By Prof. Laurence Morel, University of Lille and SciencesPo Paris (CEVIPOF)

After its intense use by the general de Gaulle at the beginning of the fifth Republic, the referendum has been an on and off issue in France. Very much like in other representative democracies, both parties of the Left and the Right were quite hostile to the device, but this was increased in France as a result of its very controversial use by de Gaulle, which was reminiscent of bonapartist methods. Such an elite consensus against the referendum was not vey different however from what can be seen in most democracies (Svensson, 2017). Thus, the only occurrence of the device has been at the initiative of the President of the Republic, when, for essentially opportunistic reasons, he decided to call a referendum under Article 11 of the Constitution, which allows him to submit a bill bypassing the parliament, or, once, under Article 89, as part of the constitutional revision (where it was optional). This happened once with Pompidou, who actually initiated the season of EU related referendums, asking the French whether they agreed that the UK, Denmark, Ireland and Norway join the Community (1972). This did not happen with Giscard, perhaps because of a personal reluctance to “appeal to the people”, but was again the case with Mitterrand in 1988, although in practice the referendum on New Caledony had been a pledge of the prime minister to the local separatist leader. The referendum on the Maastricht treaty (1992) was the only truly « Mitterrandien » referendum – maybe also the last « Gaullist style » – and the second EU related referendum in France. The third one, on the European constitutionnal treaty (2005), which came under Chirac presidency after a first referendum on presidential duration in office, was a turning point in the French experience : for the first time, the claim for referendum came from a large array of parties, personnalities, both from the left and the right, as well as from the civil society. Actually Chirac hesitated a lot to hold this referendum and was to a large extent constrained by the pressure exerted on him (Morel, 2007). Thus 2005 clearly demonstrated that the referendum was not anymore a presidential weapon under the Fifth Republic, as conceived and practiced by de Gaulle, and that it was no longer a taboo for parties, which were now spurred on by growing demands from the society.

A. Historical context

It is impossible to understand the current situation of the referendum in France without retracing its history, which goes back to the revolution and the debates of the Constituent Assembly, where it was claimed and theorized by the radical Left. Although apparently promised a bright future, in the name of sacrosanct popular sovereignty, what prevailed at the end is the fear of the people and the first French Constitution (1791) did not mention it. Admittedly the second Constitution, two years later, contained generous mechanisms of direct democracy, but it remained on the paper because of exceptional circumstances. In the end, the revolution only introduced the mandatory constitutional referendum, in accordance with the decree adopted by the Convention at its first meeting that “there can be no constitution except that approved by the people”. Indeed both the 1793 and 1795 Constitutions were ratifyed by universal male suffrage at the first two referendums of French constitutional history. The massive victory and very low turnout (26% in 1793 and 16% in 1795) show that these referendums were far from meeting the democratic requirements; but their object was not yet deviated as it would soon be under Bonapartism, where their use was pivotal to the regime. Actually the Bonapartist referendums, while continuing the revolutionary tradition of constitutional referendums, had the particularity of systematically overlapping with questions of persons. Thus, Napoleon asked the people to appoint him Consul for life (1802), and later to approve the hereditary Empire of the Bonaparte family (1804) – a question that was again the subject of a referendum during the Second Empire (1852), after Louis-Napoleon asked the French to grant him special constitutional powers following his “coup d’etat” (1851). But in practice all seven Napoleonic referendums have mixed, implicitly when not explicitly, constitutional questions and approval of the leader. This enabled them to constitute a powerful mechanism for strengthening the Executive and circumventing the Parliament, which was never involved in the elaboration of the Constitutions, as the revolutionnary assemblies had been, and was sometimes downright hostile to the proposed revisions.

While Napoleon I held ad hoc referendums, Napoleon III, it should be noted, institutionalized their use by introducing the mandatory constitutional referendum on all revisions in the Constitutions of 1852 and 1870, and, in the latter, the “appel au peuple” (Article 13: “The Emperor is responsible to the French people, to whom he is always entitled to appeal”). The fall of the Empire did not allow to experience the process, but Napoleon III had enshrined in the Constitution the referendum of confidence invented by his uncle, making it a referendum of responsibility of the Head of the state implying his resignation in case of failure. So doing he anticipated the Gaullist conception of the referendum a century later as an instrument for popular ratification of major decisions and checking the people’s confidence in the Head of State. With the exception of the last (1870), which benefited from the relative liberalization of the regime, no need to say that Napoleonic referendums were far from meeting democratic standards, as evidenced by their massive approval (over 90% and slightly below 83% in 1870). Compared with the revolutionary referendums they succeeded better however in mobilising people (around 50% under Napoleon I and 80% under Napoleon III), perhaps because of their strong personal dimension.

The referendum was of course absent from both constitutional texts and practice during the restoration of monarchy (1815-1848). But its rejection persisted during the short-lived Second Republic (1848-1851) and was confirmed under the Third Republic. It is surprising that direct legislation was so absent in 1848, after a a revolution made in the name of universal suffrage. The Constitution of 1848 did not actually provide for any form of referendum, nor was it submitted to the people. The new elected assembly was in fact haunted by the twofold fear of social revolution and “Caesarism” (the word appears under Napoleon III), and a consensus emerged in favor of strong representative institutions and against the attribution to the president of the power to appeal to the people – a temptation that might be all the stronger since he would be elected by the people. His direct election, which was part of the logic of building strong institutions, had been very controversial. His direct election, which was part of the logic of building strong institutions, had been actually highly controversial.

The Third Republic, which was born in the aftermath of the Second Empire and the revolutionnary events of the Commune, was even more marked than the Second Republic by the double specter of Caesarism and social revolution, uniting monarchists and Republicans in a common rejection of the referendum. The major fact of the period was the definitive rupture of the Left with the direct legislation. The Republicans, first, took a firm stand in favour of pure representative government. The Socialists, for their part, focused now on the leading role of the party as a privileged instrument of expression of popular will. Like under the Second Republic, the referendum was therefore absent from both the constitutional texts and the practice of the first stable regime since the Revolution. A proposal to submit to popular ratification the Constitution was rejected in January 1875. The almost exclusive defense of the referendum by Boulangists and Bonapartists until First World War was also not very conducive to the rehabilitation of the process.

Most lawyers, it should be added, viewed the referendum as not compatible with representative government (e.g. Esmein, 1896). Only in the inter-war period was there a certain rehabilitation of the referendum as a means of mitigating the power deemed excessive of the Parliament, such as the defense of popular initiative by Carré de Malberg (1931). This debate was not unique to France. Much earlier, the British lawyer Dicey had denounced “absolute parliamentarism” and the “dictatorship of the parties” in his country and demanded the introduction of the mandatory referendum on constitutional and sovereignty issues (Dicey, 1890). As in France, this plea in favour of the referendum was unsuccessful, but the new inter-war European Constitutions made a large place to the referendum (Mirkine-Guetzevich, 1931), which meanwhile continued to develop in Switzerland and in the United States under the impulse of populist movements. In some countries the referendum was also thought as a means to strengthen the Executive vis-à-vis the parliament, such as in Germany under the influence of Max Weber. In some similar way in France, former right-wing Prime Minister André Tardieu, who is known to have had a great influence on General de Gaulle (Passelecq, 1990), proposed in France a “State Reform” intended to reduce the power of parliament and which included the referendum initiated by the President of the Republic at the request of the government.

