The tension is growing in pre-election France. Stephen “Etienne” Roop, a political scientist, expert in polling, tries to enlighten us here.
Subject: One week out, is the French first round turning into a photo finish?
The French media are breathless this weekend over a new poll which seems to say of the impending April 23 first round presidential vote: it’s now a foursome, and any combination of two can win. I’d say we need to take a deep breath before buying in quite yet.
For months a familiar two-part scenario for April 23 has seemed fixed in the minds of observers and ordinary citizens. First part: the two leading candidates in the polls, newcomer Emmanuel Macron and outsider Marine Le Pen, would continue their leads and easily win the first round and face each other in the second round, which is limited to the top two vote-getters in the first round. Second part: Macron would demolish Le Pen in the second round.
But on Friday, April 14, Le Monde said there was developing “a situation unprecedented in the last 50 years….” It based its claim on the newest poll from its highly respected pollster Ipsos, hot from the field April 12-13. The poll results are very clear: the April 23 race us suddenly a foursome. Just three points separate the four top candidates. As they have for months, Macron and Le Pen still lead, but now with only 22% each. Hard right Républicain standard bearer Francois Fillon, long an also-ran candidate dogged by scandal but bouyed by incredibly loyal supporters, is an also-ran no more, at 19%. The big drama is the unexpected consolidation on the left. Jean-Luc Mélenchon, head of his own party France Insoumise, has surged dramatically to 20% while the establishment socialist party’s candidate Benoît Hamon has collapsed. The margin of error in the Le Monde poll is +2.7%. That’s a startling amount of overlap, and thus uncertainty, among the four top candidates. A clear implication of Le Monde‘s poll is that both the current poll leaders, Macron and Le Pen, could be excluded from the second round. Since Le Pen has been the cynosure of the 2017 elections almost since the previous elections in 2012, her exclusion is certainly a stunning possibility. Since Macron has been, for many (including perhaps many abstainers), the genial antidote to Le Pen, his exclusion may be no less stunning. One or both of these exclusions would make for the final boulversement of this extraordinary presidential campaign.
How persuasive is Le Monde‘s Ipsos poll? Le Monde‘s voice is disproportionately influential in France, its pollster is highly respected, and the Ipsos poll results are the talk of April 23 watchers everywhere. But there are five other quite current polls. All six agree that the socialist party candidate Hamon is foundering and far-left candidate Mélenchon is soaring. But the April 23 messages of the other five are not uniform otherwise. BVA’s poll, in the field April 12-14, also shows a three point margin separating the top four candidates. Ifop-Fidicial, in the field April 11-14, shows a four-point margin. Does the one-day-more-recent field sample explain Ipsos’s tighter result? Unlikely. Odoxa, in the field exactly the same days as Le Monde‘s Ipsos poll, shows Macron and Le Pen close but a six point spread between the top four candidates. The Harris and Elabe polls likewise show Macron or Le Pen close, but a five-point spread between the top four. Harris was in the field April 11-13, Ifop-Fiducial April 11-14. It’s hard, without torturing the data, to get a consistent pattern here–for example, that nearness to today of time in the field leads to tighter results.
Focusing on one poll, however tantalizing its import, is dangerous. Better to look at the aggregate of these six current polls. I take a simple arithmetic average of the results and do not adjust for each poll’s sample size; this average is for voters who express a preference. The two leaders are clustered together: Macron is at 23.08% support, with Le Pen just trailing at 22.41%. The two followers are clearly behind, but close to one another: Fillon is in third with 19.42%, and Mélenchon just trails him at fourth, with 19.25%. So the spectrum is a bit more than four points. (Compare the latest poll compilation result in Le Huffington Post: Macron, 23%; Le Pen, 22.7%; Fillon, 19.5%; Mélenchon, 19.3%. Almost identical both in terms of candidate support and spectrum!) I did the same computation with each pollster’s previous poll, just to measure dynamics a bit. The in-field dates are more skewed than ‘for the current six polls (thus álso making a comparison with prior composite results in Le Huffington Post impossible). The results in the immediately prior polls are the same shape: Macron barely ahead of Le Pen, 23.4% to 23.3%, and they lead the other two, who are themselves close, with Fillon at 18.92 and Mélenchon at 18.33. The spectrum here is slightly over five points. So the race is tightening slightly, based on the dynamics these six pairs of polls reveal: Macron and Le Pen are drifting downward (as they have been for several weeks–especially Le Pen); Fillon’is rebounding slightly; Mélenchon alone is really rising. But I wouldn’t, yet, quite call it a foursome.
Entering the last week, each of the four candidates has strengths and weaknesses . Le Pen’s voters are the most loyal, though some data I have seen recently suggests some desertions after the debates on March 20 and April 4, and, again, her numbers have been trending downward over the last several weeks: two months ago today her Le Huffington composite was 26%, one month ago today it was 26.1%. Fillon also has loyal voters, and he has managed to fight back extremely bad press over the no-show jobs he gave his wife, for which they are now both in deep trouble with the French legal system; nonetheless most non-supporters wish he had dropped out of the race when that system began putting the screws to him, and they are unlikely to reward his persistence on April 23. So it is no surprise he has moved barely at all over time: two months ago he was 19.6% in the Le Huffington composite, 18.8% thirty days ago, 19.5% currently. Macron is the boy-wonder, young and appealing, and widely believed, by many people for a long time, to be the likeliest next president; nonetheless he’s a newcomer, and–perhaps as a necessary correlate of his rhetoric about transcending the left-right divide of French politics and taking good ideas wherever he finds them–his platform can be annoyingly vague, so subject to onslaughts from his more ideologically predictable rivals. Late in March he (barely) overtook Le Pen to top the Le Huffington composite (24.9% v. Le Pen’s 24.8%) but Macron had already peaked and both have since then drifted down together in lockstep, Macron always just barely ahead of Le Pen. Mélanchon has been around French politics for a long time and has a lot of baggage from his far-left past; nonetheless his oratory and his crowds of late have been impressive, and he has fairly vaulted upward in the last four weeks: sixty days ago, he was 11.8% in the Le Huffington composite (with Hamon over 15%), thirty days ago just 12.1% (to Hamon’s 13.1%). Mélenchon has gotten to his current 19.3% in the Le Huffington composite in no small part because of his excellent performance in both televised debates, and very adroit followup; he seems to have neutralized most of the whiff of old party apparatchik that has long bedeviled him, though of course not everyone is persuaded. (The folksy sweater doesn’t hurt!)
Plainly the question on every observer’s mind this week will be: can Mélenchon’s rise continue until April 23?
Mélenchon’s favorability among likely voters is, no surprise, rising even faster than his poll support. That rising favorability is probably the key to any further Mélenchon poll progress. Support might continue to come from Hamon voters jumping ship. Perhaps abstainers will no longer abstain.
Of course, Mélenchon’s further progress is not inevitable. The other candidates are unlikely to sit quietly for the last week. And while some voters may shift from another candidate or abstention to support Mélenchon, others may react to stop or slow his progress.
We’ll just have to hope the remaining polls of this coming week will give us some clues about where this suddenly much more interesting April 23 contest is going.