Essay Politics

Doing Away With Presidential Elections

by Pierre Brunet & Arnaud Le Pillouer , 1 December 2011
translated by Susannah Dale
with the support of Chadoc

The election of the President of the French Republic by universal suffrage is said to represent an undeniable step forward for democracy, in that it enables the people to directly choose a leader and a type of politics. According to P. Brunet and A. Le Pillouer, however, the opposite is true: electing the president by universal suffrage destabilises our institutions and weakens political life.

For the first time in decades, comments on the upcoming presidential elections are not limited to traditional estimates of the respective chances of potential or declared candidates, but are also calling into question certain perverse effects of this procedure for appointing the head of state. Depending on the analyst, there is either regret that the system in its current form has the potential to produce a repeat of the 2002 elections when far-right leader Jean Marie Le Pen reached round two, a possibility that could affect both the left and the right, (referred to as “21st April part 2” in the French media); or else there is talk of disastrous consequences for the political health of the nation, and of a ‘personality’ contest in an era of mass media, professional communicators and glossy paper.

Nevertheless, it is clearly not enough to simply condemn these pernicious effects in order to challenge, in terms of constitutional theory, the very principle of electing the president by direct universal suffrage. Indeed, it is the root of the deficiencies that are currently affecting the Fifth Republic (personalization, loss of moral authority, conflicts of interest, public loss of confidence in institutions and political figures, and so on), which are all linked, either directly or indirectly, to the excessive concentration of power in the hands of the president. While the Revolution was criticized for replacing an absolute monarch with a despotic Assembly, and while, later on, the Third Republic was blamed for having replaced the Emperor with a House that also had absolute authority, the situation now seems to be the reverse: the omnipotent parliament of the Fourth Republic has been replaced with the president of the Fifth. It would seem that the time has also come to ask ourselves whether it might be appropriate to eliminate the basic cause of the problem: the election of the president by direct universal suffrage.

Is the Presidential Election Untouchable?

The idea might come as a surprise, given that the debate on this issue is quite simply non-existent in political circles, the media or even the academic world. It is tempting to dismiss the question out of hand, on the grounds that abolishing this election is one of those reforms that “will just never happen”. However, doing so would be to mistake a self-fulfilling prophecy for an argument, just as children think they can get out of a task by saying they “won’t be able to do it”. The other consequence of this kind of prediction is that is kills off any kind of reflection, thereby preventing people from envisaging any possible disadvantages to this type of election, and in return, any advantages to be gained from its abolition.

Rest assured, raising this issue does not mean demanding or even expecting that the upcoming electoral process should be suspended: the elections planned for 2012 will, of course, take place. However, it is extremely doubtful that institutional practice will then carry on as if nothing had happened, when the five-year term now drawing to a close will have been marked by the advent of what has been termed the ‘hyper-presidency’, or ‘omni-presidency’. Everyone will agree that the current French president’s personality has had a great deal to do with it. In fact, the hyper-presidency has been fostered by the institutions of the Fifth Republic in the form they have taken so far.

We should, nevertheless, acknowledge that a reform of the procedure for appointing the President of the Republic, as things currently stand, has very little chance of succeeding in the short term – not because of the popularity that the main political parties bring to the presidential election, but because this reform could, in any event, only be brought about by the winner of that very same election. This scenario is about as likely to happen as the Senate accepting a review of the Constitution that would see its own sphere of influence reduced. [1] Even so, we refuse to succumb to fatalism on this issue: history is full of setbacks, and there is nothing to say that the need for this reform will not be felt in the near future. When that day comes, the reasons – which we believe to be convincing – for revising the election of the president by universal suffrage will have to be explained: that is why this debate must already begin.

This method of appointing the head of state appears, nevertheless, to tick all the boxes: in general, the argument put forward is that this form of election is popular and that it would therefore be wrong to challenge it at a time when French people are turning away from politics. Sometimes an additional argument is given: it is modern, effective and democratic. That remains to be seen.