It is de Gaulle, as chief of the transitory government, who reintroduced the referendum at the Liberation. A referendum was convened on October 21, 1945, the same day as elections, asking the French whether the Assembly elected on that day should draft a new Constitution, in other words whether they wanted or not to return to the Third Republic. A second question asked whether the Assembly, in case the Yes would win at the first question, should have limited duration and powers, and provided that the Constitution would be submitted to a referendum. Officially, these referendums, as well as the two which were held the next year on the new Constitution, were meant to allow the people to choose their regime and grant the institutions stroner legitimacy. But tactical motivations were not absent. By circumventing the all-parties Consultative Assembly set up at the Liberation, de Gaulle hoped to foster his institutional choices in favor of a new Constitution with a strong Executive. Parties were actually divided on whether or not to return to the Third Republic, even though at the end only the Radical party – which was emblematic of the Third Republic – advocated the No to the first question. But above all parties were massively in favor of parliamentarism. Thus the referendums of the Liberation were used by de Gaulle as a sort of minority weapon against an Assembly which was largely hostile to their holding, for reasons of both principle – the hostility to the process – and politics – the hostility to the regime wanted by de Gaulle. Another concern was that these referendums were also meant to grant de Gaulle the undisputed legitimacy that he was still lacking in the transitional process of the Liberation.

Actually, the referendums convened by de Gaulle, although reviving the revolutionary tradition of constitutional referendums, also relaunched the Bonapartist use of the process aimed at circumventing the parliament and legitimizing their author – albeit under much more liberal conditions. The result was very mixed since de Gaulle won the referendums but the assembly elected the same day was dominated by the Communists and the Socialists, which were both supporters of strong parliamentarism; and the personal legitimacy acquired through the referendum finally proved to be of little weight compared to that of this elected Assembly, that nothing will stop on the road of a “régime d’assemblée”, causing de Gaulle to resign in January 1946 almost in the general indifference. Although the Constitution was rejected by the people in May, the second version, which only watered down the first draft, was adopted in October. The only “victory” of De Gaulle was the poor result (53% of Yes representing only 35% of registered voters), which was unfit to give a real legitimacy to the Constitution.

Although its authors claimed the legacy of the Constitution of 1793, the Constitution of the Fourth Republic gave a very limited role to the referendum: it was mandatory in constitutional matters only under certain conditions and explicitly prohibited in legislative matters (Article 3). Indeed, there was no referendum under this Constitution. It was not until the 1958 Constitution, which was established this time under the strict control of de Gaulle, that a French Constitution granted an important place to the referendum (with the exclusion of the never-applied Constitutions of 1793 and 1870). Article 3 now extended it to ordinary legislation, and Article 11 allowed the President of the Republic to submit to the people any draft law “on the organization of public authorities, entailing the approval of a Community agreement, or for the purpose of authorizing the ratification of a treaty which, without being contrary to the Constitution, might affect the functioning of the institutions “. Furthermore, article 89 granted the President the power to decide whether a constitutional amendment initiated by him would be submitted to the referendum or to the Chambers, after it was approved in a first reading – the referendum being mandatory for revisions initiated by parliament (see section on legal regulations for more details). In short, the French President could now completely bypass the Parliament on legislative matters and partially on constitutional matters. Article 11 contributed, alongside other mechanisms such as the early dissolution and, from 1962, the direct election , to forge the direct relationship between the President and the people which was the essence of the regime wanted by de Gaulle.

Until 1962, however, the election of the President was still indirect and the four referendums held by de Gaulle – the last of which actually to approve the change in the election of the President – actually acted as a substitute for the direct election (Morel, 2010). Each time de Gaulle explicitly asked for confidence and threatened to withdraw in case of a No victory. In 1958, his personal capture of the referendum aimed at verifying that he had the confidence of the Nation in spite of the controversial circumstances of his access in power – although the referendum, which was mandatory according to the law of 3 June 1958 delegating the government the drafting of a new Constitution, had also been demanded by de Gaulle to bypass the Parliament of the Fourth Republic (which remained very attached to pure parliamentarism). Thus in the end, de Gaulle’s motivations resembled much those of 1945. In the following three referendums, in January 1961 and April 1962 on Algeria and October 1962 on the direct election of the President, de Gaulle wanted rather to assert his legitimacy to govern in the first place, while Article 20 of the 1958 Constitution attributes this role to the prime minister – in other words he wanted the French to approve the de facto “presidentialization” of the regime. Admittedly, a primary goal of these referendums was also, here again, to serve as a minority weapon for the President, whose party (the Union pour la Nouvelle République) was first in Parliament but lacked the absolute majority that could ensure the approval of his Algerian policy, and even more the direct election of the President, which was unanimously opposed by parties.

After 1962, the direct election will make the referendum less functional to the regime. It will follow a decline of its practice. The President now draws his legitimacy from this election. Nevertheless, the referendum still serve to “relegitimize” him in the course of a particularly long mandate, similar to a sort of mid-term election (Parodi, 1973). This was clearly the case of the last referendum called by de Gaulle one year after the crisis of May 1968. This referendum was sometimes interpreted as a political suicide, the overwhelming victory of the Gaullist party at the early election of June 1968 having demonstrated the confidence of the nation. But de Gaulle was anxious that this confidence might have concerned more his prime minister than himself. Perhaps this fear was not totally unjustified since the French this time followed party stands and voted No, causing de Gaulle’s resignation and definitive withdrawal from political life. Similarly, the referendums convened by Pompidou in 1972 and Mitterrand in 1992 (both on Europe) can be interpreted as attempts to revitalize their position after a few years in office – although a secundary function of these referendums was also to divide the opposition (as will be seen in section B).

The advent of bipolarization and presidential majorities after 1962 will also make less necessary the minority use of the referendum by the President. Here the referendum contributed to its own loss since this process started with the division of the French in two camps at referendums and was cristallyzed through the second round of the direct election – which it introduced. But most of all, after 1962, this presidential weapon also becomes less effective: all referendums have a boomerang effect on their initiators either because of their negative result (1969 and 2005), the too short victory (1992), or low turnout (1972, 1988, 2000). Actually, referendums start to be less controlled, as a result of the democratization of the regime, leading to greater unpredictability of both their outcome and turnout. This is also the consequence of a certain “de-plebiscitarization” of the referendum after de Gaulle, his successors carefully avoiding to engage their responsibility. This played both against participation – which suffered from the absence of personnalization – and against the Yes – a No victory being no longer considered dangerous for stability, which encourages the protest vote.

The alignment of the presidential and parliamentary elections following the reduction to five years of the presidential term and the trauma provoked by the access of the Front National candidate at the second round of the 2002 presidential election made the referendum even less functional to the regime. The shortening of the presidential duration in office actually pursued what the 1962 reform had begun, namely the decline of the legitimizing function of the referendum. In parallel, the minority function of the referendum also lost its purpose, as a result of the further presidentialization of the majority, due to the parliamentary election being now held in the wake of the presidential election – something which sometimes happened in the past but is likely to become systematic. The 2017 election provided a masterful illustration of this process with Macron managing to build his majority the month after his election on the basis of the new party created by him for the occasion. Finally, the risk of failure of the referendum is now increased with the President rising in the front line in place of the prime minister, which is also a result of the presidentialization of the majority. This is likely to ruin the effort to “deplebiscitarize” the referendum and to increase its risk of failure, all the more since presidential popularity has become very volatile. For all these reasons, the future of the presidential referendum seems quite compromised in France, as confirmed by the absence of any new referendum since the last, failed referendum of 2005. Nonetheless, as will be seen in section C, the referendum is the subject of a renewed debate in today’s France, even if this interest in the process does not necessarily concern the presidential referendum.

