The 2017 presidential election and the nebula of abstentionists

Until now, the presidential election has remained a highly mobilizing election. It's the only election really capable of attracting large numbers of citizens to the polls, and bringing back some of those who abstained from other elections.
Céline Braconnier, University of Cergy-Pontoise and Jean-Yves Dormagen, University of Montpellier

As a result, abstention rates at legislative elections have doubled in the last twenty years, with each new election setting a new all-time record. If we take into account non-registered voters, far fewer than one in two citizens now takes part in choosing their MP. Less than one citizen in three participates in the European, regional and departmental elections. And municipal elections, despite the continuing popularity of mayors in surveys, have not escaped this demobilization: turnout has dropped by 15 points in 15 years, even if the stronger resistance of small towns conceals the extent of the disaffection recorded in large cities.
In this context, the presidential election is a real exception, demonstrating its ability to keep more than 8 out of 10 registered citizens voting. Certainly, on two occasions under the Ve In the French Republic, abstention was unusually high, at around 30%: in the second round in 1969, when there was no left-wing candidate (31% abstention). And in the first round of the 2002 elections (28% abstention rate), which the media portrayed as a duel that would inevitably lead to a run-off between the President and the outgoing Prime Minister, Jacques Chirac and Lionel Jospin. But since April 21st, the abstention rate has always been relatively contained: 20.3% for the second round in 2002, 16% for both rounds in 2007, and around 20% for both rounds in 2012.

Electoral intermittence and social determinants

Analyses carried out by Insee, in particular, based on cross-referencing electoral rolls with socio-demographic data, provide two important insights into the logic shaping contemporary abstention.
On the one hand, longitudinal approaches show that constant abstention affects no more than 10% of registered voters during an electoral sequence that includes a presidential ballot. In fact, electoral intermittence is the norm, rather than a definitive break with voting. The same citizens who increasingly abstain from intermediate elections have nevertheless continued to participate massively in the presidential election.
On the other hand, studies show that the traditional social determinants of abstention remain very strong: it is always the same socio-demographic categories that record the highest abstention rates - the youngest, the least qualified, the most economically fragile and the most isolated - it being understood that the electoral inequalities thus highlighted vary with the level of abstention. They are very marked when abstention is high - for example, in municipal and legislative elections - and more attenuated when abstention is low - for example, in presidential elections.
These two characteristics of abstention should help us to identify the challenges of participation in the 2017 presidential election, using the indicators available to us today to map out the space of possibilities.

More non-registered members

The data published by Insee on voter registration in 2016 supports the hypothesis that the first phase of the election was not very mobilizing: voluntary registration figures are down on those of the previous presidential election, let alone those of 2007: the "exit rate" from non-registration was just 12.4% in 2017, compared with 15.6% in 2012 and, even more so, 18.7% in 2007.
This lower mobilization for registration primarily affects fragile and isolated populations: this exit rate was only 11% among members of single-parent families in 2017, compared with 15% in 2012 and 21% in 2007. It was only 11% among blue-collar workers, compared with 14% and 17% in the two previous presidential elections. Executives, meanwhile, are still twice as likely as blue-collar workers to follow this trajectory of political integration. But like the other categories, they registered less for this year's presidential election: a 28% rate of non-registration this year, compared with 33% in both 2012 and 2007.
Not only is the proportion of non-registered voters higher for this year's presidential election than for previous ones , but the rate of re-registration after a move has also fallen, so the proportion of non-registered voters is likely to be even higher. In addition to the 6 million unregistered voters who will be unable to exercise their electoral rights this year, there will undoubtedly be around 7 million people registered in a commune other than the one in which they live. It has recently been established that they abstain three times as much as those registered next door: 28% constant abstention for the four rounds of voting in 2012 (presidential and legislative) among the poorly-registered, compared with less than 10% among the well-registered.

The abstention-sanction hypothesis

Once the possible extent of non-voting attributable to the registration procedure has been assessed, how can we account for the surplus outlined by a number of polls conducted between February and April 2017, which support the hypothesis of a historically high level of abstention, on a par with that recorded in 1969 and 2002 (around 30%), and which would in any case mark an increase of at least 10 points compared to the norm for a presidential election?
Most commentators put forward the theory of an upheaval in voting behaviour: a proportion of voters, who had previously voted, would choose not to go to the polls in this presidential election, thus practising a form of "abstention-sanction" that would call into question the political offering and, more broadly, our political system.
Before assessing the realism of this hypothesis, it's worth remembering that the record abstention figures reported in the media owe a great deal to the particular circumstances in which the surveys were carried out, and to the criteria used to define the category of "abstainers". This explains why they vary so much: from 24% to 40% abstention, depending on the time and the polling institute!

