Le Pen Fails to Land a Knockout Blow

Aaron Timms & Eric Falcon

  • Last night’s debate does not appear to have fundamentally altered the dynamics of the race
  • The immediate overnight impact of the debate was fairly balanced between the two candidates
  • Predata’s signals don’t show sufficient digital momentum for Le Pen that we think the polls are wrong in pointing to a Macron victory, but they do suggest the final outcome might be closer than the 60%-40% split that most polls continue to show
  • In particular, Le Pen has extended her “digital dominance” over Macron in the key swing regions
  • Projected turnout is low but not low enough for it to be a decisive factor tilting the likely outcome to Le Pen

Note: Access a quick visual primer on Predata’s digital momentum signals, which are the basis for most of our analysis of major political and policy contests, here. A guide to how to read Predata’s digital momentum signals for the French election can be found here.

Last night’s debate was spiky, vicious, and — for the neutral, at least — delectably watchable television. Marine Le Pen went on the attack, haranguing her opponent at every opportunity, dismissing him as a “banker” who’s “cold” and will make France “submissive” (to corporate interests, to the U.S. and Germany, to globalization). Emmanuel Macron, for his part, lambasted Le Pen as a “parasite of the system” whose “logorrhea” and “lies” conceal the fact that she is “proposing nothing” in terms of real, substantive policy and has no mastery of the facts. Le Pen’s best line was her claim that after Sunday, “France will be run by a woman — me or Madame Merkel.” Macron’s debate will be remembered chiefly for his repetition of the words, delivered with a shake of the head in a tone of weary condescension, “Madame Le Pen.”

In the digital domain, however, the debate does not appear to have fundamentally altered the dynamics of the race. Further effects may become apparent on Friday and Saturday, but the immediate overnight impact of the debate was balanced between the two candidates, who both saw their digital campaign signals rise. Le Pen may have edged proceedings slightly — eight of the top 10 drivers of the overall online conversation about the election yesterday were pro-Le Pen sources — but she failed to land a knockout blow and there was no clear “winner.” As a result, there is no change to our base case for the runoff: our signals don’t show sufficient digital momentum for Le Pen that we think the polls are wrong in pointing to a Macron victory, but they do suggest the final outcome might be closer than the 60%-40% split that most polls continue to show. We say this for four main reasons.

1. The overall digital conversation is balanced between the two candidates.

The dynamic in the overall digital conversation has not changed since last week. Le Pen dominates the digital campaign when only Wikipedia and Twitter sources are included in the signals (see first chart above) — this being the method we’ve used to analyze the digital campaign to date and the method that accurately reflected the outcomes in previous electoral contests such as Brexit, the Italian constitutional referendum and the U.S. election. However, once YouTube sources are added to the signals, the relationship is flipped and Macron dominates (second chart above).

This means, essentially, that Macron’s campaign is generating more short-term enthusiasm and buzz around each successive wave of messaging, since YouTube videos rarely garner much interest beyond the third day after their publication, while Le Pen has a larger core of true believers online. It’s worth bearing in mind that many of Macron’s closest advisors have no “online presence” in a form that would allow them to be included in our signals for his Wikipedia and Twitter campaign; this means the Wikipedia- and Twitter-only version of the digital debate (first chart above) may have a moderate Le Pen bias. Either way, it’s clear that the signals show neither candidate is truly dominating the online debate. The digital contest appears evenly balanced.

2. Momentum for Le Pen in the swing regions is clear but not election-defining.

Recall that the “swing regions” are those three départements we identified as being a key barometer of Le Pen’s second round hopes. The first round split between Macron and Le Pen in those regions, which are removed from the candidates’ traditional heartlands (Paris and the west for Macron, the north and the south-east for Le Pen), was as follows:

  • Doubs: Le Pen 23.45%, Macron 22.5%;
  • Isère: Macron 24.77%, Le Pen 22.33%;
  • Loiret: Le Pen 23.53%, Macron 23.48%.