B. EU related referendums

Table 1: EU-related referendums in France since 1958

Date & Issue Electorate Turnout
(including invalid vote)
Yes: N Yes: % cast vote No: N No: % cast vote
UK, Denmark, Ireland and Norway join EC
(Apr 23, 1972)
29 820 464 60.2 10 847 554 68.3 5 030 934 31.7
Maastricht treaty
(Sept 20, 1992)
38 305 534 69.7 13 162 992 51 12 623 562 49
European Constitutional treaty
(May 29, 2005)
41 789 202 69.4 12 808 270 45.3 15 449 508 54.7



«Do you approve, given the new perspectives opening for Europe, the Bill submitted to the French people by the President of the Republic authorizing the ratification of the treaty relative to the accession of Great Britain, Denmark, Ireland and Norway to the European Communities?»

This first post-Gaullist referendum, initiated by Pompidou, was also the first use of Article 11 to concern a treaty. Actually the constitutionality of this referendum could have been questioned since the membership of new countries did not affect the functioning of French institutions as article 11 states a treaty should do to be eligible to a referendum (Berlia, 1972). Europe was not yet an issue for most French people, and it would take long before it was. Admittedly, de Gaulle had been an opponent to UK membership, but this did not create any obligation to hold a referendum. But Pompidou had found an issue that was expected to ensure him a massive victory, since only the Communists were opposed to enlargment. Although he insisted in his declaration sent to the electors that he was not asking for an approval of his policies, but simply for a confirmation of a crucial aspect of his foreign policy, and officially justifyied his decision by the fact the French would be very affected by the entry of Britain in the EEC (press conference, March 16th 1972), Pompidou clearly hoped to reinforce his position indirectly through this referendum.

This did not prove a very successful strategy since abstention (39,8%) and invalid votes (7,1%) reached the highest level ever, roughly double the normal rate (Leleu, 1976). This was the combined result of lack of personalization and dramatization, the foreknowledge of the answer, discontent with the government and, most of all, indifference regarding Europe. Thus one month before the vote 72% of the polled said they were not or little interested in the referendum (IFOP, 25-29 March, 1972), while 22% declared that they did not clearly perceived the consequences of the question asked and another 27% that the entry of new members would have no consequence (IFOP, April 1972). Moreover only 51% approved the holding of the referendum (IFOP, March 21, 1972) (IFOP polls published in journal Sondages).

But the very low turnout also reflected the abstention of half of the Socialist electors, whose party, although approving the enlargment, had recommended abstention as to avoid jeopardizing the on-going process of alliance with the Communists – this was followed by a joint statement of the PSU (parti socialiste unifié) and other eight left-wing non Communist organizations in favour of abstention . Actually Pompidou’s also hoped the referendum would divide the Left and help him pursuing his tactic of extending the Gaullist majority to the Center parties (Centre démocrate, Centre républicain, Progrès et démocratie moderne and the Radical party), which were the most pro-European among French parties (Criddle, 1972 ; Morel, 1996). Here again he failed, since the Socialistsand Communists signed their « Programme Commun de Gouvernement » only a few months after the referendum. Moreover the Left substantially increased its score at the parliamentary election a year later.

The massive abstention of Socialists also made the victory much smaller than expected, voting intentions for the Yes falling from 89% on March 22, shortly after the announcement of the referendum by the President (March 16th ) to an actual 68,3% in the ballot (IFOP). The defection of part of the centrist electorate (Criddle, 1972 :247) also responds for this drop and was a blow to Pompidou’s strategy of alliances – the centrist leaders making clear throughout the campaign that their Yes was in no way a Yes to the government. Thus the President came out weakened from this referendum, not the least on the European scene, where he failed to appear with the enhanced authority he had hoped to gain at the Paris European summit scheduled six months later, with the result that he had to renounce his claim to fix the agenda at this summit (Leigh, 1975 :167). Yet, as with the Gaullist referendums, the campaign had been dominated by the government, which still controlled fully the audiovisual. Only the press expressed more balanced views, but the real protagonists were now the radio and the television (Morel, 1996:78). The tutelage on audiovisual will be suppressed progressively from 1974 on the initiative of Giscard, but for the time being the journalists campaigned openly on the government side.

Until 1969 disloyalty to party recommendations at referendums could not really be interpreted in terms of an autonomous voter since it mostly reflected loyalty to de Gaulle against the parties (Morel, 1996). Thus the bulk of it consisted of electors of the parties advocating the No chosing instead to vote Yes to de Gaulle (admittedly this was less true in 1969 where half of the Centrist electors voted No against their party recommendation). “True” disloyalty really starts in 1972, though in a rather modest way (Criddle, 1972:247; Leleu, 1976 :32). It consisted mostly of electors choosing abstention rather than following their party stand. Thus 35% of the Communist electors abstained but only 5% voted Yes (the PCF was for the No); while 33% of the Centrist electors and 29% of those of the UDR and the small RI (“Républicains indépendants”) abstained and respectively 6% and 3% chose to vote No against their party official endorsement of the Yes. Only Socialist and PSU electors proved to some extent more disloyal with half of them going to the polls while their party recommended abstention. But in practice 27% voted Yes, in line with the actual position of the Socialist party (see above). Admittedly the 23% who voted No represented a contingent of “true” disloyal voters larger than for any other party (IFOP).

The Left-Right cleavage was reflected in the territorial distribution of the vote, the Yes been stronger in the Right-oriented departments and particularly low in the Left-oriented ones (Leleu,1976:36). Thus it was highest in the West (Bretagne) , in the East (Alsace-Lorraine) and in the Southern part of the Centre, which were strongholds of the Right at the time. And it was the lowest where the Left was particularly strong: in the Paris basin, in the Center and on the Mediterranean shore. Notwithstanding the call for abstention of certain parties, the pattern of non participation was also very usual, with the North going much more to the polls than the South, showing that abstention stemmed more from indifference than from party recommendations. And not surprisingly, the Yes was massive among social groups traditionnally the most favourable to the Right, that is, peasants (90%) and higher tertiary employees (“cadres supérieurs”) and liberal professions (85%). Turnout was also higher in these categories. Thus « the electoral behaviour of the French, despite the extent of non-participation which was largely due to the indifference of the elctorate towards the European question, remained characteristic of traditional patterns » (Leleu, 1976).

Indifference to the issue caused abstention but was also not favourable to issue voting. Thus, although Pompidou had carefully avoided to engage his responsibility (see above), breaking with the Gaullist tradition, voting intentions were mixed: only 52% declared they would pronounce on EC enlargment, while 13% said they would vote on government policy and 11% on the President of the Republic (IFOP poll, 25-29 March 1972). As a reminder, in 1969 60% of Yes voters wanted to support de Gaulle and 46% to avoid a crisis, while 32% of No voters expressed their opposition to de Gaulle and 47% voted for change (multiple answers; SOFRES poll quoted in weekly newspaper Le Nouvel Observateur, 21/04/1969). Beside the issue, no doubt that the “deviated” vote at the 1972 referendum was also the consequence of the personal initiative, which prevents to completely eliminate the plebiscitary bias of the French presidential referendum (Parodi, 1973).


«Do you approve the Bill submitted to the French people by the President of the Republic authorizing the ratification of the treaty relative to the European Union?»