This form of election may well be popular, but in what sense exactly? There is deep-rooted ambiguity here, which, it must be acknowledged, doubtless makes the argument more effective [2] given that the obviousness of the word conceals two highly dubious theories that nonetheless have a formidable rhetorical effect when used in combination. The first theory is perhaps the most widespread: it maintains that, at a time when the French people are turning their backs on politics, it would be paradoxical to abolish the only election that continues to have a high turnout. This makes peculiar use of the concept of ‘popularity’, because it reduces it to the (practical) success of this form of election, with no concern for the (theoretical) support it generates. Yet there is nothing to say that the two are inevitably linked. This is not the most important thing, however. Obviously, as the French public feels passionately about presidential campaigns, there is thus a much lower level of abstention than in other elections, and therein lies the problem: by taking voter turnout as the only indication of the popularity of this form of election, we are disregarding the disappointment it causes, while the personalization and excessive drama to which this election inevitably leads feed the myth of the providential leader. And yet, on the one hand public disenchantment is always equal to the foolish expectations we are obliged to have in order to recover from previous disappointments. On the other hand, by focusing all the attention on the public, this contest hides the very real importance of other elections. What is more, the election of the president by direct universal suffrage is damaging to France’s political health, in that it leaves room for the idea that democracy can be reduced to the will of a single individual. We should not, therefore, be surprised that levels of abstention increase in all the other elections that, receiving less spectacular media coverage, seem less decisive in political terms. The election of the president is therefore only popular on the surface: underneath, it insidiously undermines the French people’s belief in their institutions.

The second theory consists in using the word ‘popular’ in the more common sense of the term to say that the French public are ‘very attached’ to the election of their president by direct universal suffrage, that it symbolises their capacity to act democratically, and that it would be wrong, if not inappropriate or, in any case, impossible, to suggest it should be abolished. This idea is deeply paradoxical: in a society that claims to be democratic, the only (legitimate) way of establishing that ‘attachment’ would be to call on the electorate to deliver its opinion. The inappropriateness thus lies in invoking the popularity of an institution as a way of refusing to hold a debate on its continuation. It would be irresponsible to object that the public already gave its opinion on the matter in 1962, because that would be tantamount to arguing that what is already done cannot be changed. Even more so, although this is no doubt the aim, it would mean hiding the extent to which the function of this election has undergone profound changes since 1962, particularly since the length of the presidential term was brought into line with that of the National Assembly members. And yet nobody can claim that the public wanted what this election has become: like any legislator, the public can, when required, undo what it has already done. Taking these changes into account, we maintain that questioning the relevance of appointing the president by direct universal suffrage is necessary today: firstly in the public sphere and then one day, if need be, in the polls. That would be the moment to measure its real popularity. In the absence of such a consultation, we are limited either to positing that popularity based on fairly limited personal experience (although always generalized to the extreme, based on vague prejudice), or else to trusting ‘opinion polls’ which they are in name only, given that they are restricted to asking the French people their opinion without their having had the chance to form one (assuming the question has never been put to them publicly).


The supposed modernity of this form of election is based on the idea (which was often put forward during the reform of the five-year term) that most contemporary democratic systems are organized around a very similar structure, which consists in allowing the electorate to appoint a leader at regular intervals, who has the freedom and time required to carry out the policies for which he or she was elected. This is the case in the United States, of course, as well as in parliamentary systems, in which legislative elections have become a personal contest between candidates for the position of prime minister – and which, in some respects, end up resembling France’s presidential elections.

However, if we look a little more closely at these great democracies, such similarities are minimal when compared with what distinguishes them from the French system: neither parliamentary systems (in which all of the prime minister’s legitimacy, and therefore power, is dependent on the majority that supports him in parliament) nor the American system (in which the president must continuously reach a compromise with Congress, which is elected separately and either dissenting or hostile by turns), nor even the systems in which the president is elected by direct universal suffrage (Austria, Finland, Portugal, [3] etc.) claim, as France does, to place all the democratic legitimacy in the hands of one man (or possibly woman) who is supposed to have enough authority to act (almost) alone. The French system therefore has the appearance of a modern regime – if this can be considered a virtue: it is the only one of all the great contemporary democracies to organize the induction of an individual by universal suffrage in order to grant that leader the maximum level of power. Every aspect of the 1958 system tending to deflect that logic of concentrated legitimacy has been deliberately brushed aside – particularly, once again, in the wake of the reform of the five-year term (even though the seed of this change was present from the very beginning, and had started to flourish well before the revision of October 2000). In any case, if this is what is meant by the ‘modernity’ of a system – selecting a guide to follow in all circumstances – then we would willingly plead in favour of a little archaism in French institutions.


