The French "Petit Oui":
The Maastricht Treaty and the French Voting Agenda

Despite expectations of a landslide, the French public barely approved the Maastricht Treaty in 1992. Traditional explanations for the surprise outcome that rely on socioeconomic conditions, partisanship, regional characteristics, and domestic issues are inadequate. Analysis based largely on the French National Election Study of 1995 shows that the voters were largely concerned with foreign policy. Citizens who were friendly to Germany and supportive of the European Union greatly favored the treaty. French nationalists and those possessed of strong anti-immigrant sentiment were vehemently against it. These disparate views, united in their concern about the sovereignty of France, dominated the Maastricht vote and appear to have persisted in the European Union Constitutional referendum of 2005.

On September 20, 1992, the impossible almost happened. The French public nearly rejected the Maastricht Treaty, with its provisions for new and serious commitments by France to the European idea. The referendum to ratify the treaty received 51 percent of the valid ballots, a "petit oui" indeed, surprising virtually everyone. Leading up to the referendum, politicians and pollsters were confident that support was overwhelming. For example, a SOFRES (Societé Française d' Enquêtes par Sondage )/Le Figaro poll in June showed 76 percent of those with an opinion in favor. The evaporation of this support when the actual vote was taken sent a shock through the European political and economic community. Clearly, the French had made an extraordinary choice. Why did French citizens vote as they did in this referendum? What happened during the summer to bring about such a change in public opinion? This article offers an alternative to conventional explanations of the unforeseen outcome, based on a rich data set from a neglected survey, and draws out appropriate implications regarding the May 2005 referendum on the European constitution.

The Events of Summer 1992

From the moment when President François Mitterrand took office in 1981, he worked assiduously to move the "European project" forward, specifically the Single Market Program of 1985 and Economic and Monetary Union (EMU). These major integrationist commitments met with little effective opposition. On the far left, the Communists were increasingly isolated, as their vote totals plummeted. The Gaullists, keen to return to power, based their attack on the ruling Socialists on [End Page 65] strictly domestic issues, partly to stave off the rise of the National Front. The citizenry itself seemed unconcerned with these remote institutions, which had scant reality for their daily lives. Commenting on the public's opinion about the construction of Europe, Mayer concluded that "the very content of the European idea does not seem very clear to them." As Ross observed, "renewing European integration was cost-free for a time," an especially useful circumstance for Europeanists like Mitterrand, "who wanted to keep it off domestic political screens." In sum, among the few who thought about Europe, there was a positive consensus, but it did not appear high on the agenda of everyday politics.1

The June announcement of the referendum on the Maastricht Treaty, however, placed Europe directly at the forefront of the political agenda. Was the referendum necessary? No. In France, treaty ratification could have happened via a referendum or a vote in parliament. The parliamentary option required a three-fifths majority in both the National Assembly and the Senate. Though a demanding hurdle, a parliamentary quota was likely, as evidenced by an April National Assembly vote on a relevant constitutional amendment, which passed 398 to 77. Clearly, Mitterrand could have settled business quickly through parliamentary procedure. But he declared for a referendum. Why?

First, Mitterrand was determined to maintain French leadership on the continent. Key European issues were on the table, among them the European Free Trade Area and the EMU. Of particular importance was the fact that the Danes had just rejected the Maastricht Treaty. A resounding French "oui" would renew the momentum for Europe, and place it more firmly under his direction. Additionally, at the time of the referendum announcement, most political observers had little reason to think that ratification would not prove a resounding success. Second, Mitterrand wished to advance his domestic political causes. He was faring poorly in popularity polls. In March and April, the percentage of those "satisfied" with him in the monthly Institut français d'opinion publique (IFOP) survey was only 26 percent, close to his lowest rating. [End Page 66] More generally, the Socialist party was fearful of defeat in the 1993 legislative elections. Mitterand believed that a referendum win would ultimately benefit the Socialist candidates in the forthcoming National Assembly contests.

