i.e. April 21, 2002, the day of the first round of the 2002 French presidential election, when Jean-Marie Le Pen overtook Prime Minister Lionel Jospin—the Socialist party candidate—to finish in second place, thereby moving on to the second round two Sundays later to square off against incumbent President Jacques Chirac. The 21 avril 2002 was the darkest day in recent French political history. Practically no one saw it coming (though I did; more on that below). The shock of Jospin’s elimination and Le Pen’s making it to the second round cannot be understated, and not only on the left. Even Chirac and the mainstream right seemed dazed. They did not exult. As for Le Pen, he was manifestly not expecting it. While there were tearful celebrations that night at Front National HQ Le Pen himself was subdued, which was out of character for him and particularly given the enormity of his success.
The outcome of the election was a foregone conclusion. There was a union sacrée of the left and parliamentary right, with voters of the left casting ballots en masse for Chirac in the second round to bar the route to Le Pen (who didn’t have a prayer of winning in any case). So Chirac won reelection and with 82% of the vote. Chirac did not campaign during the two weeks between the two rounds. There were no rallies and no debate with Le Pen. What was supposed to have been a closely fought election between Chirac and Jospin turned into a Bonaparte-like plebiscite, and with the entire Socialist party leadership calling for a vote for Chirac. (An American equivalent was the 1991 gubernatorial election in Louisiana, when David Duke, running as a Republican, made it to the run-off, causing the national GOP to endorse the corrupt Democratic incumbent Edwin Edwards). So in 2002 the French people were in effect deprived of a presidential election. Chirac was given another five years in office on a silver platter, and with his reelection paving the way—in a coattail effect—to a large victory of the right in the legislative elections the following month (which would most likely not have happened had Jospin faced off against Chirac in the second round and won, which he may well have).
There has been much talk over the past year, in view of Marine Le Pen’s rise, of a repeat of the 21 avril in the upcoming presidential election—of Marine LP overtaking the Socialist candidate to face off against President Sarkozy—or of a 21 avril à l’envers, i.e. a reverse 21 avril, where the candidate of the mainstream right is eliminated—in this case Sarkozy—and with Marine LP going head-to-head with the candidate of the left in the second round. The former prospect is now pretty much excluded given François Hollande’s consistently high poll numbers. But the latter possibility, of Sarkozy being eliminated, is very much on the table in view of his weak poll numbers and high unpopularity. This is, in any case, the conventional wisdom in the political class, media, and among the public in general. Just about everyone who weighs in on the matter believes that a repeat of the 21 avril is at least a possibility this coming April 22nd, if not an outright danger.
I have been arguing strongly against this, though. I have been insisting for years—and particularly over the past one—that there is almost no chance of a repeat of the 21 avril, that Marine LP will not make it to the second round in the upcoming vote. The reason why I insist on this has to do with an analysis of what actually happened on April 21, 2002, of why the infréquantable candidate of the extreme right made it to the second round and the major candidate of the mainstream left was eliminated—and why this will most certainly not be repeated in the future. There is a fundamental misunderstanding, or lack of understanding, in France as to why the 21 avril happened, including among political professionals, political scientists, and pollsters. I will set the record straight.
I will say straight out that the 21 avril was an accident. It should have never happened. And it would not have happened had the election been in America (more on that below). The situation in early 2002 was thus: France had gone through five years of cohabitation government, with a President of the Republic on the right—Chirac, who had decisively defeated Jospin the 1995 presidential election—and a government of the left led by PM Jospin, that had come to power in the early legislative elections called by Chirac in 1997 and which blew up in his face (in a political miscalculation of epic proportions). Both men were actually popular during the years following the ’97 elections, as the economy was doing well and unemployment declining. Jospin’s gauche plurielle government was well-regarded and not just by French public opinion but also by normally skeptical free-marketeers abroad, including The Economist magazine, which had editorially attacked Jospin as a socialist dinosaur when he took office in June ’97 but changed its mind later that year. But when the economy started to slow down after three good years, Jospin’s popularity dropped in tandem, particularly as delocalizations and stock market-driven layoffs increased. And then there was Jospin’s famous “l’Etat ne peut pas tout” (“the state can’t do everything”), which is what he imprudently told protesting, laid-off workers outside a Michelin plant in 2000, who demanded to know why he, as Prime Minister—and of the left no less—, could not intervene with the company and save their jobs. Jospin’s words were a statement of fact but given the way things work in France—in terms of discourse and expectations of citizens from the state—it was not a politic thing for him to say, particularly at that time and place. And, above all, as the then uncontested leader of the French left.