Fragile declarations of intent to participate

It is certainly risky to anticipate the level of participation in such an election. The very nature of a presidential campaign is that it can escalate in intensity right up to election day - not least because of the mechanisms that encourage participation within families, groups of friends and colleagues, but also because of the candidates' campaign strategies and tactics, and the increasing dramatization of the issues by the media.
All the more so as polls - and this is the second point - are not very well suited to measuring abstention. Samples always include far too many voters, and declarations of intention to participate are even more fragile than declarations of intention to vote. Experience shows that asking potential voters whether they will vote is not enough to measure future abstention levels.
The "abstentionists" in polls are not citizens who declare themselves as such, but - more often than not - the aggregate of those who do not say they are sure they will vote at the time they are questioned.
In other words, it's difficult to predict the abstention rate next Sunday. Will it be 20%, 25%, 30%, 35% or more? We'll only know... on polling day.

The risk of an explosion in electoral inequalities

On the other hand, it is already possible to anticipate the sociological effects of a high abstention rate. The chances of people finally turning out to vote differ widely, and depend not only on the electoral offer, but also on their age, their level of education, the professional environment in which they evolve, their economic affluence and the composition of their household.
In 2012, unskilled workers abstained twice as much as skilled workers, and blue-collar workers as a whole twice as much as senior executives. The least qualified abstained three times more than the most qualified. Voter turnout among 18-25 year-olds was 10 points lower than among 50-64 year-olds. And as the abstention rate rises, these gaps widen considerably. Two months later, in the legislative elections, the gap in turnout between 18-25 year-olds and 50-64 year-olds reached 30 points.
As a result, if abstention were to approach or even exceed 30% in the first round of the presidential election, the inequalities in voter turnout, usually reduced for this ballot, would undoubtedly return to a level equivalent to that which has become their norm for intermediate elections.

The impact of business and left-right blurring

As far as political factors are concerned, it's clear that the 2017 presidential election is not conducive to the crystallization of choices. The level of uncertainty in the declared potential electorate of the main candidates, with just one week to go before the vote, is unprecedented in its scale, and obviously feeds the hypothesis of disoriented citizens who may not turn out, for want of finding an offer that matches their vote or even a slightly stable reference point for making their choice.
The prominence given to cases questioning the probity of certain official candidates probably only reinforces the already widespread feeling that the political world is inward-looking, unconcerned with the difficulties that ordinary citizens encounter on a daily basis, and exonerates itself from respect for common rules. By fuelling a poisonous climate, these affairs undoubtedly increase the risk of abstention, but not necessarily primarily or only among the electorate of the personalities most frequently implicated.
The over-representation of young retirees in his electorate - an age category that votes (almost) systematically - certainly protects François Fillon from a real risk of abstention (but not from the risk of some of his potential voters fleeing to Marine Le Pen, Emmanuel Macron or the blank vote). On the other hand, the suspicions that affect all contenders for elective office - "they're all the same, the others just haven't been caught yet" - could well hinder the mobilization of the least captive voters, i.e. first and foremost young people - whatever their electoral orientation - from the most working-class categories and the least qualified.
Let's not forget either that the increase in turnout recorded so far for the presidential elections is largely due to the mobilization of citizens who are the most distant from institutional politics, who may vote out of duty, habit, social pressure or conformism, but are not necessarily driven by a strong conviction or interest in politics and the issues at stake. For many of them, the left-right divide in particular still functions as a framework organizing a minimal understanding of candidacies. The blurring of these reference points, by making voting more complex and costly, can also discourage participation.

The myth of passionate abstentionists

Are these political factors really enough to make abstention a new form of expression that carries "messages", as we've been hearing on the airwaves, social networks and television screens for the past few weeks? Obviously not. Indifference, disenchantment and skepticism have for several years now been ingredients in the relationship with politics of many citizens, the vast majority of whom have nevertheless continued to vote, at least for the presidential elections. Disillusioned voters, they don't turn into passionate abstentionists at intermediate elections.
In fact, abstentionists make up a particularly heterogeneous conglomerate, including both highly politicized citizens and individuals who are totally indifferent to politics. The fact remains, however, that if abstention were to rise significantly next Sunday, it would probably still be fuelled mainly by the least politicized segments of the population - young people, blue-collar workers, the uneducated and so on. - which the particular context of this election could tip over into a sort of electoral exit.
To focus solely on the "political" dimension of abstention, and interpret it as a means of expression equivalent to voting, would be to overlook a fundamental fact: our democratic order is based on growing inequalities in politicization and political participation, which are no less worrying than the social and economic inequalities they encourage to reproduce.


The ConversationCéline Braconnier and Jean-Yves Dormagen have co-authored La démocratie de l'abstention (Folio Actuel).
Céline Braconnier and Jean-Yves Dormagen are Professors at Faculty of Law and Political Science theUniversity of Montpellier
Visit original version of this article was published on The Conversation.