The immediate consequences of this electoral mobilization were several. When it finally surfaced, elite discord was virulent. The debate caught the popular eye, resulting in serious defection from party appeals. The campaigners for Maastricht had made serious miscalculations. Before the referendum, a quiet, if uneasy, consensus had obtained among leading party elites regarding the value of European integration. This consensus, fragile as it was, was simple to maintain. But the announcement forced contestation. The right was sorely divided. Within the Rassemblement du people français (RPR), spokesmen Charles Pasqua and Phillipe Séguin were openly hostile, the latter condemning the treaty as a "historic stupidity." Other RPR leaders were supportive, but to varying degrees. Edouard Balladur was unabashedly positive, but Jacques Chirac waffled. In contrast, the heads of the Union pour la démocratie française (UDF) united solidly in favor of the treaty. The National Front continued its strident opposition. The left did not escape its own divisions. Within the Socialist party, Jean-Pierre Chevènement actively protested against approval. The Greens, split internally, took no official position on the issue. Only the Communists were steadfastly opposed, in keeping with their long- held view that Europeanization was an instrument of capitalism.2

The dissent about the treaty within and across party elites spilled into the public debate that arose in anticipation of the approaching vote. The sudden, looming presence of the referendum framed the question of Europe anew, moving it from abstract consensus to real politics. As its highly partisan nature became clear, the electorate showed considerable movement. Many party followers broke with their leaders' call to support Maastricht. On the traditional right, defection was especially heavy among the RPR supporters. On the traditional left, about two-thirds of Socialists [End Page 67] reported a "yes." Slightly more than one-half of the Greens and of the UDF gave their assent.3

The campaign itself was little short of a disaster. As Appleton observed, its errors "bordered on the comic." The Socialist government did little to advance the cause until well into the summer, either because it felt assured of a win, because of the president's shaky health, or because the "official" campaign was not scheduled to start until September 7. In August, it began an active pro-Maastricht initiative, led by the well-known Jack Lang; it was attacked on the grounds of being superficial. On September 3, Mitterrand granted a television interview on the subject. The leaders of the traditional right, Valéry Giscard d'Estaing and Chirac, jumped squarely onto the Treaty bandwagon in the closing days, seized with the sudden fear that their European dream might be dashed. This fear was not unfounded; August and September polls showed "oui" support at about 50 percent.4

The events of summer—the calling of the referendum, the elite and mass political reactions, and the management of the campaign—caused the widespread public support for Europe in the spring to vanish by the fall, setting the stage for ratification by the slimmest of margins. Voters in the polling station had many factors to consider. Which of them turned out to be the most important?

Conventional Explanations of the Treaty Vote

Standard explanations for the surprising Maastricht numbers draw on the political-sociology traditions that have generally served scholars of French elections well. The core literature relies on three types of variables: socioeconomic conditions (which French election scholars sometimes call "the heavy variables"), partisanship, and regional characteristics. In her examination of exit-poll data, Mayer reports clear percentage differences in the "yes" vote by occupation, education, and income. Two occupational groups indicated a majority in favor—professionals/upper management (70 percent) and middle management (57 percent). All other major occupational groups were opposed. Farmers were least in favor (29 percent) and workers next to least (42 percent). Both education and income are ordinally and positively related to support, at [End Page 68] about the same degree. For example, 46 percent of those with an elementary-school education and 71 percent of those with a university education voted "yes." Only 43 percent of those earning less than 5,000 French francs (FF) monthly supported Maastricht, as opposed to the 64 percent of those earning more than 20,000 FF monthly.5

The same exit-poll data show that both left–right ideology and political party were related to the treaty vote. Of those who placed themselves on the political left, 75 percent were in favor and of those who placed themselves on the right, 43 percent were. The relationship of party identification to vote demonstrates a similar pattern. The two biggest parties, the Socialists (76 percent) and the rpR (20 percent), represented the dominant partisan rivalry. The Communists and National Front members, as expected, gave few "yes" votes (16 percent and 15 percent, respectively). Disarray within the traditional parties on the right was reflected in the 59 percent "yes" recorded by UDF backers.6