So by the time the campaign for the 2002 presidential election began, there was discontent among voters of the left with Jospin and his government. The base was unhappy. There was also an exceptionally high number of candidates that year who qualified for the ballot: 16, compared to nine in both 1995 and 1988. In the 1995 election, there were three other candidates of the left in addition to Jospin: Robert Hue of the Communists, the Green candidate, and the perennial candidate of the flaky Trotskyist sect Lutte Ouvrière, the pasionaria Arlette Laguiller (whom the majority of Frenchmen and women found charming and adorable—her personal popularity ratings were always high—, despite her amusingly archaic Marxist rhetoric, which practically no one paid any attention to). The presence of candidates from these formations was normal and there was minimal competition between them and the Socialist candidate for votes, as their electorates were distinct and the first two—of the PCF and Les Verts—could be counted upon to rally to the Socialist in the second round against the right.
In 2002, though, there were an additional four left-wing candidates and which complicated matters for Jospin’s first round campaign. Two of the new candidates were Trotskyists—for a total of three Trots, which was certainly a world record for a presidential election anywhere—, one being the charismatic, well-spoken Olivier Besancenot, who was 27 but looked younger, and with distinct appeal to younger voters (and particularly women in the 18-22 cohort). The Green—the outspoken, media savvy Noël Mamère—was a stronger candidate than the one in 1995. But two candidates of the left ended up posing a problem for Jospin (though this was not perceived until after the fact): the left-wing souverainiste Jean-Pierre Chevènement and Christiane Taubira, associated with center-left Parti Radical de Gauche (PRG). Chevènement was a political heavyweight and a longtime leader of the PS left-wing before forming his own party in 1993. But he remained allied with the Socialists and was Jospin’s Minister of Interior before quitting in 2000 over major policy differences. Normally Chevènement would have supported the Socialist candidate—and particularly Jospin, with whom he had collaborated politically since the 1970s—and despite his euroscepticism, but he was so unhappy with the Jospin government—and convinced that he himself had the stature to be President of the Republic (which he did)—that he decided to throw his hat into the ring. As for Taubira—a black deputy from French Guiana—, her candidacy showed how fractious the governing coalition had become, as the PRG was a small party permanently allied with the Socialists and that had supported the Socialist candidate outright in every previous presidential election save one.
So the final phase of the 2002 campaign had a circus atmosphere—as all 16 candidates were legally entitled to equal coverage on television and radio, and equal ad time in the final two weeks—, though it was absolutely assumed by almost all that Chirac and Jospin would finish in first and second place respectively, and move on to the second round, where a tight race was expected. This is what the polls were projecting in any case. Furthermore, Jean-Marie Le Pen—who had won 15% of the vote in the 1995 first round—had suffered a major setback with the Front National’s internal crisis and split in late ’98-early ’99, that caused the FN’s vote in the ’99 European Parliament elections to plummet. For the first time since the mid-80s there was relatively little hand-wringing in the usual quarters (the left, intellectuals, media pundits, etc) over Le Pen and the FN. During the winter of 2002 Le Pen’s poll numbers were lower than usual. He did not appear to be a threat. So going into the first round of the election, left voters who would normally have voted Socialist from the outset but who were unhappy with Jospin, or who simply didn’t feel like voting for him in the first round—Jospin, though solid and serious, was not an exceptionally exciting, charismatic politician—, dispersed their votes, casting ballots for Chevènement, Taubira, Mamère, or Besancenot. No one expected these four to do any better than they did (see bar graph below); as just about everyone assumed that Jospin would go through to round two, voting for one of the others was seen as a no cost way of sending a message, expressing discontent, or simply amusing oneself. Some voted for Chevènement because they liked his militant républicain discourse—there is no equivalent of Jean-Pierre Chevènement in American politics, of a man of the left who wraps himself in the flag—, others voted for Taubira in a symbolic gesture, as she was the first ever presidential candidate of color to qualify for the ballot. The important point here is that the Chevènement and Taubira voters—who constituted almost 8% of the vote—would have voted for Jospin almost to a man or woman had their candidate not been on the ballot, and practically all—along with the Mamère and Hue voters, plus the majority of Trot voters—intended to vote for Jospin in the second round against Chirac.
