Frédéric Micheau is director of the opinion and politics department of the OpinionWay institute and teaches at the Paris Institute of Political Studies. He publishes
Le sacre de l'opinion - A history of the presidential election and the polls (editions du Cerf, 2022).
FIGAROVOX. - When did polls enter French political life? What influence did this have on the 1965 election?
Frederic MICHEAU. -
Polls appeared at the end of the 1930s, following the American example of George Gallup. Interrupted by the world conflict, their development resumed after the Liberation. But the parliamentary regime of the Fourth Republic hindered their activity. The transition to the Fifth Republic will create the conditions for their development. General de Gaulle's repeated use of the referendum to establish new institutions and periodically consolidate his political legitimacy is conducive to polls: the yes/no alternative facilitates the measurement of the electoral balance of power and limits the risk of error. The institutes are achieving their first successes. The presidential election of 1965 will give them their letters of nobility.
When it takes place, this first election by universal suffrage since 1848 is unknown in its operation. Most observers see it as an additional plebiscite and many anticipate the election of General de Gaulle in the first round, as indicated by the first voting intentions. But the electoral campaign, in particular on television, transforms the balance of power. Jean Lecanuet, who uses the new methods of political marketing, of which polls are a part, makes a breakthrough. The score of the man of June 18 drops sharply and the polls predict his tie-up.
This information meets with the greatest skepticism.
But on election day, the first estimate of the results validates the forecasts of voting intentions.
Against all odds, the institutes triumph.
By having correctly announced the occurrence of an event that seemed unrealistic, they gain legitimacy and impose themselves on the political and media scene.
Is the emergence of private institutes a guarantee of impartiality?
Today, as in the media landscape, the pluralism of the institutes constitutes a guarantee for the citizen: the diversity of the polls limits the risk of electoral maneuvers.
Frederic Micheau
This is because the institutes were put forward in 1965: the announcement of the head of state being put on the ballot was presented as a sign of independence from political power. No public body could have authorized this crime of lèse-majesté. It should be recalled in this respect that the results of the last electoral study carried out confidentially by the Central Directorate of General Intelligence of the Ministry of the Interior have been modified by order of the Minister. The publication of these official but counterfeit figures was imposed on
France-Soir
with the aim of giving credibility to the scenario of a victory for the General and thus to counterbalance the announcements of the institutes.
Today, as in the media landscape, the pluralism of the institutes constitutes a guarantee for the citizen: the diversity of the polls limits the risk of electoral maneuvers.
We can still see it today, the dynamics of polls can encourage a candidate to run for the supreme election, or on the contrary discourage him from doing so. Aren't polls dangerous for the full exercise of democracy?
Historically, several politicians have been encouraged to participate in the presidential election by their performance in popularity barometers and/or polls of intention to vote.
This is the case of Raymond Barre in 1988 or Édouard Balladur in 1995. Conversely, many personalities have given up declaring a candidacy because of disappointing figures.
Let us quote here the example of Valéry Giscard d'Estaing, dissuaded in 1969. Note especially that the majority of the potential candidates did not listen to the polls.
In 2011, François Hollande, nicknamed "
Mr. 3%
” because of his score at the time of the undivided domination of Dominique Strauss-Kahn, maintained his candidacy with determination against the polls.
The observation also applies to Jacques Delors, who remained deaf to the figures promising him an easy election in 1995.
Read alsoPresidential 2022: Macron remains in the lead, Le Pen is slightly ahead of Pécresse, Zemmour is stagnating, according to a poll
These various examples remind us that polls produce information on the electoral potential of a candidate, but do not provide information
a priori
on the effects that an electoral campaign can have.
It is therefore always up to the candidates to take this into account or not in their decision to participate in the election, a decision which is their own.
What will the law of July 19, 1977 change?
The law of 1977, which regulates the polls, brings two fundamental novelties.
First, it founded a polls commission, “
responsible for studying and proposing rules tending to ensure, in
the field of electoral forecasting, the objectivity and quality of the polls published or broadcast
”.
The action and control exercised by this independent administrative authority have made it possible to reduce fraudulent practices and manipulation operations based on the publication of opinion polls.
Polls [...] are still perceived as a potential vector of influence rather than as information in its own right.
Frederic Micheau
Then, this law establishes a ban on the publication of electoral polls in the week preceding the vote.
Wanted to preserve the sincerity of the ballot, this reserve period actually opened the door to all rumors and all attempts to intoxication voters.
In 2002, the legislator was forced to act to remedy this situation: he chose to reduce the period for banning the broadcasting of polls to 48 hours, rather than simply abolishing it.
