What does the social media crystal ball say about the biggest global event of the day, the French presidential campaign? On social media, Marine Le Pen and Jean-Luc Mélenchon seemed to be the favorites, running neck in neck on Twitter the day before the election. This is surprising since they are the “outsiders” (especially Mélenchon, who returns to the big leagues after an interesting showing in 2012), representing the ends, rather the middle of the political spectrum. More important, in opinion polls, Le Pen seems to be in slight decline, while Macron (a centrist) and even Fillon (the traditional “Republican” candidate) are inching up a bit.
Yet, this is not what happens in social media, or not on Twitter, YouTube and Google. In each of these media spaces, the candidates of the two main parties, the Socialists, and the “Republicans,” are trailing behind. Marine Le Pen and Jean-Luc Mélenchon seem, on the other hand, to be doing very well.
Both Le Pen, a populist and nationalist with a checkered ideological pedigree, and Mélenchon, who receives support from quarters as far as the remains of the Communist Party, garner a very high level of engagement on Twitter. Engagement is reflected by the number of retweets, especially since the April 10, when the campaign entered in a straight line. (Note 1)
I believe that social media engagement is quite revealing, since it was useful (albeit in retrospect), in explaining Donald Trump’s success. At least one French colleague, professor Bevort, seems to agree with the idea that social media can be a good predictor for political campaigns, if not in terms of outcome, at least in terms of process. Furthermore, he comes to the same conclusion, namely that Le Pen and Mélenchon are in an apparent tight competition and that mainstream candidates (Fillon, Hamon) or the never-elected Macron might get less support than what the opinion polls say because they are not as adept at or active in the social media game. Of course, only the results of the first round of the elections can validate his opinion or this analysis. Until then, let us consider the numbers below.
Le Pen, much to everyone’s surprise, did very well in opinion polls, too. Yet, she was a real player on Twitter throughout the entire campaign. As the chart below shows, she garnered more retweets per tweet both during and after the official opening of the campaign.
At the same time, Mélenchon came very strongly from behind, almost (but not quite) catching up with Le Pen in the last two-three weeks of the campaign (after April 10 – ie, the “campaign” cluster).
More important, calculating the total number of retweets for all the tweets of each candidate, we notice a significant reversal of rankings. In the pre-campaign period (before April 10), Fillon accumulated the largest volume of retweets, followed by Le Pen. Mélenchon, however, took the lead on April 10, followed closely both by Le Pen and Fillon. (Note 2) Macron, who for many seemed like a good centrist alternative to the leading contenders that represent two extreme opposite political visions, trails behind. Hamon, the socialist candidate is the weakest of them all.
Mélenchon’s lead might not be as strong as it seems, though, especially when we compare the overtime performance on Twitter of each of the five candidates. Looking at how many average retweets each candidate got per tweet, we notice that although Mélenchon came strongly from behind, in the last two weeks he seemed to have run out of steam. It is noteworthy to observe Le Pen’s (blue line) aggressive engagement (retweeting) presence.
Zooming in on the last three weeks, we see some things even clearer. All candidates, including Le Pen, started getting a lot more retweets; they were activating their basis very energetically. Fillon seems to have done a particularly good job in this respect. Only Mélenchon, while remaining at a relatively high retweet level, declined.
At the same time, Mélenchon might have one more arrow in his quiver. On YouTube, which is replete with his campaign clips, his name is searched far more often than the other four candidates’.
All in all, the French presidential campaign reminds us how important social media has become in activating the public for political action. The numbers presented above might announce a political surprise similar to that we witnessed in the American elections. In France, as our colleague Antoine Bevort announced, observers tended to ignore social media as an early warning signaling system for political change. This is similar to what happened in the US, where Donald Trump’s strong showing in social media, where he dominated Hillary Clinton in terms of engagement both on Twitter and Facebook, was discounted because the opinion polls constantly put him in the second place.
At the same time, France is the not the United States. Political activism in France depends on many more resorts and channels of communication, some of them outside and unseen from social media. Only 60% of French adults used social media in 2016. But, again, surprises have been seen. Stay tuned… Will come back again with an update on Monday, after the results are announced.
(1) The analysis reflects 11,978 tweets I collected on Friday, April 21, representing all the tweets issued by each campaign, except for one (Mélenchon), since January 1st. For each tweet I calculated the number of times it was retweeted. In addition, I calculated the average number of retweets for each tweet by time period (weeks and pre and during the “official campaign,” which started on April 10.) ForMélenchon I collected all available tweets from February 5th (latest available).
(2) Mélenchon’ s numbers for the pre-April period are partial, missing January. Thus, he might’ve in fact lead in that part of the race, too. If I get the numbers, I will redo this part of the analysi.