In the post-mortem analyses after the first round—and in the mutual recriminations and finger-pointing—, the principal explanation of Jospin’s failure focused precisely on the plethora of candidates and dispersal of left votes. Jospin himself never forgave Chevènement for having run (though in view of Chevenèment’s stature, discourse, and programmatic differences with the Socialists, his candidacy was perfectly legitimate). And while Taubira’s candidacy may have been symbolic, she had a particular message, so why not? Political analysts in the media and academia also dwelled on Jospin’s mediocre, uninspired campaign, his numerous missteps, and projecting himself into a second round face-off with Chirac before solidifying his base, as well as Chirac’s seizing of the insécurité issue (i.e. criminality, law and order), which became the centerpiece of his campaign—and a club with which to bash the Jospin government, under whose watch crime had supposedly risen—, but that also contributed to Le Pen’s late surge. Insécurité is one of the FN’s fetish issues—linked to immigration, of course, as a disproportionate number of muggers and louts are of immigrant origin (“immigrant” in France being a code word for North Africans and blacks)—, so whenever the media and politicians focus on it, the FN’s and Le Pen’s fortunes rise accordingly. As for the reaction of left voters who had “dispersed” their votes, there was huge regret that their failure to vote Jospin had enabled Le Pen to overtake him. This was reported in the press and I heard it from numerous people—including quite a few friends and acquaintances—over the subsequent period. The collective reaction of these voters—and particularly those who voted Chevènement and Taubira—was “If only I had known!” (si on avait su !).
If only they had known… The near totality of the French electorate went to the polls on the 21 avril with no idea that Le Pen could overtake Jospin. When the result was announced at 8 PM, the country was collectively blindsided. The question was WHY the voters—not to mention the politicians and pundits—did not see it coming. Well, I did. I had been worried about Le Pen overtaking Jospin for four days prior to the vote. This was on account of my unique American perspective. Voilà, je m’explique.
Until February 2002 there had been a ban on the publishing of public opinion polls a week prior to an election. With the modification of the law, polls could now be published up to two days before the vote. As there is usually a lag of three or four days between the conducting of the poll and its publication—after the data has been processed and weighted—this means that movements in voting intention during the final five days of a campaign cannot be tracked—or if polling institutes do so, they cannot legally announce the results in France (though the results can be published in the Swiss and Belgian press, whose web sites now have heavy traffic on election day). In the 2002 election the final poll was announced on Wednesday, April 17th, based on numbers collected over the previous weekend. It had Chirac in first, Jospin in second at 18%—which was a drop for him—and Le Pen in third at 14%. This was Le Pen’s highest poll number of the campaign. When I heard the numbers on France Inter that day my alarm bells went off, based on one aspect of my American way of thinking.
In the US, public opinion surveys are based on random samples of the population and with a statistical margin of error, which is normally 3% for samples of around 1,000. Polling methodology in France is different, based on quotas of the population by profession. As the statistical model for this method does not have a confidence interval, no margin of error is given, though there is in fact an implicit one and of around 3%. So when I heard the final poll numbers, I thought that Jospin could in fact be at 21% or 15%, and Le Pen at 11% or 17%. In other words, the two were inside the margin of error and with Jospin dropping and Le Pen surging. My alarm bells went off again the day before the first round, when the lead story on the television news was a fait divers in Tours, where an elderly man was assaulted by two young thugs, who burned down his little house before taking flight. In view of the Le Pen surge that I perceived—though which the media had not—, this was the kind of story tailor-made to add another point or two to his vote (and after the 21 avril a number of analysts did think that it had had that effect). On the day of vote I was quite nervous about Le Pen and expressed it to a colleague—a far-left Besancenot voter—over the phone, who dismissed my concerns. When France 2’s David Pujadas said in the hour preceding the 8 PM projection that there was going to be a “grosse surprise” and that would “faire couler beaucoup d’encre” (over which a lot of ink would be spilled), I knew that Le Pen had knocked off Jospin.