Last week Le Pen and Macron appeared to be neck and neck in the digital campaign in the swing regions, which we took as a bad sign for Le Pen. In the last few days, however, the contest has swung decisively in her favor, as the chart above shows. It’s clear that this is more to do with a drop in intensity from Macron’s campaign than any significant rise in digital engagement with Le Pen’s. The Macron campaign’s official Twitter account for the Doubs region, for example, has tweeted 20 times since the first round 10 days ago; Le Pen’s Twitter account in Doubs has tweeted close to 500 times over that period. A similar pattern is observable in the other swing regions: Le Pen’s digital campaign is far more active than that of her rival, and her on-the-ground cadres are doing everything they can to broadcast her message.

The urban-rural divide — really a divide between big cities, on one side, and small cities, towns and the countryside, on the other — was, we know, a major factor in both the Brexit vote and the U.S. election. The issue has received comparatively little attention throughout the French election but could have an important influence over the outcome on Sunday. Our signals suggest Macron’s team may be slightly complacent about his prospects of winning the vote outside the big cities. It’s possible, of course, that he won’t need to carry the swing regions to win more than 50% of the vote. It’s also true that the “digital gap” that’s opened up between the two candidates in the swing regions reflects a straightforward divergence in output — Le Pen’s campaign is simply producing more stuff — and that this surfeit of content may not necessarily reflect or anticipate a divergence in patterns of voter behavior on Sunday.

That said, we know two things from previous electoral contests with a heavy digital dimension, such as Brexit and the U.S. election:

  • it’s dangerous to overlook the importance of the regional vote; and
  • victory in the battle of content — the contest to see who puts more material online — can be an important factor in the final outcome.

The furious intensity with which Le Pen’s local representatives are continuing to broadcast her message in these parts of the country, and the comparatively lackadaisical pace of messaging from Macron’s camp, creates some basis to suspect the polls may be underestimating Le Pen’s true levels of second round support.

3. The digital messaging of the Fillon and Mélenchon camps shows no evidence of a big swing to Le Pen.

Endorsements have not broken convincingly in Le Pen’s favor since the second round. Of the first round contenders, only right-nationalist Nicolas Dupont-Aignan has explicitly endorsed Le Pen, who has promised her erstwhile rival the prime ministership should she win on Sunday. The real second round intrigue surrounds how the supporters of François Fillon and Jean-Luc Mélenchon will vote. Fillon himself has endorsed Macron, but the latest polls suggest a quarter of his supporters could vote for Le Pen. Mélenchon has refused to make an endorsement but urged his supporters not to back Le Pen; two-thirds of them will apparently register a blank vote or abstain on Sunday, according to the results of a poll released by his own party.

If Le Pen is to stand any chance of winning on Sunday, she will need many Fillon and Mélenchon voters to mobilize on her behalf. The social media conversation shows no sign of a movement in that direction. Online, most major campaign advisors and surrogates of the two defeated first round candidates have used the last 10 days for two purposes: 1) to lament the absence of their candidate from the second round campaign, and 2) to urge a vote against Le Pen.

Last night’s debate offered a good illustration of this dynamic. The debate was defined primarily in the negative, as a confrontation of two critiques rather than a contrast of two positive policy visions. The candidates’ performances, in line with that overriding negativity of tone, were memorable mostly for their weaknesses than their strengths: Macron was arguably weaker on social and political questions (the question of whether France committed crimes against humanity in Algeria, the endorsement of his campaign by Islamic organization UOIF), while Le Pen was probably weaker on the economy (her increasingly confusing project for leaving-but-not-leaving the Eurozone, the question of how she will fund a reduction in the retirement age to 60, and so on).