This type of election is certainly effective, but to what end? If it is about organizing the parliament’s submission to the president (the latter being able to continually remind members that they are each the elected representative of a small constituency, whereas the president represents the whole of France), then that is the kind of effectiveness we could certainly do without if democracy in France is to improve – we shall return to this. Or does it perhaps simply mean that electing the president by direct universal suffrage leads to the bipolarization of political life, which itself helps to stabilize the government? In that case, the theory is more questionable still: the grouping of political forces into two opposing camps is in fact the result of the process of holding a majority vote in two rounds, which has long been the system used for legislative elections. There is also little doubt that this bipolarization becomes particularly acute during those elections, because it is vital for any political group wishing to obtain seats in parliament to form alliances with the major parties. Is it true to say that the presidential election, now linked to the legislative elections, at the very least creates a knock-on effect and guarantees the president a comfortable majority? That would be like saying the presidential election was needed without taking the trouble to prove that a stable majority was not possible without it. Nothing, therefore, shows that re-establishing a different procedure for appointing the president would lead us to a more varied multi-party system, and running that risk would mean invoking a spectre in order to make inaction legitimate. More fundamentally, however, we would also need to question the actual comparative advantages brought about nowadays by this form of bipolarization: does the mere desire to stabilize the government – the major issue at the start of the Fifth Republic – still justify sacrificing everything for it (the diversity of policy agendas, the quality of public debate, the culture of compromise).


Finally, we believe that the democratic nature of this procedure for appointing the president is a highly debatable platitude. It is based on the idea, widespread but entirely questionable, that not only can democracy be reduced to the fact of electing society’s leaders, but even that any election (by universal suffrage, of course) is by nature ‘democratic’. The democratic nature of a country would therefore by measured according to the number of its institutions that are directly chosen by the people. However, nothing could be further from the truth: a system in which each minister was elected by universal suffrage would be in no way democratic, but simply chaotic. It is our belief that, because of the way in which its president is elected, the system of the Fifth Republic has the exact opposite problem (the concentration, rather than the dilution of power – we shall return to this). However, the deliberately exaggerated example of ministers enables us to understand that considering any election by universal suffrage to be automatically democratic is in fact tantamount to mistaking the means for the end. If we agree to define democracy (modestly, but taking into account the etymology of the term) as the system in which the major political decisions (particularly laws) are, as far as possible, the result of the will of the majority of citizens, then elections can be seen for what they really are: a means of establishing democracy, not democracy itself. In other words, the question of the nature of democracy can only be raised in relation to the system in its entirety, and it is important to identify the best way of making that system democratic.

Under such circumstances, it is necessary to ask whether the fact of electing their president by direct universal suffrage allows citizens to choose the form of politics they wish to have implemented during the five-year term. Unfortunately, the answer has to be ‘no’, because of the serious confusion surrounding the actual purpose of that election.

Firstly, people would be hard-pushed to know whether the presidential election consists in appointing a candidate or choosing a political agenda. This, of course, is a defect that affects every election process: the voters speak out in favour of the candidate’s ideas as well as what they believe to be his or her capacity to implement them, and usually a vague combination of the two. The same phenomenon can be observed in legislative elections. In this case, the consequences of that defect are largely mitigated by the effect of numbers: the party’s agenda is the same everywhere, and therefore the degree to which the parliamentary candidates’ (different) personalities influence the outcome is greatly reduced, so that the result of the legislative elections gives a reflection (which is fairly accurate) of voters’ basic preferences. During presidential elections, the influence of the candidates’ personalities is, on the contrary, at its highest point, because each political agenda is linked to one candidate alone. Confusion over the purpose of the vote then reaches a peak. Under these circumstances, it is difficult to state (as is often done) that, in the presidential election, the French people vote for the policy they wish to see implemented during the five-year term.

