Category Archives: Swings, trends and political evolution

2007: Jean-Marie Le Pen’s collapse

On April 21, 2002 the far-right’s standard bearer, Jean-Marie Le Pen, had made history by placing second and qualifying for the runoff with 16.86% of the vote. It was an historic night for the French far-right, which had won its best result in its existence. However, five years later, Jean-Marie Le Pen failed to repeat his feat. With 10.44%, he placed a distant fourth and won a result which marked the end of his reign as the patriarch of the far-right in France. Few had expected such a result: in fact, with polls placing him at 12-14% before the vote, most casual observers had expected him to pull at least 16% given how polls underestimated his vote in the past. While the FN has roared back to prominence making talk of its imminent death silly, one of the main lessons of the 2007 presidential election had been Nicolas Sarkozy’s ability to grasp, by the first round, a sizable share of Le Pen’s April 21 voters.

Comparative evolution of the right and FN (2007) by department

Between 38% (Ifop) and 21% (Ipsos) of Jean-Marie Le Pen’s first round voters from 2002 voted for Nicolas Sarkozy by the first round of the 2007 election. The old patriarch kept only between 53% (Ifop) and 64% (Ipsos) of the voters who had made his spectacular feat of April 21, 2002 a reality. After having been deeply ingrained in French politics and society for over 20 years, after having weathered through the crippling split of 1998 with Bruno Mégret and after having resisted well in the 2004 regional elections, how did the FN suffer such a sudden and violent collapse?

The answer lies in the unique personality of Nicolas Sarkozy. The French right, after having tried unsuccessfully to court Le Pen’s electorate in the late 1980s, the right led by Jacques Chirac had cut short its attempts to seduce his electorate. Sarkozy, on the other hand, while not openly embracing or courting the far-right as a political entity, built himself an image as a law-and-order tough on crime populist which was quite different from that of the traditional right, led by Chirac. An ironic image for a man who was as recently as 2002 considered as one of the most liberal (in the French sense) politicians in France, but Sarkozy harboured deep presidential ambitions. Sarkozy’s strategy was to conquer Le Pen’s electorate by reclaiming control over certain themes which had until then been the exclusive property of the FN (ideas such as too much immigration, insecurity in the suburbs and so forth). To counter the old patriarch, his strategy was to show himself as an energetic Interior Minister in touch with reality who “gets stuff done” as opposed to Le Pen, portrayed as an archaic leader with radical positions out of touch with reality. Sarkozy struck at a moment which was perfect. Following Le Pen’s underwhelming performance in the 2002 runoff, an increasing number of FN voters were growing desperate for action and change while harbouring mounting doubts about Le Pen’s ability to conquer power and affect those changes himself.

While the FN criticized Sarkozy’s action as mirages and Le Pen often repeated how voters preferred the original to the copy, the party’s electorate appreciated Sarkozy’s action and positions taken in his role as Interior Minister in the fight against criminality and delinquency. Looking at Ifop polls over the course of campaign, the turning point seems to have been the riots in the Gare du Nord on March 27. Following those incidents, Sarkozy increased with Le Pen 2002 voters from the lows 30s to the high 30s. Ironically, Sarkozy’s announcement on the creation of the Ministry of Immigration and National Identity in early March actually led to a slow slide in his support with Le Pen’s electorate. It seems as if Le Pen’s electorate responds poorly to moves which seem as overt pandering from the UMP but respond far better to circumstances and events (such as the Gare du Nord riots).

Jean-Marie Le Pen lost about 6.42% of his 2002 result in 2007. On the other hand, Sarkozy increased by 6.2% the combined performance of Jacques Chirac, Alain Madelin and Christine Boutin in 2002. It is almost as if both go hand in hand: Le Pen’s lost vote almost all flowed to Sarkozy. At a departmental level, excluding the DOM-TOM and Corsica (because they voted weirdly), the correlation between the FN and the right’s evolution between 2002 and 2007 is very strong at 0.81. Of course, when you include the DOM-TOMs and Corsica, the correlation drops a whole lot to 0.58, but we’re looking at places where voting is very parochial, where the FN is very weak (the DOM-TOMs) or where it grasps a rather unique electorate (Corsica). In metropolitan France, the only main exceptions to the pattern is Corrèze where Sarkozy lost 8.9% of the combined right’s vote in 2002 (Le Pen lost 1.3%). There is thus a striking symmetry between the evolution of the right’s vote with that of the traditional FN map.

However, Nicolas Sarkozy and Jean-Marie Le Pen’s gains and loses were not spread out homogeneously over the country. Where did the FN lose the most, and where did the FN show the strongest resistance? The map below shows the evolution of Jean-Marie Le Pen’s support between 2002 and 2007, drawn up by 1986-2009 constituency. Brown indicate constituencies where Le Pen’s support increased between 2002 and 2007, while varying shades of blue indicates constituencies where his support decline between 2002 and 2007, with darker shades indicating a larger decline.

% change between Jean-Marie Le Pen’s 2007 performance and Jean-Marie Le Pen’s 2002 performance by legislative constituency (1986-2009 redistricting)

The Three/Four Worlds of FN Voters

In general, Jean-Marie Le Pen lost the most support where the FN traditionally does best: the Mediterranean coast, the Garonne valley, Rhône-Alpes, Alsace-Moselle and the greater Parisian basin. From the map, we pick out three or four major traditional FN voters, which progressively morphs into another type.