Analysis of the electoral geography of the treaty vote, by region, contradicts some of the survey findings. The "yes" vote was greater in regions where Catholicism was strong—for example, Brittany and Alsace. This regional result is also compatible with a linguistic explanation; regions identified by language, such as Brittany and Alsace, might have expressed support for the EU because they saw it as a vehicle for their regionalist aspirations. In another aggregate data study, the "yes" vote was found to be greater in regions with many farmers and few unemployed people. One team of analysts makes a forceful argument that the Maastricht vote was decided by regional considerations: "Thus Europe gave rise to a regional referendum: every land voted according to its political history, its economic development, its religious traditions, all together producing a specific perception of the construction of Europe."7 [End Page 69]

These conventional explanations of the Maastricht vote are informative but inadequate in several senses. First, they tend to be bivariate, increasing the risk of the spurious inferences that multivariate controls would have ruled out. Second, the regional data are susceptible to the problem of the ecological fallacy. Following these aggregate studies, Catholicism and farming increased Maastricht support, but individual-level survey data fail to uphold such conclusions. Last, but most important, all of the studies, survey or aggregate, offer a structural explanation of the treaty vote. That is, the sociological, political, or regional causes (or implied causes) behind voters' choices were assumed to have been fixed conditions, at least in the near-term. In other words, the occupation, education, income, ideology, party, and region of nearly every treaty voter would have remained the same on September 20 as they were six months earlier.

The political sociology of the treaty vote may have been stable, but change, not stability, is what warrants attention in this matter. This dramatic shift in the electorate's opinion between May and September cannot be explained by the standard, fixed structural elements. As the referendum approached, and the debate grew more intense, voters began to weigh the pros and cons of the treaty seriously. The explanation for their change of heart lies with an examination of the issues, attitudes, and values that eventually counted for Maastricht voters.

The Decisive Issues in the Maastricht Vote

Foreign-policy issues, in particular, shaped the Maastricht vote, but this proposition is controversial in two senses. First, it runs counter to the conventional political-sociological interpretations popular at the time. Second, its stress of foreign concerns over domestic ones is exceptional. Foreign-policy issues may center on nationalistic feelings, with perhaps an emotional fear of "outsiders," or they may involve a calmer assessment of international situations and commitments. Fortunately, the 1995 French National Election Study (FNES), based on a large national sample (N = 4078), posed many questions relevant to voting and political agendas. Table 1 relates vote choice on the treaty to the foreign-policy issue of "building the European Union." Among those who considered the EU "most important," 77 percent claimed to have voted to ratify the [End Page 70] treaty, in contrast to only 32 percent among those rating it "least important." Those who make Europe a priority seem likely to champion projects that further integration.8

The Relationship of the Importance of Europe to the Treaty Vote
Click for larger view
Table 1
The Relationship of the Importance of Europe to the Treaty Vote
The Relationship of Attitude toward Germany to the Treaty Vote
Click for larger view
Table 2
The Relationship of Attitude toward Germany to the Treaty Vote

Regarding more specific nationalistic attitudes, Table 2 shows that indviduals who were "very positive" toward Germany were highly likely to approve the treaty (75 percent), whereas those "very negative" were not (37 percent). Anti-German sentiment could have been behind many "non" votes. France dominated Europe from about 1951 to 1990, when Germany reunified. Mitterrand was keenly aware that French leadership of the European community depended on holding Germany in check, which became increasingly difficult after German unification. French aversion to German direction in Europe has deep roots in the foreign [End Page 71] office. According to a recent critic, French Foreign Minister Hubert Védrine held the view "that while the European Union is founded on a Franco-German partnership of equals, France—as Orwell might say—is more equal than Germany."9