Although the 1972 referendum may have given the impression that France was particularly willing to hold referendums on European treaties – it was the only country to vote on accession treaties of new countries – the Single Act treaty was not submitted to popular ratification. Thus Mitterrand’s decision in June 1992 to call for a referendum on the Maastricht treaty under article 11 – which was this time constitutionnally irreprochable – came as a surprise. The motives of the president were threefold (Portelli, 1992:5). First, he wanted to avoid the ratification process by national Parliaments to be stopped after the Danish No on June 2nd. Actually, Mitterrand announced the referendum the day after the Danish referendum. Europe was the main axis of his second mandate and the abandonment of the treaty would have meant a political failure for him.This introduces Mitterrand’s second motive: in the tradition of French Presidents, there is little doubt that the referendum was intended to boost his declining populary at mid-mandate, through popular approval of a treaty he co-authored with the German Chancellor Helmut Kohl. Beyond short-term effects, the President wanted to build his image of European founder for posterity, after having renounced to personify the break with capitalism (Grunberg, 2005:130).

Finally, a third motive was party oriented, in the view of the forthcoming parliamentary election (March 1993). The Socialist party had scored dramatically low at the March regional elections and could have benefited from the strong divisive potential of the referendum in the Right. The Gaullist RPR was actually split on the issue – the bulk of its elected officials and party activists being in favour of the No in contrast with the leadership – and the Giscardian europhile UDF was also experiencing internal dissidences. While the Single European Act had been a non issue, the sleeping consensus of government parties on Europe was actually reaching its end with the growing hostility of peasants to the common agricultural policy. Moreover, the European currency was raising doubts, all the more since the Danes were seemingly reluctant to approve it and the British had managed an opt out. In this context, the Constitutional Court’s decision in April declaring unconstitutional the provisions of the treaty allowing resident European citizens to vote in local elections and the transfer of monetary powers to Brussels fueled party divisions. These became particularly salient during the constitutional review process which immediately followed in order to adapt the Constitution. Finally, the rejection of the treaty by the Danes ( the first country to vote) could only further intensify internal divisions of Right-wing parties, especially in the context of a referendum.

The referendum proved uneffective but for its first aim, countries continuying to ratify the treaty as expected by Mitterrand in spite of the Danish No (Ireland on June 18th, Luxemburg, Belgium and Greece in July). Admittedly, it also succeeded in shaping Mitterrand as a great man of the European construction. But the too short victory of the Yes was unable to produce the intended authority-enhancing effect. In spite of the overwhelming superiority of the pro-Maastricht camp in the official campaign (see D/ on legal regulations and also Criddle, 1993:231-2 and Appleton, 1992:7), the Yes actually fell from 65% in voting intentions in early June to 51% in the ballot (Portelli, 1992: 9).

This was due to several factors. The first was Mitterrand’s unpopularity and the inconspicuous campaign of Giscard d’Estaing and Chirac, who did not want to offer a victory to Mitterrand and whose parties were divided. The RPR even went so far as to share its speaking time on public waves with dissidents (which the PS and the UDF did not do). Second, while the campaign of opponents to the treaty raged during the summer and mobilized charismatic tenors such as Philippe de Villiers, Philippe Séguin, Charles Pasqua or Jean-Marie Le Pen on the Right, and Jean-Pierre Chevènement on the Left, the pro-Maastricht Establishment was on holiday. Finally, the Yes side had obvious difficulties to explain its support for the treaty, while its opponents just relied on the wave of discontent (Criddle, 1993:232; Portelli,1992 :8; Appleton, 1992:7-8). Without the in extremis mobilization of the Yes after the holidays – with Giscard this time not sparing his efforts – and Mitterand’s performance against Seguin in a debate at the Sorbonne in September, the No, which was leading in late August, would have probably prevailed.

The referendum was also unable to seriously damage the Right. Although it exacerbated divisions within Right-wing parties, Chirac was re-endorsed by 95% at the RPR National Council soon after the referendum, and the alliance with the UDF overwhelmingly won at the next parliamentary election. Nor was the referendum able to rescue the Socialist party, which failed to improve its poor score in the previous March regional election and ended up with half as many votes as in 1988. Ironically, the main impact of the referendum was maybe to split the Socialist party, with the breaking away of the anti-European current led by Chevènement.

While twenty years earlier the referendum had not transgressed the Left-Right cleavage (see above), the Maastricht referendum brought on the scene the European cleavage as we know it today, with on one side pro-European government parties and on the other side the Extreme-Right and the Extreme Left campaining for the No, albeit on different grounds – a cleavage which clearly appeared as opposing the Establishment, also supported by the Church and the business community, and the excluded and the peripheral. Given the diversity of the No side, the cleavage was along two lines: the Communists, but also part of the Ecologists and dissident Socialists, denounced the “mercantile and liberal Europe”, while the Front national and the RPR anti-Maastricht leaders focused their attacks on the infringment of state sovereignty. Party loyalty was massive among Communist and Frontist electors (respectively 81% and 92% followed the party line), and was lowest among RPR electors (41%) and the Ecologists. But these two parties had left freedom of vote, thus the most significative disloyalty was maybe that of the 22% and 39% electors of the PS and the UDF who voted No against party recommendation – though it should also be relativized since these electors might simply have followed the dissident party leaders (BVA exit poll published in Libération 22/09/92).

The rise of euroscepticism was well reflected in the cartography of the vote, with now more than half of France (13 regions out of 22) voting No, while the Yes had won everywhere in 1972. Nevertheless, the Yes was still highest (over 55%) in the traditional conservative strongholds, especially in Brittany and Alsace, and in Paris, which had been more mitigated in 1972. Similarly, the No was dominant in Left-wing areas and highest (over 55%) in Center France, the Mediterranean coast (where it also came from electors of the Front national), and this time also the North. From a sociological point of view, it particularly affected rural areas and old industrial regions, where it reflected the anxiety of economically fragile and lower educated categories. While peasants had been the stronger supporters of Europe in 1972, there were now its stronger opponents, with 62% of farmers voting No. More in line with their previous vote, 61% of workers rejected the treaty (half of them had voted No in 1972). Conversely, the Yes was urban, rich, and scored the highest, like in 1972, among high tertiary employees (67%) and liberal professions (66%), and most of all intellectual professions (71%) (BVA exit poll, quoted in Portelli, 1992:11). Actually, the pro-Maastricht vote was first of all the vote of well-educated categories, with 72% of higher education graduates voting Yes (Habert, 1992:874-6). The influence of the level of education was found among all social categories, quite independently from the social and income category, as evidenced by the highest scores of the Yes among teachers, whether in the primary, secondary or higher education (Criddle, 1993: 237). If 1992 marks the structuring of a powerful divide on Europe in French society, it was also for the first time the expression of a vote on Europe, contrasting with European elections and of course the 1972 referendum on enlargment. Voting studies show that voters expressed themselves prevalently on the issue at stake and not, as in the previous referendums under the Fifth Republic, on the action of the government or the President. Admitedly, this was less true of No voters, who counted a substantial proportion of people dissatisfied with François Mitterrand and government action (39%) or rejecting the political class (31%). But Yes voters voted at 72% “to assure a lasting peace in Europe” and 63% because the treaty was “indispensable for the building of Europe”. Only 14% said they wanted to support François Mitterrand (BVA exit poll, multiple responses, quoted in Criddle, 1993: 238). Furthermore, the analysis of the correlation between the vote and the index of approval of the treaty showed that 97% of No voters had low approval while 88% of Yes voters expressed high approval (Habert, 1992: 877-880). Thus attitudes toward Europe, in particular the cleavage between “Sovereignists” and “Europeists”, were the main explainatory factor of the vote in 1992 (Cautrès, 2005:149-50). The European issue was built up gradually during the campaign. Whereas in 1972 the campaign had been powerless to arouse interest in the matter, in 1992 on the contrary people interested in the referendum grew from 58% at the end of June to 67% in early September and 71% on the eve of the vote (Figaro-Sofres poll quoted in Habert, 1992:871 and Piar and Gerstlé, 2005:44 ). Moreover, the campaign was essential for the formation of opinion, 40% of voters saying they decided in the course of it (Piar and Gerstlé, 2005:144).


