti-immigration candidate, Sarkozy.

Un roi sans divertissement est un homme plein de misères

by linca (antonin POINT lucas AROBASE gmail.com) on Tue Apr 24th, 2007 at 06:12:03 AM EST
[ Parent ]
The Huffington Post (left-hand column) has a strong warning by Tony Hendra about the misery a Sarko victory will cause to France and the EU:

A George Bush with French subtitles!

by Quentin on Tue Apr 24th, 2007 at 07:05:00 AM EST
It would open the floodgates of neoliberal reform in both France and the EU. By the time the French get around to getting rid of him in 5 years' time, there won't be much of the social EU left to save.

Regarding France, there is still the chance of cohabitation if instead of cannibalising a losing Royal, the Socialist Party builds on her momentum for the parliamentary elections. Also, Jerome (and others?) have predicted great social unrest if Sarkozy gets a parliementary majority and starts implementing "reforms".

Now off to read the Puffington Host...

"It's the statue, man, The Statue."

by Carrie (migeru at eurotrib dot com) on Tue Apr 24th, 2007 at 09:35:45 AM EST
[ Parent ]
Huffington Post: An Open Letter To My French Friends: Don't Let Sarkozy Turn France into a Red State! (by Tony Hendra on 04.23.2007)
Très chers amis,

I'm worried. Really worried. How can you have made Nicolas Sarkozy, the pint-size tough guy of the hard-right the leading contender to be next President of France? Not that I want to meddle in your internal politics, but since he's gotten there in large part by claiming to be pro-American, I think I should explain exactly what that means.

"It's the statue, man, The Statue."
by Carrie (migeru at eurotrib dot com) on Tue Apr 24th, 2007 at 09:39:54 AM EST
[ Parent ]
I do think its possible, if Bayrou leans left a bit without losing all the UDF deputies and if a good bit of his electorate goes for Royal yet Sarkozy wins narrowly with strong support of the LePen electorate and appoints a government that reflects his rightward turn (and his vindictive nature, ie a government without the UDF), then we could see a lot of triangular or even quadrilinear legislative elections.

In that case, there could be the makings of a PS-UDF majority, if its based on an agenda of institutional reform that would culminate in new elections. No one will vote for a five-year left-right cohabition.

I think thats only envisageable though if the perception is that Sarkozy's government will be precisely what most of her expect -- a hard-line neo-liberal, hard-line authoritarian government. He's too damn smart though to do that, and the PS and UDF aren't nimble enough.

I expect he'll co-opt the center and the hard-right leave the PS to splinter.

by desmoulins (gsb6@lycos.com) on Tue Apr 24th, 2007 at 01:06:51 PM EST
[ Parent ]
Here's the Royal vs. Sarkozy matchup:

Royal polls in the 46% - 49% range, and Sarkozy thus in the 51% - 54% range...

"The basis of optimism is sheer terror" - Oscar Wilde

by NordicStorm (m<-at->sturmbaum.net) on Tue Apr 24th, 2007 at 02:56:05 PM EST

Go to: [ European Tribune Homepage : Top of page : Top of comments ]