Even if ulterior motives decided the Maastricht vote, foreign policy was not the only possible prime mover. Possible alternatives were economic conditions, domestic politics, social-psychological group identification, racial and ethnic considerations, and general openness to social change. The FNES included items touching on all of these issues; the measures used are listed in Appendix B. Table 3 is a correlation (Pearson's r) matrix relating fifteen issues to the treaty vote. Nine of the fifteen are statistically significant at the .05 level, and almost all of the signs are in the expected direction. Foreign-policy matters correlate relatively well with the vote; as cases in point, for "Importance of Building the European Union," the r = .35 and for "Attitude toward Germany," the r = .21. However, rival perspectives also register comparable coefficients—for "Sympathy with [Lionel] Jospin," the r = .26 and for "Too Many Immigrants," the r = .27. However, reliable comparative judgments vis-à-vis the foreign-policy paradigm require multivariate analysis.

Different Issue Models of the Maastricht Vote

Use of multivariate analysis to explore support for five issue paradigms— economic voting, domestic politics, European partisan, anti- immigrant, and openness to change—permits control of the key sociological variables of income, education, class, religion, and gender. Such control allows us to rule out any spurious association between issues and the treaty vote, as a result of common prior influences from these variables. Further, control on political interest and first-round vote participation increases the reliability of these treaty-vote recall data. With regard to estimation procedures, both ordinary least squares (OLS) and logit regressions are presented. Although logit is generally more efficient than OLS, given the dichotomous dependent variable, the OLS results are common coin and more readily interpretable. In any case, because of the relative evenness of the distribution on the dependent variable, [End Page 72] the pattern of statistical and substantive significance is about the same with either technique (see tables 4 and 5).

The economic-voting model has been applied with some success to the study of French national elections. The core argument is that French voters reward or punish the government according to evaluations of economic conditions. Applied to the referendum, the hypothesis would be that those who saw the treaty as bringing economic good news would vote for it and those who foresaw economic bad news would vote against it. Column 1 (tables 4 and 5) partially supports that hypothesis. Pocketbook issues—unemployment or anxiety about household finances—do not relate to the treaty vote. This null finding is in line with the economic-voting literature, which consistently finds pocketbook issues of little consequence for voters. However, that voters optimistic about the economic future of France tended to be more favorable to the treaty is compatible with the general conclusion that French voters are more prospective, as opposed to retrospective, on economic issues.10

A rival argument maintains that the Maastricht vote reflected domestic politics. Indeed, as already discussed, Mitterrand's campaign for the referendum was founded on the partisan agenda of promoting the Socialist cause and embarrassing the political right. His calculation was that a well-played party game could confine voting to national affinities rather than European concerns. Some hard scientific evidence supports this perspective. In 1991, a survey that asked the French public to rank the importance of election types—presidential, parliamentary, municipal, cantonal, regional, and European—placed European elections last. In a study of vote determinants in French legislative and European elections, Lewis-Beck and Nadeau found that domestic political variables alone could account sufficiently for the results. That is, French voters did not take European issues into account even in European elections, [End Page 73]

Correlation Matrix of Explanatory Variables
Click for larger view
Table 3
Correlation Matrix of Explanatory Variables
[End Page 74]
Competing Explanatory Models, OLS Results
Click for larger view
Table 4
Competing Explanatory Models, OLS Results
[Begin Page 77]
Competing Explanatory Models, Logit Results
Click for larger view
Table 5
Competing Explanatory Models, Logit Results
[Begin Page 79]

relying instead on traditional party cues. As the second column of tables 4 and 5 shows, the Maastricht vote confirms this finding. Those placing themselves ideologically on the right were more likely to vote against the treaty, whereas those sympathetic to Jospin, the presidential candidate for the Socialists (and, to a lesser degree, Édouard Balladur, the stand-in for the UDF, which split from the RPR over Maastricht) were more likely to vote for it.11

