RTS. For this presidential election as for the previous ones, the RTS is one of the Swiss and Belgian media to provide estimates of the result before 8 p.m., freed as they are from French law. Two weeks ago, Swiss Radio-Television was not deprived…
Presidential results near you
[Mis à jour le 24 avril 2022 à 18h47] “Macron and Le Pen neck and neck at 24%, according to a first poll.” It is 6:15 p.m. and not 8 p.m. this Sunday, April 10, when the RTS publishes this title in its live dedicated to the French presidential election, on its website. Swiss Radio-Television thus leaked an estimate of the result of the ballot even before the polling stations were closed in France. And the Swiss media, of public service, finally says very little about the source of this estimate, which will prove to be eccentric to say the least less than two hours later. The RTS is content to evoke a “first poll at the exit of the polls” then “a survey carried out by one of the major French polling institutes”.
In these RTS figures, Emmanuel Macron and Marine Le Pen were therefore neck and neck at 24% ahead of Jean-Luc Mélenchon at 19%. The information will be so significant for the Swiss media that the latter will take the trouble to post it on its Twitter account a few minutes after its publication on its website.
An hour later, the RTS revised its initial estimates. “Three new polls still give Emmanuel Macron the lead,” she said, citing other French-speaking media and then relaying her info once again on the social network.
Unsourced and… false RTS estimates
These estimates before 8 p.m. from RTS and many other Swiss and Belgian media are impossible in France, where it is forbidden to communicate on the outcome of the ballot before the end of the vote at the national level. If these foreign media are freed from this ban, it is sometimes for the better, but often for the worse. In 2017, some had indeed been right. But it is clear that for the first round of this presidential 2022, if the two finalists announced have been confirmed, the “elbow-to-elbow” promised at the end of the afternoon was a pure illusion. Or, to use the term of the Survey Commission in France, which had published a statement before the ballot, “the fruit of rumors or manipulations” to which “no credit should […] be granted”.
The reliability of the RTS estimates can indeed be called into question when we know that the 8 main polling institutes (BVA, Elabe, Harris Interactive, Ifop, Ipsos, Kantar, Odoxa, OpinionWay) have assured the said commission that no of them would carry out “exit polls” during the election. Of course, one of these institutes was able to carry out this kind of investigation in secret by questioning the voters about their choice via the Internet or “physically”, at the exit of the voting booth. The Survey Commission has also indicated that it had contacted the Harris Interactive Institute on April 10 “to ensure that the commitment made not to carry out an exit poll had been respected”. But exit polls are nevertheless used less and less by pollsters. Deemed unreliable after years of practice, in particular because of their methodology based on declarations, they have been gradually abandoned.
Are there reliable estimates?
So where do the figures that circulate and that certain Belgian and Swiss media such as the RTS can publish come from? It has been ten years now, since the 2012 presidential election at least, that the question remains unanswered. That year, the Survey Commission, angry to have found leaks in the middle of the afternoon, had seized the Paris prosecutor’s office which had opened an investigation. Two Belgian media, a Swiss media, a website based in New Zealand but also AFP, which had decided to relay the estimates, and a Belgian journalist who had revealed himself on Twitter were cited. But the investigation will not identify the source of the figures.
However, there are reliable estimates established by polling institutes on the day of the presidential election. These are even essential for television channels to be able to deliver a solid result from 8 p.m. But nothing to do with “exit polls”. This time, the pollsters are taking a survey of the actual results in several hundred polling stations deemed representative of the vote in France. From 7 p.m., closing time of many polling stations, investigators will communicate the results of the counting of all 100 ballots to the institute. A powerful algorithm then establishes national projections. More reliable than the usual declarative method of polls, this calculation of estimates has gradually imposed itself in practice on polling days. But with this method, it is impossible to obtain reliable estimates with 7:45 p.m., according to the pollsters themselves.
The leaks of estimates, national sport in Switzerland
Under the law of July 19, 1977, the publication of any indication of the results of a political election is prohibited before the closing of the last polling stations at 8 p.m. French law provides for a reserve period for the media which began Friday at midnight and does not end until the polling stations are closed, with the aim of preserving the neutrality of the ballot and not influencing those who do not have not yet fulfilled their civic duty. Polling stations closing at 8 p.m. in major cities, this French legislation is thus a boon for all French-speaking foreign media. But for about ten years now and at each election or almost, estimates of the election results have been posted online by the Swiss and Belgian media (Le Temps, La Tribune de Genève, RTS, but also, in Belgium , Le Soir, La Libre Belgique or RTBF) from the middle or even sometimes the beginning of the afternoon.
Like their Belgian colleagues, the Swiss newspapers, Time and “TDG”, but also the RTS are turning their Web edition upside down to follow the ballot from Paris, thanks to their correspondents based in the French capital. This is how they happen to spread leaks on their sites as well as on the Web, in other words the first results of the French elections in the form of polls, before the legal time of 8 p.m. All while on the other side of the Alps, faced with the fear of a heavy fine, we impatiently await the verdict of the ballot box.
These figures, which water thirsty social networks, which can no longer wait since the start of the day for clues on the ballot, are often quite precise, giving for example during the last presidential election the names of the two finalists with a percentage of vote very close to what had been established during the final result. Information on the election of such and such a Member of Parliament or such and such a European or regional councilor is also regularly given during legislative or intermediate elections.
Pioneer of this circumvention of French law, which obliges the media of France to wait 20 hours before revealing figures, the RTS had, from the presidential and the legislative elections of 2012, set up a special edition on the Web, proposing at the end of the afternoon the first estimates on its Twitter account and on its website. During the French legislative elections in particular, it had received such an influx of visitors to its site that the latter had ended up experiencing delays. Since then, for Swiss media that are not subject to French law, poll leaks have become a national sport. Sport in which they again excelled in the first round.