“Do you approve the bill authorising the ratification of the treaty establishing a Constitution for Europe?”

The difficult victory and almost boomerang effect of the Maastricht referendum are certainly not unrelated to the fact that neither the treaty of Amsterdam nor the treaty of Nice were subsequently put to the referendum. And the 2005 referendum on the constitutional treaty, as said in the introduction, was a first case under the Fifth Republic of a “politically obligatory” referendum (Grunberg, 2005; Morel, 2007). For the first time, the President did not fully control the decision to hold the referendum, which was largely imposed on him by parties and civil society. Not that tactical motives were completely absent. There were in a way symetric to those of Mitterrand with the Maastricht treaty: although not guaranteed, a victory could revitalize his authority at mid-mandate and counterbalance the poor results of the Right at the recent regional and European elections; it could also enhance the President’s image as an historical Gaullist and builder of Europe; finally, it was likely to divide the Opposition, especially the Socialist party (e.g. Martin, 2005). There is no doubt either that the President’s decision also answered to a logic of appropriateness, thus reflected his belief that such an important text had to be ratified by the people. He was constrained by a chorus of voices, either from the Left and the Right, which claimed for a referendum. The pressure grew stronger whe Tony Blair officially announced in Spring 2004 that he would submit the European Constitution to the British people. This was followed by the leaders of the Socialist party (François Hollande) and of the UMP (Alain Juppé and Nicolas Sarkozy) officially asking for a referendum. Former Presidents of the Republic Giscard d’Estaing and of the European Commission Jacques Delors declared that the referendum had become unavoidable. At this point, had he opted for parliamentary ratification, Chirac would have jeopardised the future legitimacy of the Constitution and its acceptance by the country. Moreover, he would have exposed himself to the criticism of being a poor democrat, a pseudo-Gaullist afraid of universal suffrage and obstinately deaf to the people after the double sanction of the regional and European elections. After lengthy hesitation, he thus announced the referendum to be held on May 29th 2005

As we know, Chirac’s fears were grounded since like in 1992 voting intentions dramatically dropped, from an initial 65% for the Yes still in January to an actual 45,3 % in the ballot, which was fatal to the treaty since in the aftermath eight countries declared postponing its ratification (Laurent and Sauger, 2005: 161). This was the second referendum lost by the President under the Fifth Republic, but unlike de Gaulle, Chirac did not resign – just as he had not resigned in 1997 after his early election failed “coup”, which had brought a Left majority to the low Chamber. He just replaced the prime Minister. Nonetheless, the impact of the referendum was clearly devastating for him, as it was for François Hollande, who ended up with a severely divided party and lost the primaries for the presidential election one year later. The effects of this referendum were long lasting for the Socialists, since it was the catalyst of the internal rebellion against the “culture of government”, which will disrupt the whole Holland presidency and end up with the party implosion in 2017.
But the referendum, like with the Maastricht treaty, raised a genuine interest in the population, almost a frenesy, as shown by the intense large-scale debate in the country and the almost 70% participation rate (same as in 1992 and much more than the respectively 66% and 43% at the previous year regional and European elections ). No doubt that interest and participation were also boosted, like in 1992, by the high level of controversy and the uncertainty about the result, which in return created strong media coverage. Never before during the Fifth Republic had a campaign proved so volatile. Voting intentions in favour of the Yes had also shrinked dramatically in 1992 (see above), but stabilized at some point, while in 2005 the curves of the Yes and the No crossed three times in nine weeks. And 44% of voters made up their mind during the campaign or in the last days (Piar and Gerstlé, 2005:44-45; Grunberg, 2005: 132-5)

The referendum also contributed to remap the cleavage on Europe: while sovereignty remained the main issue for Right-wing opponents to Europe, on the Left anti-capitalism turned into a denunciation of socio-economic drawbacks of the single Market, such as firms’ relocations and social dumping, which was in part shared by the populist Front national. This led electors in principle favourable to Europe, which had approved the Maastricht treaty, to vote this time No, not to Europe, but to Europe “as it works today”. These electors, a substantial part of which were Socialists, made up a third of No voters and greatly contributed to the failure of the treaty (Brouard and Sauger, 2005:135-6; Brouard, Sauger and Grossman, 2007:77-116; Cautrès, 2005:151).

Like in 1992, the vote quite reflected the situation in the various parties (Brouard and Sauger, 2005: 123-125; Grunberg, 2005: 135-137). Thus the PS and the Greens, which were split, had their electors divided. Both parties had held internal referendums to decide the party official position, and the No had lost by very little, with the result that its leaders felt encouraged to campain openly against the treaty (which was tolerated by the leadership). It had reached 41% in the PS, quite less however than the 56 % of Socialist electors who voted No (people declaring themselves close to this party, IPSOS exit poll quoted in Grunberg, 2005:137). Conversely, parties which were united had a united electorate, and the maximum loyalty was found among those advocating the No, namely the PC (98% of No voters) and the National Front (93% of No voters). The UDF and the UMP, which were both for the Yes, had respectively 76% and 80% of electors voting Yes.

In 2005 the vote returned fairly homogeneous troughout the country, like in 1972, but this time in favour of the No, which was first everywhere, with the sole exception of Paris and the traditionally pro-European strongholds of Brittany and Alsace. In spite of this “irresistible nationalization of the vote on Europe” (Boy and Chiche, 2005), it was possible still to identify areas of particular strength of the No (over 63%) , which were roughly the same as in 1992: the North, the South and the Center. Again, the Yes was highest in large cities (over 100 000 inhabitants) and in the most privileged areas (typically in the richest Parisian districts), while the No was leading in the working class cities of the North, the Parisian belt, and in the South East areas of strength of the Front national (Boy and Chiche, 2005: 99).

Such geographical distribution clearly points to the importance of social and educational factors. Just as in 1992, the level of education was the key-variable, the only categories voting Yes being the “bac+2” (54%) and “bac+3” (64%), the high tertiary employees and liberal professions (65%), and the students (54%) – to which must be added people over 60 (57%) (IPSOS exit poll, quoted in Grunberg, 2005:136. Note bac is equivalent to A level.). Conversely, 72% of those without diploma voted No, and, unsurprisingly, peasants were again massively against the treaty (70%), just as workers and unemployed people, whose opposition to Europe increased compared with 1992 (respectively 79% against 61% in 1992, and 71% against 59% in 1992). But the strongest novelty was undoubtedly the hostility to Europe of social categories which voted Yes in 1992, namely intermediate professions (53% of No against 38% in 1992) and employees of the public (64% of No against 49% in 1992), the latter voting this time more than employees of the private (56% of No against 50% in 1992) against the treaty.