The third perspective argues that the key to the Maastricht outcome was voters' social-psychological identification with Europe. For certain individuals, the question of European citizenhip is not a short-term issue but an enduring attitude that trumps French citizenship. Long-standing membership in the European Union could tap into "feelings of belonging to some unit greater than the nation-state," a "supranational identity." European identification is not unlike party identification, but its political compass points more toward an international allegiance than a national one. The FNES asked respondents whether they felt "only French," "more French than European," "as much French as European," "more European than French," or "only European." Column 3 (tables 4 and 5) reveals that the variable registers strong statistical significance.12

The anti-immigrant paradigm takes its inspiration from earlier work on ethnocentrism in France. In her analysis of the 1988 FNES, Mayer found responses to the following four items formed a unified scale: "Jews have too much power in France"; "It is only fair for Muslims in France to have mosques to practice their religion"; "Nowadays we do not feel as at home as we used to"; and "There are too many immigrants in France." Scores on this ethnocentrism scale clearly relate to party preference. As Mayer comments, "While people who identify with the NF [Front National] beat all the records for ethnocentrism, those on the extreme left come in second, though at a good distance behind." The evidence surrounding the Maastrict vote suggests that ethnocentrism tends to act as a break on European integration. The item in the 1995 FNES—"There are too many immigrants in France"—is but one component of the ethnocentrism index. [End Page 79] Nevertheless, anti-immigrant sentiment by itself emerges as having been a significant factor against Maastricht.13

The final paradigm, openness to social change, is based on the unexceptional but solid argument that those who were disposed to change in general, or to reform in particular, were more likely to favor the treaty. However, those who professed openness to change in society were actually less likely to support the treaty, as the negative coefficient in column 5 (tables 4 and 5) shows. Moreover, this negative sign does not appear to be a product of collinearity, since it exists at the bivariate level (see Table 3). This curious finding remains unexplained (even after checking the direction of the coding to ensure no mistakes). With respect to attitude toward reform, the finding is not unexpected. Individuals more positive toward reform were slightly more apt to vote for the treaty and its reforms, but the significance level is marginal (the logit coefficient is barely three times its standard error, unimpressive given the sample size). The notion that a broad disposition toward change results meaningfully in support for any and all policy change, such as Maastricht, is, in the end, unconfirmed.

The survey items assessed foreign policy at three levels of generality: France versus the world, France versus Europe, and France versus Germany. All three areas touch on the question of sovereignty, but in different contexts. Thus, attitudes at each level of vision would seem to contribute independently to a vote for or against the treaty, as multivariate analysis demonstrates (column 6, tables 4 and 5). Citizens who were not proud of France's place in the world, appreciated the importance of European integration, and valued Germany's presence were significantly more likely to vote "oui."

To answer the question of how well the foreign-policy model explains the vote, compared to the other models, goodness-of-fit statistics should be evaluated. With regard to the OLS results, the standard error of estimate (SEE) of the foreign-policy model has a lower value than any of the others. Likewise, its adjusted R2 is clearly larger. Indeed, this fit is about one-quarter larger than that of its nearest rival, the European partisan model, which arguably partakes of a similar theoretical base (more on that point below). [End Page 80] The logit results are parallel. The "percent correct" predicted for the foreign-policy model is 71 percent, which shows more accuracy than the other models. Also, it records a pseudo-R2 larger than the others. In fact, this logit fit is the only one to break the .10 barrier. Once such comparisons are made, the foreign-policy model clearly outdistances the rival explanations. So far as the Maastricht referendum is concerned, unlike other national votes in France, the implication is that the principal issues were foreign not domestic.