This half-measure speaks of a persistent suspicion with regard to polls, which remain perceived as a potential vector of influence more than as information in its own right.
You come back at length to the Balladur-Chirac war. Is this the first rout of the polls in the history of the Fifth Republic?
Errors in assessing the score of certain candidates had already been observed before 1995. For example, in 1981, the result of Georges Marchais had been overestimated by the latest voting intentions, which had not detected the onset of electoral decline. of the Communist Party.
But the 1995 presidential election was the scene of the first major failure of the polls in France.
The order of arrival of the candidates was not correctly anticipated.
Lionel Jospin's unexpected first place plunged the Chiraquian camp into despondency, which was convinced that its champion would be in the lead as the polls indicated.
At the same time, Édouard Balladur was not as far behind as the latest figures suggested.
Associated with the frustration of defeat, this realization that qualification for the second round was within reach had aroused the ire of Nicolas Sarkozy.
The spokesman for the beaten Prime Minister had made a virulent charge against the institutes on the very evening of the election, live on television.
Have we learned since?
Every failure of the polls - how not to mention the earthquake of April 21, 2002?
- was the occasion for the institutes to question their methods.
The media, which have a major responsibility in the way poll results are presented to citizens, have also made pedagogical progress.
Do you see any parallels between this episode and the war between Le Pen and Zemmour in the polls?
Presidential campaigns have often been the subject of rivalries within the same camp.
The confrontation between Valéry Giscard d'Estaing and Jacques Chaban-Delmas in 1974, the competition between Raymond Barre and Jacques Chirac in 1988 or more recently the opposition between Jean-Luc Mélenchon and Benoît Hamon in 2017 resulted in an inversion of positions of the two candidates concerned.
The success of the former CNews columnist comes at the expense of the National Rally candidate, who sees a significant part of her 2017 electorate turn away from her.
Frederic Micheau
The competition between Marine Le Pen and Éric Zemmour is not strictly comparable to the configuration of 1995. Édouard Balladur and Jacques Chirac had the same mentor - Georges Pompidou -, belonged to the same party, exercised major responsibilities and had sealed a pact in name of their friendship.
We do not find today the tragic thickness of this frontal shock, which was increased by the personal heartbreak of the two men.
The only common point, the porosity of the electorates is evident in both cases.
RPR voters hesitated for a long time between the two enemy brothers.
The success of the former CNews columnist comes at the expense of the National Rally candidate, who sees a significant part of her 2017 electorate turn away from her.
You explain that Nicolas Sarkozy relied on the polls to try to influence the course of the 2012 campaign… in vain. Do the polls have a real impact on the election or do they remain a thermometer of opinion at a given moment?
It is all the ambivalence of the polls: on the one hand, they want to be the neutral and objective witness of the evolution of the balance of power.
They are in a way the scribe of electoral campaigns, which they help to document in great detail.
On the other hand, they have become an object of political debate: the interpretation of their figures is a source of confrontation between the candidates.
Everyone is trying to impose a reading favorable to their electoral interests.
Politicians are both more exposed to the results of the polls and more attentive to their variations, insofar as they inform them about their own destiny.
They naturally come to assume that their knowledge and interest in politics are shared by all citizens, which does not correspond to reality.
More generally, it should be remembered that the question of the impact of polls on elections is as old as the polls themselves.
After more than 80 years of reflection and research, there is no exhaustive, objective and unequivocal scientific report, in France or abroad, concerning the various effects of the publication of opinion polls on the behavior of voters.
This vacuum is the condition for the perpetuation of the polemics around the polls, which reappear in the same forms with each presidential campaign.
Should our political model evolve to conform to this "democracy of opinion"?
The growing place taken simultaneously by the polls and the presidential election in political life since 1965 reveals the structuring thread of the history of the Fifth Republic: the rise in power of the democracy of opinion, that is to say the attention paid by governments to public opinion formalized by polls and expressed by the media.
Read alsoDistrust is growing between the French and politicians
This development has gradually given rise to a tension: the horizontality of listening to public opinion's expectations comes into conflict with the verticality of the classic conception of presidential power, which must lead public opinion and not follow it. To resolve this contradiction, some propose to abolish the presidential election when others suggest liquidating the polls: none of these remedies seems beneficial. The solution remains to be invented. It goes through the institutionalization of democracy of opinion to avoid a solitary face-to-face between the Head of State and public opinion. Criticized on several counts, can the Great National Debate or the Citizens' Convention on the Climate be summary drafts?
The Rite of Opinion.
A history of the presidential election and polls, Frédéric Micheau, ed.
du Cerf, January 20, 2022, 312 p., €22 Cerf