In keeping with the tone of the event, criticism of Emmanuel Macron by the entourages of Fillon and Mélenchon on social media throughout the debate was fierce, and much of that criticism mirrored the very message pushed by Le Pen; for instance, Yves d’Amécourt, a Fillon surrogate, retweeted a comment from his old boss to the effect that Macron represents the continuation of François Hollande’s administration and his candidacy is a “marketing” gimmick. But that criticism was in most cases mixed with an unambiguous signal to voters to do everything on Sunday to block Le Pen’s path to power.

Interestingly, that signal emerged most strongly from the Mélenchon camp — and from key advisors such as Danielle Simonnet, Sophia Chikirou, and Alexis Corbière in particular. (The tweet below, from Chikirou, reads: “Tonight, the mediocrity of the two candidates risks making us vote for the one who’s not racist: the ultimate argument.”) Le Pen has worked furiously to woo Mélenchon’s supporters in recent days, and her repeated description of Macron as “submissive” last night was perhaps a subconscious appeal to the vision of an “unsubmissive France” promised by the name of Mélenchon’s own political movement. There’s no evidence, beyond isolated cases, that these efforts at seduction have proved effective.

4. Turnout looks likely to be low, but not low enough to give Le Pen a decisive advantage.

Abstention has always loomed as the key risk weighing on the second round. In theory, a low participation rate should benefit Le Pen, whose electoral base is more committed than that of Macron. But turnout needs to be drastically lower than normal for Le Pen’s chances of winning to increase meaningfully. If the latest polls are correct, only 75% of the electorate will turn out to vote on Sunday — well below the 80% mark that most presidential runoffs in France over the last few decades have reached, but still above the historical low of 69% registered in the second round of the 1969 election. Our signals support this view.

The level of Predata’s France sector rollup signal — a measure of country-level political volatility built from equally weighted component sectors — gives an indication of the impact of the election on the overall digital conversation related to France. Because the signal is stable in its composition across its history, we can also do credible comparisons to past signal activity (see chart above).

As we noted last week, the signal began rising much later before the first round this year than it did in 2012, but eventually reached the same level on the day of the vote, consistent with turnout figures that exceeded many pollsters’ expectations. Since the first round, however, the signal has not reached the heights it attained in 2012, indicating that this year’s election has had a smaller impact on the French digital landscape than the last presidential race. The signal has remained stagnant over the past few days; by this point of the 2012 campaign, it had already begun a sharp ascent. This pattern is consistent with the low turnout projections in the polls.

More than anything, France heads to the polls unenthused — but not so unenthused, it seems, that turnout will hand a decisive advantage to Le Pen.

Bottom line: Even if Macron wins, Le Pen will be a bigger threat in 2022.

For Le Pen to overcome her deficit in the polls and win the second round, by this stage of the campaign there would need to have been evidence of one or all of the following:

  • a clear swing in overall digital momentum her way;
  • the threat of mass abstention or a catastrophically low projected turnout rate; or
  • a wave of pro-Le Pen endorsements by her first round opponents.

In the event, none of these factors has materialized; projected turnout is low but not low enough for it to be a decisive factor tilting the likely outcome to Le Pen. Le Pen does have two things in her favor, however:

  • the prospect that Mélenchon supporters will stay away from the ballot box in large numbers; and
  • a dominant digital presence in the key swing regions.

This suggests the final outcome, while still handing the presidency to Macron, may end up being closer than polls today suggest.

Even if the polls are right and Le Pen only wins 40% of the vote on Sunday, that would still represent, assuming a participation rate of 75%, around 12 million votes in her favor. In the first round of the 2012 election, Le Pen won 6.4 million votes; in this election’s first round she gained 7.7 million votes. To progress to 12 million votes in the runoff, from that base, would mark spectacular electoral progress for the Front National over the last five years, and set the party up for a serious tilt at the presidency in 2022. ⏪

Aaron Timms is Predata’s Director of Research. Contact: aaron@predata.com.

Eric Falcon is a Predata analyst. Contact: eric@predata.com.