A Supermodel of Issue Voting on the Maastricht Treaty

Issue voting in the Maastricht referendum was considerable, but the final verdict on whether foreign-policy issues trumped others depends on direct statistical competition in one model. The many issue items complicate testing with threats of collinearity (for example, the negative coefficient of the variable, France in the world, already appears suspicious [column 6, tables 4 and 5]). Including all of these issue items blindly in a single equation makes little sense. As an alternative strategy, this article includes, along with the core political-sociology variables, the most statistically significant variable from each paradigm (that is, the variable yielding the highest absolute value from the logit coefficient divided by its standard error). Other variables were also added to the central paradigms of domestic politics and foreign policy—sympathy for other leading candidates (Balladur, Chirac, and Jean-Marie Le Pen), to give national political forces full play, and attitude toward Germany (though not France in the world, because of collinearity concerns). These supplementary variables seem prudent given the centrality of these two paradigms in the treaty-vote debate.

For clarity, the variables can be grouped into three sets: sociological dispositions (gender, income, religion, participation, education, rank in society, and interest in politics), political anchors (sympathy for Balladur, Chirac, Jospin, Le Pen, and left–right political loyalty), and policy incentives (attitude toward Germany, importance of building the European Union, identification with France/Europe, attitude toward immigrants, attitude toward change, and prospective evaluation of the economy). Table 6 presents the OLS and logit estimates for this "supermodel," which fits the data well, given the dichotomous nature of the dependent variable (as a base of comparison, consider that the R-squared are [End Page 81] in the range typical for survey-based models that attempt to explain national vote turnout, measured as a recalled yes–no).

Supermodel of Voter Decision on the Maastricht Treaty
Click for larger view
Table 6
Supermodel of Voter Decision on the Maastricht Treaty

The more rigorous analysis brings considerable clarification. The "openness to change" paradigm continues to falter, and should probably be dropped, at least in the form argued and presented herein. Economic voting makes a contribution, but only marginally. For example, taking the significant OLS coefficient as an estimate in a linear-probability model, it suggests that anyone who saw the economic future as "better" was only 3 percent more [End Page 82] likely to vote "oui," compared to anyone who saw it as "the same." Surprisingly, the domestic-politics model is also weak. Ideology—as measured by left–right affiliation and sympathy for Jospin, Balladur, and Le Pen—are barely significant. For example, the coefficient of the left–right variable is just twice its standard error (β/S.E. = .018/.009 = 2), not a large value for this sample size. In addition, linear-probability translation of this slope coefficient suggests only mild effects. For instance, individuals who had moderately strong sympathy for Jospin (a score of 75 on the scale) would only be 7.5 percent more likely to favor the treaty than individuals who were indifferent toward Jospin (score of 50 on the scale).

The domestic-politics model fares badly, but the foreign- policy model does well. The two variables, attitude toward Germany and importance of Europe, are easily significant. Furthermore, effects appear strong, even after applying the linear- probability interpretation with some caution at the tail values of the scales. A voter who was "very positive" toward Germany was 20 percent more likely to vote "oui" than a "very negative" voter. A voter who rated the importance of building the European Union as "8" appears 49 percent more likely to vote "yes" than a voter who rated it as "4." Because of the implicit curvilinearity of the probability for this vote, these values are undoubtedly exaggerated. Nevertheless, as a baseline method for comparison across the coefficients, they indicate that the foreign-policy model dominates.

Interestingly, the OLS slope coefficient of the European- partisan model suggests that citizens who saw themselves as "more European than French" were about 24 percent more likely to vote "yes" than those who saw themselves as "only French." At first blush, European partisanship might seem to be another foreign-policy issue. However, it is not that simple. First, it is statistically independent of the variables, "attitude toward Germany" and "importance of building Europe," easily achieving significance in the face of their control. In particular, its OLS coefficient is more than six times its standard error, as is its logit coefficient (see Table 6). Second, it theoretically composes a different dimension, tapping the most profound feelings of national (or international) identity. In that sense, it addresses values rather than issues, although as a social-psychological identification, it could certainly [End Page 83] be expected to shape responses concerning a host of particularly European controversies.