In his address of 26 May 2005, Chirac had declared: “We should not mistake the question. The decision before us goes far beyond traditional political cleavages. It is not a matter of saying ‘yes’ or ‘no’ to the government. It is about your future and that of your children, the future of France and Europe. ” This shows the persisting fear of French Presidents that the vote is turned into a vote of confidence. Actually Mitterrand had made a similar statement in 1992 and made clear, like Chirac, that he would not resign in case of a No victory. Moreover Chirac was afraid that the vote could deviate on the issue of Turkish membership. For this reason, he had secured the adoption by Parliament in the months preceding the referendum of a constitutional amendment requiring a referendum for every new membership.

Chirac’s fears were not confirmed. Turkish membership was not a central motivation in the No vote (brouard and Sauger, 2005:132). Like in 1992, domestic problems played a greater role in the rejection of the treaty, with 52% of No voters saying they had voted thinking more of national problems rather than Europe, while 81% of Yes voters said they had thought more of Europe (CSA exit poll, quoted by Piar and Gerstlé, 2005:60, Grunberg, 2005:140). Similarly, other surveys showed that those who intended to vote No were also the most critical with the government and the President, and the less confident in the improvement of the standard of living (Brouard and Sauger, 2005:128-9; Ivaldi, 2006). But, just as the Maastricht vote was correlated to attitudes toward the treaty (see above), the vote on the constitutional treaty was strongly related to attitudes toward Europe (Brouard and Sauger, 2005: 130-5; TNS-SOFRES poll, 11-17 May 2005, quoted in Grunberg, 2005:141). Thus, Europe was actually “at the top of the head” on May 29th (Grunberg, 2005:142) and “electors have decided, as a matter of priority, on the question put to them” (Cautrès, 2005:145).

C. Current debate

The current French paradox is that although absent since 2005, the referendum has been present as never before in the political debate. For sure, it is still not a big issue in French politics. To a large extent it is limited to being a campaign issue at presidential elections. But there was a clear increase in its saliency from 2007 to 2012 and then 2017. Admitedly, 2007 was marked by the 2005 trauma and the referendum was not popular among the two finalists : although being in favor of a new referendum on European institutions (proposal 91), Ségolène Royal did not mention the device in her crusady for participatory democracy, preferring tools such as participatory budgets, citizen juries, or the agenda initiative (proposal 73); while Nicolas Sarkoky, in spite he had urged Chirac to hold the referendum on the constitutional treaty as the UMP General Secretary, had turned around and publicly adopted a very critical tone about the device. Once elected, he failed to submit the Lisbon treaty to the voters as well as the large constitutional revision of 2008. The introduction of a joined parliamentary and citizen agenda initiative (see section D) as part of this revision was largely forced on him, notwithstanding the fact that the popular initiative had figured out in his 2007 electoral program. And he did not ensure the implementing legislation of this new mechanism to be passed during his term in office. This was only done under Hollande – although the new elected president was far from a supporter of referendum. Finally Sarkozy pressed to abolish provisions for a mandatory referendum on new EU memberships introduced by Chirac in 2005 (in an effort to dissociate the issue of Turkish membership from the vote on the constitutional treaty), but was only able to make such referendum optional (the decision relying on Parliament).

2012 was the big return of the referendum, as a result of Sarkozy’s new turnaround on the matter. The outgoing President campaigned this time as the “People’s candidate” with the referendum as a flagship proposal of his program.This allowed him on one side to differ from François Hollande, who claimed “the importance of respecting representative democracy” (interview in weekly newspaper Marianne, 17/02/2012) and defended an exceptional use of referendums on institutions and European issues, and on the other side to compete with the referendum-oriented populist candidate of the Front national Marine le Pen, who predilected referendums on migratory and islamic issues. Sarkozy remained very vague about which questions should be put to the referendum. He just declared that he would call the French to decide on « major issues » and « each time there will be a deadlock », insisting on the responsibility of intermediary bodies in blocking reforms. Among the main six candidates, François Bayrou (Center-Right), who was considered a serious challenger because of his score very close to that of the two leading candidates in 2007, also participated in putting the referendum on the campaign agenda. He announced that if he was elected he would hold a referendum on a package of institutional reforms on June 10th, the same day as the first round of parliamentary elections. This was clearly a tactic intended to help him, as the leader of the small MODEM (Mouvement Démocrate), to win a majority at this election. But it contributed to draw the outlines of a pro-referendum presidency, all the more since Bayrou, although not a Gaullist, had a Gaullist conception of institutions, and was the fiercest critic of Sarkozy’s decision not to submit the Lisbon treaty to the voters. Clearly enough, one will never know whether Sarkozy would have this time respected his commitment to referendums, but the victory of François Hollande unsurprisingly meant a halt to this referendum momentum. The only exception was the the President’s stance in favour of local referendums following the death of an ecologist activist in a demonstration against the Siven’s dam in 2014. This was followed by the announcement of a local referendum on the planned new airport of Notre-Dame-des-Landes. But the Socialist President did not organize any nationwide referendum, like his predecessor.

The Brexit vote, which just preceded the launching of the Republican party’s primaries for the 2017 presidential election, brought the referendum back into the public debate. To a large extent, the question of holding an EU-related referendum became a Left-Right divide since prime minister Manuel Valls clearly stated on June 28th in a speech in the National Assembly about Brexit’s consequences that a referendum on an amended treaty or a new European project was not an issue for the Executive (« a referendum cannot be the way to get rid of a problem, let alone a backdoor to solve internal policy problems ») ; while on the Republican side, the main candidates to the primary, namely Sarkozy, Fillon, Juppé and Le Maire, all advocated a referendum. Admittedly, only the two latter kept this point in their program a few months later – Juppé asked for a pan-European referendum and Le Maire for a strictly national one. But in the meantime five out of the seven candidates had developped generous proposals for referendums on various subjects : Sarkozy on the reduction of the number of parlamentarians, the prohibition of multiple mandates (which was just passed by the Left, the end of “automatic” family regroupment for migrants and the preventive internment of people classifyied as “dangerous”; Juppé on a new European project; François Fillon on the constitutional requirement of budgetary equilibrium, the reform of local authorities, the end of special pension schemes, the introduction of immigration quotas and the reduction in the number of parliamentarians; while Bruno Le Maire, in a Bayrou-like tactic (see above), claimed he would hold a referendum the same day as the second round of the legislative elections on a four issues project of institutional reforms for the moralization of political life; and the almost unknown Christian-Democrat candidate Jean-Frédéric Poisson advocated a referendum on the right of foreigners to social assistance. The Brexit was indeed the catalyst of this unprecedented rise of the referendum in the public debate, which reflected a double crisis: on one side, a crisis of confidence toward parties and politicians against a backdrop of growing demand for referendums by citizens (e.g.; on the other side a crisis of performance of the political system and the widely shared view that the country was not reformable through “normal means” leading most candidates to advocate the recourse either to the referendum or to decrees (see

The referendum became an all-parties issue in the weeks prior the presidential election. Most candidates were now proposing specific referendums and five out eleven were in favour of popular initiative. On the Right, Fillon sticked to his five referendums to be held under existing constitutional provisions (see above). On the Extreme Right, Marine Le Pen and Nicolas Dupont-Aignan, who allied between the two rounds, advocated a referendum on “European refoundation” and the introduction of popular initiative. On the Left, the stronger advocate of the referendum was now Mélenchon, who clearly boosted his direct democracy program compared with that of 2007, being now in favor of popular initiative and mandatory referendums on all constitutional revisions and new EU treaties (in addition to demanding the recall for any elected representative). But Hamon was also not short of ideas with his proposal to introduce the agenda initiative and the rejective referendum at the initiative of 1% of registered voters, and his announcement of referendums on various subjects, such as voting rights to foreigners or the blank vote.