I mentioned above that such an outcome would have never happened in America. This is for the simple reason that there are no restrictions on the publication of polls in the final days of a US campaign. Moreover, pollsters in the US have introduced tracking polls in the final phase of presidential campaigns, which have smaller samples—and therefore a higher margin of error—but interview the same persons over time, thereby detecting late shifts in opinion. In the cliffhanger 2000 election between Gore and Bush, politicized Americans were riveted to the daily poll results in the final weeks of the campaign, and particularly the tracking polls in the final days (including the final numbers in the early hours of election day). Likewise in 2004 with Bush and Kerry. If polling-wise the French had done things à l’américaine—with tracking polls and without all the legal restrictions and regulations—, the banner headline in the news all day Saturday and on Sunday morning would have been “Jospin et Le Pen à l’égalité dans les sondages!“. The screaming headline in the weekend Libération would have likely been “Danger Le Pen!!” With this knowledge, the panic on the left would have been such that those intending to “disperse” their votes would have shifted massively to Jospin. This is a certainty. Jospin would have thus not only overtaken Le Pen—who edged him by a mere 0.7% of the vote—but possibly Chirac as well. Had Jospin finished ahead of Chirac, he would have gone into the second round with a head of steam and no doubt won—and with the Gauche plurielle logically besting the new UMP in the June legislatives—, as his approval ratings in the polls were already higher than Chirac’s to begin with: In the April 2002 SOFRES/Figaro-Magazine baromètre measuring the president and PM’s “cote de confiance,” published prior to the first round, Chirac was at 40-58 positive-negative, whereas Jospin was at 47-51 (their March 2002 numbers were 43-54 and 48-48 respectively); in fact, Jospin had, since taking office in June 1997, never dropped below 50% prior to March ’02 in the SOFRES/Fig-Mag baromètre, whereas Chirac had, with the exception of the last three months of 2001—in the wake of 9/11 and France’s participation in the US-led operation in Afghanistan (overwhelmingly supported by French public opinion)—, been below 50% for a full year prior to the 21 avril. Given Jospin’s consistently higher approval numbers, he would have likely beaten Chirac in the second round even if he had finished in second place in the first. And the rest would have been history.
One other thing. SOFRES had the Socialist Party’s positive/negative rating (bonne/mauvaise opinion) at 51-41 in April 2002, whereas the the RPR’s was at 41-50 and the UDF’s at 35-53. Moreover, the PS’s positive number had dipped below 50% only once from 1997 to 2002 in the baromètre, whereas the RPR’s had gone above 40% only three times during that period—the UDF’s numbers were largely the same, when not worse—and with the RPR not having topped 50% since August 1995. Ergo, the Socialists were viewed more positively by the electorate in 2002 than was the right-wing opposition; in view of the discrepancy here, Jospin would have thus normally bested Chirac.
Conclusion: the 21 avril happened for one sole reason, which was the absence of late polls that would have provided critical information to voters and helped them cast their ballots strategically. But so far as I know, I am the only person who had this analysis. I was continually struck over the days, weeks, and years following the 21 avril that absolutely no political analyst brought up the French law on polls as a factor. For the anecdote, in 2003 I attended a small round-table discussion on the 2002 election cycle at the office of a major Parisian intellectual revue—whose political orientation and general sensibility I completely share—, which was conducted by two specialists: a well-known political science specialist of French electoral politics and the then political director of the LH2 polling institute. Neither mentioned the absence of late polls. I wasn’t able to get a word in during the discussion but afterward offered my analysis to the pollster (who is well-known and with a high media profile). After listening to me his response was: c’est très intéressant ce que vous dites, vous avez raison. It hadn’t occurred to him.
But last-minute polls or not, there won’t be a repeat of the 21 avril, as voters now go to the bureaux de vote with that experience in mind. There was no danger of a 21 avril repeat in the 2007 election—Sarkozy and Ségolène Royal were solidly ahead of the pack—but, as mentioned at the beginning, the prospect now exists for it with Sarkozy and Marine Le Pen. Polls last month had Sarkozy in the low-mid 20s and MLP in the high teens, separated by five to seven points. The latest IPSOS poll has the gap widening, but that’s happened before. If in the final week of the campaign the gap between the two is too close for comfort—if there is a perceptible danger of a 21 avril à l’envers—, Sarko-disappointed UMP voters contemplating defection to another candidate, such as François Bayrou or even MLP herself, will “come home” and cast their ballots for Sarko, to avoid a second round face-off between Hollande and MLP. If the latter occurs, Hollande will win with 65 to 75% of the vote. A Sarkozy elimination—which would be the first ever first round elimination of the parliamentary right in a presidential election in the Fifth Republic—followed by an Hollande landslide would all but guarantee a large victory of the left in the June legislative elections. The debacle for the right would be total. Voters of the right—who remain, malgré tout, more numerous than voters of the left—won’t risk it. So it won’t happen.
There is a possibility that this may be a moot subject, as it is not entirely certain that Marine Le Pen will even qualify for the ballot. This will be the subject of a follow-up post in the coming days.