In opposition to the European partisan is the anti-European partisan, who can be coded as anti-immigrant. Those opposed to immigration were as likely to vote against the treaty as were those who strongly identified with France nationalistically. These findings may not be surprising per se. But what is surprising is their statistical and theoretical independence. Contrary to supposition, French nationalism is not merely a disguise for anti-immigrant feeling. Instead, it connects, at least in part, with older, broader, feelings about the French patrie and its place in the world.

Hence, voters did not respond heavily to the usual variables of sociological disposition. Instead, they responded to policy incentives, almost entirely foreign ones. This feature distinguishes the treaty referendum from other popular votes, where the issues are invariably national or local. These diverse foreign-policy concerns are united at their source—worry about sovereignty. However, the multivariate analysis demonstrates that these concerns manifested themselves in two types of expression—specific issues, like building Europe and negativity toward Germany, and general values, like European identification and anti-immigrant feeling.

Although the two sets of foreign-policy concerns overlap to some extent, they remain separable. Opinions about foreign- policy issues—which tend to involve more immediate considerations about the mechanics of constructing Europe, and in which Germany is the key opponent—almost certainly showed the biggest shifts throughout the course of the Maastricht campaign. Longer-term values may have shifted to some extent as well, but, because they represented a more general European orientation, they inclined voters to adopt more consistent positions. Moreover, these values pulled in two directions. An identification with Europe in general moved voters to further the European project in specific political ways. In contrast, an extreme displeasure about the presence of immigrants, or ethnocentrism, moved them to support policies that closed the borders.

* * *

Conventional explanations of the Maastricht vote indicate that the "petit oui" was a product of the usual political sociology of French elections—socioeconomic status, partisanship, and regional forces. The bourgeoisie and the ideological left wanted the treaty; workers, [End Page 84] peasants, the poor, and the ideological right did not. The more Catholic regions showed greater support. All of these characterizations have some truth, but they are more descriptive than explanatory. They cannot account for the enormous change in opinion that occurred during the campaign period.

What swung the vote were foreign-policy concerns—short-term issues and long-term values. Domestic politics mattered to a degree; the Socialist-led push for Maastricht left its mark. However, that campaign was dwarfed by the French electorate's worries about sovereignty. The weight that such worries carried in the Maastricht campaign should come as no surprise, since the treaty aimed precisely at altering the parameters of French sovereignty. The EMU proposed by the treaty "involved major pooling of sovereignty. . . . It would dilute sovereignty over monetary policy by removing the national government's ability to revalue national currencies, manipulate interest rates, and use deficit finance to stimulate growth." In other words, powerful tools that the French state had used for regulating the business cycle were to be taken from its hands. Hence, the vote reflected a "profound bifurcation in the French electorate, one that transcends the Maastricht Treaty and that encompasses conflicting views of the national state in the modern integrative environment."14

The dynamics of the 1992 Maastricht Treaty referendum provide insight about that the 2005 European Union Constitution referendum, which, unlike its predecessor, failed, though both exhibit a similar equilibrium of forces in their margins of victory. On the basis of the analysis in this study, the 2005 result was unremarkable. Before that ballot, the question of Europe was not on the agenda during regular elections. For example, only 2 percent of the respondents in the national study before the 2002 presidential election viewed European unity as the primary issue, and only 9 percent rated it as first, second, third, or fourth. It ranked tenth in importance, at the bottom of the list.

The political flurry of the 2005 referendum, for which President Chirac aggressively campaigned, returned it to the top of the agenda. Open debate on its passage broke the fragile, multi- partisan, pro-European elite consensus that originally obtained. Some of the voters welcomed a European opening, but more of [End Page 85] them feared it. On balance, the referendum presented the French people with too many foreign-policy risks. They voted to keep Europe at arms length. Although most of them lacked details on the bureaucratic intrusions of the EU into French life, they knew enough to refuse any further ones. This referendum, like that of Maastricht, touched a deep chord in the psyche of French voters concerning the sovereignty question.15

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.