However, a major exception in this general infatuation for the referendum was Macron, who carefully remained mute on the subject. It looked as if he had forgotten his post-Brexit commitment in favour of a transnational referendum on a new European project after its discussion by national citizens’ conventions (“I believe that we must today consult the peoples, ask them their opinion. (…) But this must be done in an appropriate framework, not on the basis of a referendum tomorrow on this Europe, but first by building with the European peoples this new project and then submitting this new roadmap, this new project, to a referendum”. Speech at Sciences Po University, 25 June 2016 Thus, like Hollande, the new elected President has so far rather cooled the referendum frenzy, although he declared in July in his message to the Congress that he would not hesitate to submit his constitutional reform to the people if necessary (i.e. in case the parliament would not reach the qualified majority). Clearly enough, the referendum is conceived by Macron as a last resort. In a Pericles-like speech in Athens on September 7, 2017, he has revived his idea of democratic conventions, but now seems to conceive them rather as an alternative to a referendum: “I want us to get out of this type of infantile dilemma in which Europe is now plunged, with on the one hand those want to go and get the people and make them say a ‘yes’ or a ‘no’ that they will manipulate for months, when the referendum becomes the weapon of the populists, the anti-Europeans, and on the other side those who really believe in Europe end up afraid of their peoples and hide behind their own doubts and say to themselves: ‘Let’s move forward, but let’s never change the treaties for fear of returning to a referendum. Let’s move forward, but by little steps between us, people will not understand that’. We must choose another path, a third one, the path invented here, the path invented in the very place where we find ourselves, which was not that of demagogy: it was that of democracy, controversy, debate , construction by criticism and dialogue, which consists in returning to the intimacy of each question and its complexity, namely what we want for the common city. That is what I want in the first half of 2018 in all the countries of our continent, of our Europe”. ( It still remains to be seen then whether the Macron presidency will or not be the trigger of a new referendum on Europe.

D. Legal regulations

The various types of nationwide referendums – with regard to their basic features such as the initiative, the object, the form of the text, approval requirements or the legal impact of the vote – are provided for in the Constitution in articles 11, 89, and 88-5. All these standards can be and have actually been changed through an amending procedure which does not require the referendum. A special « organic » law (a law which requires a qualified majority) regulates in detail the so-called « joint initiative » of a parliamentary and popular minority , which has been introduced in article 11 since 2008 (and still never used); but this is actually an agenda initiative, not a referendum initiative – although it passes to be one – since it is enough that the proposal has reached the number of signatures is debated in parliament within a certain lapse of time for it not to be submitted to the people.

Article 11, which field has been enlarged in 1995 and 2008, now provides for an optional referendum on a large range of issues. It may deal with any «  draft law on the organization of public authorities, on reforms relating to the economic, social or environmental policy of the nation and to the public services that contribute to it, or for the purpose of authorizing the ratification of a treaty which, without being contrary to the Constitution, might affect the functioning of the institutions ». This is not clear whether the “organization of public authorities” can include revisions of the constitution. Most lawyers reject this possibility on the ground that there is a specific article in the constitution for constitutional revision (article 89). But de Gaulle used twice article 11 to revise the constitution. The referendum is initiated by the President of the Republic under proposal of either the government or the two chambers; but in practice, it was always initiated under proposal of the government and amounted to a discretionary power of the President, who managed to control the majority and the choice of the prime minister – a situation which was frequent before the 2000 reform and is now almost guaranted as seen above. When the referendum is proposed by the government, the prime minister must make a statement followed by a debate in each assembly, but no vote is required. All referendums since 1958 have been held under article 11, apart from the founding referendum of the Fifth Republic, whose legal basis was a special law, and the 2000 referendum on the reduction to 5 years of the presidential term (article 89).

Article 89 provides for two types of referendums on constitutional revisions. First, the optional referendum on a draft introduced by the President of the Republic (on proposal of the prime minister). After the draft has been approved by both chambers, the President can decide either to submit it to the voters or to the chambers convened in congress, where it needs a majority of three-fifths of the vote cast to be adopted. Second, a mandatory referendum on a draft initiated by MPs after it has been approved by both chambers.

Finally, a referendum on new accession treaties was introduced in 2005 under the impulse of President Chirac (see above) but was never used. While the referendum was mandatory in the first place, it became optional after the constitutional revision of 2008. Thus, according to Article 88-5, any bill authorizing the ratification of a treaty of accession of a State to the European Union is submitted by the President of the Republic to the referendum unless each Chamber decides by a majority of three-fifths that the treaty will be ratified by the two Chambers convened in congress with the same majority (as for the constitutional amendments).

All these types of referendums are binding and have a direct legislative effect if approved since they bear exclusively on draft laws (they cannot relate to a question of principle or a non-drafted proposal). They can propose new legislation but also be suspensive, in the case of a constitutional revision adopted in first reading by the parliament, or abrogative, when the draft proposes to abrogate a law or a part of a law. There is no quorum requirement and voting is not compulsory. Finally , there is no obstacle to the result of a referendum being challenged by a procedure excluding the referendum; thus for example the statutes of the French territory New Caledonia approved by referendum in 1988 were substantially modified in 1999 by parliament, or the Constitution for Europe rejected by the French in 2005 was approved, albeit in a modified form, by Parliament in 2008 (Lisbonne treaty).

Beyond these basic characteristics, the referendum procedures are not laid down in a specific referendum legislation and there is a lack of codification. The detailed organization of referendums is mainly based on five ad hoc presidential decrees, which mostly transpose to the referendum the rules of the electoral code, especially regarding the campaign and the vote. Thus the referendum tends to be treated as an election, which sometimes appears arguable. The main issue in this respect is the attribution of public resources to parties, in part proportionnally to their strength – seats in parliament or electoral results. Until the last referendum these resources consisted exclusively in speaking time on public waves, but in 2005 the government decided to grant also public funding. Like for elections, campaign funding must be transparent and is controlled by the “Commission nationale des comptes de campagnes et des financements politiques” (CNCCFP), but contrarily to elections there is no maximum expenses. Government is also not allowed to campaign or to use public resources for the purpose of propaganda, as was made clear during the Maastricht campaign when the president of the “Conseil Supérieur de l’Audiovisuel” (CSA), which controls compliance with rules for expression on radio and television, asked it to renounce its pro-Maastricht spots – or during the 2005 campaign, when the president of the Constitutional Court this time invited a minister to stop to campain too openly. The question of the formal and material validity of the text is completely unregulated, the jurisprudence of the Constitutional Court being the only reference. Thus recent judgments of the Court, since it agrees to rule on the decrees organizing the referendum (see below), put forward the requirement of clarity and loyalty of the question, but these notions remain little explained. Possibly loyalty could refer to the criteria of unicity and unbiasedness of the question and unity of the subject, which are common in the legislation of other countries (Roche, 2012).