Even if we agreed to overlook this first problem, others immediately arise.

For example, let us accept the theory that the presidential election consists merely in appointing a candidate, in this case the one who seems most apt to lead the country thanks to his or her persuasive skills, dynamism, intellectual and moral competence, etc. Nowadays, should we not raise the issue – particularly in view of certain recent examples (in the plural) – of the potential discrepancy between the candidates’ actual personality and their media ‘persona’, created under the guidance of communications experts who are increasingly numerous and influential? We have to be allowed to insist on the fact that this gap may come to light before the election – or not until afterwards: in our view, these two examples can be illustrated quite easily, without needing to go back too far in the history of the Fifth Republic. Admittedly, there is no doubt that voters are (mostly) aware of the artificial nature of a candidate’s media persona. Nevertheless, to return to constitutional theory, it has to be said that this issue is problematic for an election that continues to be presented as a ‘meeting’ between an individual and the nation.

Let us now consider the presidential election as a way for the citizens to choose a political agenda (the most favourable hypothesis, although certainly not the most plausible). What agenda is this, exactly? The candidate’s agenda, or the one put forward by the party that supports that candidate? Paradoxically, a presidential election candidate cannot have the exact same agenda as his or her party (for, in order to be elected, the candidate must show ‘personality’), nor one that is too different (for everyone knows that the candidate belongs to a party, and needs that party in order to obtain the parliamentary majority required to govern). And yet the divergence between those two agendas (which it would be wrong to condemn, since it is the mechanical result of the current system) is problematic, because it makes the electorate’s reason for voting even less certain.

In short, during a presidential election, we never really know for whom we are voting, or for what we are voting, or even if we are voting for someone or something.

The Presidential Election: a Source of Instability

Therefore, none of the reasons that are usually put forward to justify continuing with the current procedure for appointing the head of state is convincing. On the other hand, there are solid reasons that lead us to believe it would be appropriate to do away with it.

With his customary vision, Georges Vedel explained in 1964 that adopting this procedure for appointing the president, without making any other changes to the constitution, left the institutions halfway between a return to the pseudo-parliamentarianism of the Fourth Republic and an increased focus on a plebiscitary Consulate. [4] Today, we can see that the experience of the five-year term has definitively paved the way for the plebiscitary Consulate. However, the rot had already set in.

We can find all kinds of explanations for the president’s considerable influence over French political life, but the primary cause still remains the election of the president by direct universal suffrage, since, at best, it creates representational consistency between the president and party members or, at worst, a hierarchy to the benefit of the former, who can claim to be the sole representative of the entire nation. The architects of the 1848 constitution, opposed to the election of the president by direct universal suffrage, had well understood those who said that an elected president would declare himself to be the ‘true’ representative, thereby relegating members of parliament to their electoral constituencies. What is more, this is precisely what Charles de Gaulle did, describing legislative elections as “487 local contests” [5] and stating that ‘deep-rooted legitimacy’ did not come from parliament, which he considered to be a ‘multiple, uncertain and vague representation of the issues that divide the nation”, but rather from the president elected by universal suffrage, who alone was capable of uniting the nation.

From then on, everything was at stake. While de Gaulle’s successors themselves never expressed that same hierarchy of legitimacy, it is not because they had distanced themselves from it but simply that they did not feel the need. One initial vote, in 1965, was therefore enough to establish once and for all that this election would determine all others and that the ‘unity of the nation’ could only lie with a president elected by direct universal suffrage. We have now grown accustomed to telling ourselves that the unity of the nation still depends on that election, not because we have verified this but because we refuse to imagine its disappearance. In order to justify the fact that we refuse to envisage it, we convince ourselves that its disappearance would jeopardize the nation’s unity. We are trapped in a vicious circle.

The upshot of this is that, regardless of everything that constitutionalism teaches us (which is essentially a theory based on balance), our entire institutional system tends towards imbalance – a profound, structural imbalance to the sole benefit of the President of the Republic.