Finally, the judicial control remains very uncertain, insofar as it not explicitely provided for and can only work on request (of one or several electors). If the appeal against the decisions of the authorities in charge of controlling the campaign uncontroversially fall under the Council of State (the higher Administrative Court), things are much less clear with regard to the five decrees calling for the referendum and defining the rules of the campaign and the vote. Here the competences of the Constitutional Court and the Council of State are neither clearly defined nor clearly distributed. It is mostly a matter of jurisprudence. Before 2000, the control of these decrees was regarded as falling within the competence of the Council of State, which however refused to rule on the first decree, which convenes the referendum and states the wording of the question. There was therefore no control over the initiative and the question. But since 2000 (Hauchemaille jurisprudence), the Constitutional Court has accepted to rule on all five decrees, including the first. This points indisputably in the direction of increased control of the referendum. Thus, the doctrine now considers that the Court may dismiss a referendum on procedural grounds or because the question does not respect the requirements of clarity and loyalty. It would also probably reject a referendum which question clearly does not fall within the referendum domain, knowing however that the latter remains quite vague. It is more arguable whether the Court would accept to control the compliance of the proposed law to superior law (Fatin-Rouge Stefanini, 2004 and 2005). In the absence of an appeal, it is only required that the Court expresses an advice on the decrees, which since 2000 must be published. It should be stressed that in any case the control can only take place before the vote, since the Court has reiterated several times its refusal to control the referendum law. Diversely, nothing prevents the Council of State or any administrative court from controlling on request that the referendum law complies with treaties or Community law.

E . Web-links and citations

Digithèque de matériaux juridiques et politiques:
(complete results)
Constitutional Court: (see also folders on specific referendums in “dossiers thématiques”:

APPLETON, Andrew (1992), “Maastricht and the French party system: domestic implications of the treaty referendum”, Journal of Comparative Politics, 10/4, pp 1-18.

BERLIA, Georges (1972), « Le référendum du 23 avril 1972 », Revue française de droit public, vol 88/4, pp 929-943.

BOY, Daniel et CHICHE, Jean (2005) « Le référendum français du 29 mai 2005 ». « L’irrésistible nationalisation d’un vote européen », in Pascal Perrineau (dir.), Le vote européen 2004-2005. De l’élargissement au référendum français, Paris, Presses de Sciences Po, Coll. “Chroniques électorales”, pp. 229-244.

BROUARD Sylvain, SAUGER Nicolas (2005), “Comprendre la victoire du ‘Non’: proximité partisane, conjoncture et attitude à l’égard de l’Europe”, in LAURENT, Annie et SAUGER, Nicolas (dir.), Le référendum de ratification du Traité constitutionnel européen : comprendre le « non » français, Les Cahiers du CEVIPOF, n°42.

BROUARD Sylvain, SAUGER Nicolas, GROSSMAN Emiliano (2007), Les Français contre l’Europe?, Paris, Presses de Sciences Po, 352 p.

CARRE DE MALBERG, Raymond (1931), « Considérations théoriques sur la question de la combinaison du référendum avec le parlementarisme », Annuaire de l’Institut International de Droit Public, vol. II.

CAUTRES, Bruno (2005), “Les clivages socio-politiques sur l’intégration européenne et le vote du 29 mai 2005”, in LAURENT, Annie et SAUGER, Nicolas (dir.), Le référendum de ratification du Traité constitutionnel européen : comprendre le « non » français, Les Cahiers du CEVIPOF, n°42.

CRIDDLE, Byron (1972), « Politics by plebiscite in France », The World Today , Vol. 28, No. 6, pp. 240-248.

CRIDDLE, Byron (1993), « The French referendum on the Maastricht treaty September 1992 », Parliamentary Affairs, Volume 46, Issue 2, pp 228-238.

DICEY, A.V. (1890), “Ought the Referendum to be Introduced into England?”, Contemporary Review, vol. 57, April 1890.

ESMEIN, Adhémar (1906) Eléments de droit constitutionnel français et comparé, Paris, Sirey,(1st edition).

FATIN-ROUGE STEFANINI, Marthe (2004), Le contrôle du référendum par la justice constitutionnelle, Paris, Economica, Coll : Droit public positif, 381 p.

FATIN-ROUGE STEFANINI, Marthe (2005), “ La décision du Conseil constitutionnel du 24 mars 2005, Hauchemaille et Meyet : un nouveau pas en matière de contrôle des référendums “, Revue française de droit administratif, pp. 1040-1048.

GRUNBERG, Gérard (2005), “Le référendum de ratification du Traité constitutionnel européen”, French Politics Culture and Society, 23/3, pp 128-144.

HABERT, Philippe (1992), “Le choix de l’Europe et la décision de l’électeur”, Commentaire 60, pp 871-880.

IVALDI, Gilles (2006), “Beyond France’s 2005 Referendum on the European Constitutional Treaty: Second-Order Model, Anti-Establishment Attitudes and the End of the Alternative European Utopia”. West European Politics, 29:1.

LAURENT, Annie et SAUGER, Nicolas (dir.) (2005), Le référendum de ratification du Traité constitutionnel européen : comprendre le « non » français, Les Cahiers du CEVIPOF, n°42, juillet 2005.

LEIGH, Michael (1975) , « Linkage politics : the French referendum and the Paris summit of 1972 », Journal of Common Market Studies 14 :157-170.

LELEU, Claude (1976) , « The French referendum of April 23, 1972 », European Journal of Political Research, 4 : 25-46.
MARTIN, Pierre (2005), “Le choix de Chirac ou pourquoi Jacques Chirac a-t-il chois la voie référendaire pour la ratification du traité constitutionnel?”, in LAURENT, Annie et SAUGER, Nicolas (dir.), Le référendum de ratification du Traité constitutionnel européen : comprendre le « non » français, Les Cahiers du CEVIPOF, n°42.

MIRKINE-GUETZEVITCH, Boris (1931), « Le référendum et le parlementarisme dans les nouvelles constitutions européennes », Annuaire de l’Institut International de Droit Public, vol. II.

MOREL, Laurence (1996), “France: towards a less controversial use of the referendum?”, Michael Gallagher and Piervicenzo Uleri (ed.), The Referendum Experience in Europe, London, Macmillan, pp. 66-85.

MOREL, Laurence (2007), “The Rise of Politically Obligatory Referendums. The 2005 French Referendum in Comparative Perspective”, West European Politics, vol. 30, N° 5, pp.1041-1067.

MOREL, Laurence (2010), “La Ve République, le référendum et la démocratie plébiscitaire de Max Weber“, Jus Politicum, n°4, 55 p

MOREL, Laurence (2014) , « The Referendum in the Overall Balance of the Fifth Republic”, lecture at Collège de France, Paris, March 12th, 2014 :

PARODI, Jean-Luc (1973), « La Cinquième République et le système majoritaire. Le référendum ou l’apparition d’une question de confiance présidentielle », Thèse de doctorat de l’Institut d’Etudes Politiques de Paris.

PASSELECQ, Olivier (1990), « De Tardieu à de Gaulle : contribution à l’étude des origines de la Constitution de 1958 », Revue française de droit constitutionnel (3): 387-407.

PIAR, Christophe and GERSTLE, Jacques (2005), “Le cadrage du référendum sur la Constitution européenne: la dynamique d’une campagne à rebondissements”, in LAURENT, Annie et SAUGER, Nicolas (dir.), Le référendum de ratification du Traité constitutionnel européen : comprendre le « non » français, Les Cahiers du CEVIPOF, n°42.

PORTELLI Hugues (1992), “Le référendum sur l’Union Européenne”, Regards sur l’actualité, sept.-oct. 1992, pp 3-12.

ROCHE, Jean-Baptiste (2012) « Le contrôle juridictionnel du référendum en France – Première partie », Revue Juridique de l’Ouest 2012-1; “- Seconde partie », RJO 2012-2.

SVENSSON, Palle (2018) « Views on Referendums: Is there a Pattern?”, L. Morel and M. Qvortrup (eds) The Routledge Handbook of Referendums and Direct Democracy, Routledge (ch. 5).