This manifests itself first of all in the very nature of the functions carried out by the head of state – whose basic peculiarity is their acute ambiguity. Whereas in traditional parliamentary systems, the president (or king) embodies the national consensus with regard to the rules of the constitutional game, leaving the government and its majority with the task of doing battle with the opposition in the political arena, the president of the Fifth Republic, due to the procedure by which he or she is appointed, is expected to fulfil both roles at the same time: that of moral authority (as the president of the French people, according to the ritual formula), and that of political authority (as the leader of the majority that supports the president’s actions). Who could fail to spot the advantages to be gained from this ambiguity? The president is a moral authority that holds political power, and a political authority that claims to have the intangibility of the moral authority! Certainly, the practices of the current French president and, we should remember, the transition to the five-year term, have clearly tipped the president in the direction of ‘party leader’. However, that ambiguity has not disappeared, and is showing no signs of disappearing, since the head of state can at any time hide behind his status of moral authority in order to avoid controversy or prepare for an election. We cannot discount the idea that the French people’s distrust of their institutions, and their political figures in general, might have something to do with the confusion that reigns at the top of the state with regard to these different roles.The ambiguity that arises from the election of the president by direct universal suffrage not only has a symbolic impact: it is also at the heart of another imbalance, this time institutional, which seriously affects the function of the head of government. We should remember the extent to which this function has been reduced to the minimum since the beginning of the Fifth Republic. As we know, the prime minister has little say in the nation’s politics (even though this is provided for in the Constitution), because the president is mostly in charge; the prime minister appears at best as the person who skilfully implements the president’s policies, and at worst as an easy scapegoat who is treated carefully until no longer needed. Admittedly, the experience of political cohabitation showed the limitations and boundaries that restricted the president. However, this configuration seems to be ruled out a priori by the institutionalization of the electoral calendar.

In this regard, it is important to clear up an ambiguity. It is often said that when the five-year term was adopted the calendar was ‘inverted’. This is not quite true, and it has been rightly pointed out that in the Fifth Republic the presidential election has never followed the legislative elections [6]. While the expression ‘inversion’ is incorrect, the fact remains that the calendar established by the reform of the five-year term has the aim of making the results of the presidential elections determine those of the legislative elections as mechanically as possible.

The purpose – avoiding political cohabitation by means of electoral mechanics, because experience shows that virtue is not enough – is perfectly legitimate. However, there is another side to this: in the context created by the presidential election, this subordination strengthens parliamentary members’ dependence on the president, which has been one of the characteristic features of the Fifth Republic ever since the beginning. This third imbalance is doubtless the most serious, and is also a result of the procedure for appointing the president.

To be sure, one could argue that this dependence can work both ways, or become a double-edged sword, because the president needs parliament in order to adopt the policies that he or she has helped create. However, this plays with appearances: the order set by the calendar creates a feeling of dependence while also leading people to interpret any difference of opinion between the president and members of parliament as a sign of the latter’s discontent, as they are expected to obey – indeed, that would seem to be their function – in such a way that any divergence that could potentially lead to a legislative bill is taken to be a shift in mood or, worse, a rebellion.

They could also be criticized for their poor attendance at debates, the primary reason for which could be the fact that they hold more than one position. Admittedly, combining multiple posts is perhaps not the best way to be influential, but has nobody ever questioned their reasons for holding several posts concurrently? Has nobody wondered to what extent parliament’s increasing weakness can justify – not morally but politically – that combination of posts?

The president, therefore, is not limited in any way, at least not by members of parliament.

The president is unlikely to encounter any limitations in the area of justice, either, despite what some would hope. The reasons for this are entirely different, and we should first establish what is actually meant by justice. If we are talking about magistrates in general, both public prosecutors and sitting judges, we fail to see how their judicial authority can restrict the president’s political power. Should we therefore focus on the Constitutional Council? In most democracies today, the constitutional court plays the role of an influential counter-power. However, it is doubtful that the Constitutional Council will manage to be up to this formidable task, given the difficulty with which constitutional justice takes root in France: we only need to look at the archaism of both its composition (particularly the absolutely astounding right of former presidents to have a seat for life) and the procedure for appointing its members, with regard to its new jurisdictional powers.

The French president encounters few limitations in the press, even though we like to think this constitutes – or ‘would constitute’, to be more precise – a counter-power. This is not to challenge the critical role played by the written press, which is sometimes very effective, nor to deny the freedom it enjoys. However, the fact remains that we can all attest to the media immunity bestowed upon the elected president, who has the astonishing privilege – incomprehensible to France’s neighbours – of being able to request journalists to be so kind as to ask questions that he (or she) would like to answer, and neither the president, nor any member of the presidential entourage, is ever forced to attend any press conference that is even the slightest bit heated.

It is therefore possible to measure the degree to which – far from guaranteeing institutional stability, which was its initial raison d’être – the presidential election has become a destabilizing force, given that the obstacles intended to oppose it can seem so fragile. There is no use talking of the powers that members of parliament ‘would only have to’ enforce in order to loosen the presidential grip. As Vedel said as early as 1964 [7], “it is taking a rather abstract form of refuge to argue that a text has every virtue apart from that of being applied”. We have to stop imagining that by granting new powers to other institutions we will put an end to their dependence on the president: it is that very dependence that must be attacked, and its ultimate cause: direct universal suffrage.

First published in Translated from French by Susannah Dale with the support of the ChADocs (chercheurs associés et doctorants du Collège de France).


by Pierre Brunet & Arnaud Le Pillouer, 1 December 2011

Further reading

Also in Books & Ideas:

 The Nature and Limits of the Majority Principle, by Pasquale Pasquino

To quote this article :

Pierre Brunet & Arnaud Le Pillouer, « Doing Away With Presidential Elections », Books and Ideas , 1 December 2011. ISSN : 2105-3030. URL :

Nota Bene:

If you want to discuss this essay further, you can send a proposal to the editorial team (redaction at We will get back to you as soon as possible.


[1This article was written well before 25 September 2011 and the Senate’s swing to the left. That event, which had seemed so unlikely until recently, would appear to confirm our theory on two counts. Firstly, because the Senate could well, in the end, see its own powers reduced: the left owes its majority to an unlikely combination of circumstances over the last few years, and will therefore no doubt be keen to reform this institution before another change can take place. Secondly, because if the left wins the 2012 elections it will find itself in a position to carry out unprecedented reforms of the institutions of the Fifth Republic. The simple fact that they are able to make reforms might well create a strong desire to do so… Who knows, perhaps long-standing Socialist reservations about the election of the president by universal suffrage might resurface.

[2It would appear that it is on account of this popularity that, in their work outlining their reform proposals for the Fifth Republic, Arnaud Montebourg and Bastien François, to name but two, failed to mention the reform of the process for electing the president – one they nevertheless deemed important by their own admission.

[3These countries (as well as Romania, Ireland, Serbia, Poland and even Russia) provide for the election of the president by direct universal suffrage. However, as Olivier Duhamel notes when he points out the singular nature of the French case in relation to those countries, “Everywhere else, everywhere, except in Putin’s Russia, the government obeys only the superior authority of the prime minister, and in no way that of the president. Everywhere else, the popular election of the president appoints only the holder of a representative public office that is, at most, influential but never has actual political power. Nowhere else does the presidential election challenge the government in place, nor the prime minister who leads it. In short, everywhere else the semi-presidential regime is just another parliamentary system like any other” (own translation of O. Duhamel: “Une démocratie à part”, Pouvoirs, 2008/3, n° 126, pp. 17-26, here p. 23).

[4Georges Vedel: “Vers le régime présidentiel?”, Revue française de science politique, 14th year, n°1, 1964. pp. 20-32, here p. 32.

[5De Gaulle, “Conférence de presse du 16 mai 1967”, in Discours et messages, vol. 5, Paris, Plon, 1970, p. 157.

[6Cf. Guy Carcassonne : “Immuable Ve République”, Pouvoirs, 2008/3, n° 126, pp. 27-35, here p. 29.

[7Ibid., p. 24. Own translation.

Our partners

© - Any replication forbidden without the explicit consent of the editors. - Mentions légales - webdesign : Abel Poucet