The first type we can distinguish is a petite bourgeoisie (lower middle-classes), in many cases retirees or small employees, concentrated mostly along the Mediterranean coast. They are not all that poor, but they are not traditionally considered as being part of the more affluent elites, and by their social status as petit bourgeois they are strongly individualist and deeply conservative. The region was too urban and industrialized to be Poujadist in 1956 (except Vaucluse, less industrial, and Poujade’s best department in 1956), but in their concerns the first type is somewhat Poujadist. In Provence, a fair number of these voters may be of pied noir ancestry, but in 2007 it is perhaps a bit ridiculous to say they’re all pieds noirs. Living in a region with a large North African population and with particularly high crime rates (especially Marseille or the Alpes-Maritimes), the FN has always been – since 1984 – very strong with this first type of voters. To an extent, their vote for the FN might be a protest vote, but it is not entirely that and in large part it could be constructed as a conservative vote concerned with North African immigration, insecurity and diametrically opposed to left-wing conceptions of the state and society.

In the Vaucluse, traditionally the FN stronghold by excellence, the FN tradition is born out of historical factors (less industrial, an old reactionary-conservative base, a vibrant Poujadist movement, the Algerian war and the OAS) and of contemporary social factors (an important agricultural sector employing farm workers, a petite bourgeoisie and lower middle-classes).

The first type is strongest on the Mediterranean coast, especially in the Var and Alpes-Maritimes (where it is the most conservative and affluent). A similar type of electorate (let’s call it type 1-bis), lower middle-class and equally concerned with immigration and insecurity, can be found in suburban or exurban communities, especially in the Rhône-Alpes region. It is especially strong in old working-class hinterland, but which has increasingly been transformed into average income middle-class bedroom communities. Unemployment in areas such as Meyzieu (Rhône) or Nord-Isère is not particularly high – in fact it is below average – so it is not the protest vote of poor suburbs with high unemployment – rather it is a conservative vote about immigration and insecurity (they are located close to working-class suburbs with high immigration such as Vénissieux), like in the Var or Alpes-Maritimes. The left-wing roots of these regions have been dropping like flies in recent years, as the contest in places such as Meyzieu becomes increasingly UMP and FN.

In terms of social categories, the first type is largely composed of employees and professions intermédiaires (a blanket term for broadly middle-class people). But these categories, like that of ouvrier, is far from homogeneous. They are all divided by some fairly key schisms in terms of their comparative political attitudes. Employees, generally the second lowest step on the “social ladder” in France behind the broad ouvriers category, are divided between those who work in small businesses (PME-PMI in France, including construction – BTP or agrifood) and in commerce (vendors, cashiers) versus those who work in the public sector (education, health, social services). The FN performs strongly with the first type of employees (small businesses) but performs very poorly with the second category. A similar public-private divide is found with the middle-class categories. Again, the FN performs well with those middle-level employees in the private sector (construction, small businesses, commerce) where the fear of losing their job is pretty big. On the other hand, the FN usually registers its worst results with middle-level public employees including teachers. The FN, like the traditional right, performs well with non-salaried self-employed workers (including the old, stereotypical FN-voting shopkeepers). The FN may use populist quasi-statist rhetoric, but its base often reflects some of the most economically liberal, anti-statist views out there.

The second type is somewhat similar to the first type, but it is less affluent and less urban. We can call this type a rural conservative vote, a phenomenon which is particularly pronounced in Alsace where the FN has performed well in lily-white small towns (often more Protestant than Catholic) with an older population particularly touched by concerns over immigration (which is particularly important in a border region like Alsace and in cities like Mulhouse). While there is a very strong working-class base throughout Alsace, unemployment is very low in some of these areas, so it is hard to see it as a protest vote of economically declining regions. It is, however, because of its rural element, more Poujadist in its orientation than the first type was.

The third type is an old white working-class vote, higher in in communities which concentrate both high unemployment (industrial decline) and proximity to large immigrant communities. The third type is particularly low-income, and it is the most left-leaning of the FN’s three/four types. The third type is important in the Nord-Pas-de-Calais, Moselle, parts of the Haut-Rhin (in the potash mining communities), the Montbéliard area and some regions such as the Yssengelais in the Haute-Loire.

Some people want to make you believe that the FN is strong in all working-class communities, no matter what. I would like them to explain to me why the FN is weak in Carmaux, Decazeville, Saint-Nazaire and even Longwy. The other misconception is that the FN’s working-class vote is made up in large part of old left-wingers or former PCF voters. This is truer now than it was in 1984, but PCF voters are actually those least likely to vote FN. A micro analysis looking at FN votes across working-class areas shows that old PCF strongholds are the least receptive to the the FN. In the past, a good part of the FN’s working-class electorate came from the 30% or so of working-class voters who were traditionally right-wing. Right-wing working-class communities such as Cluses, Saint-Amarin, Freyming-Merlebach or Stiring-Wendel are some of the FN’s rock-ribbed strongholds. The Marxist left never gained a foothold in these regions, and the FN emerged as the main alternative to the right in those regions. Of course, since 1995, there has been an increasing element of gaucho-lepénisme, denoting a certain kind of traditional white working-class left-wing voter who votes for the FN yet often returns to his left-wing roots in the runoff.

The third type’s vote is far more likely to be a protest vote. Unemployment is not necessarily high, but it is on average probably highest in these type of regions than in the first two types. In the stereotypical community of the third type, Hénin-Beaumont, unemployment in 2006 was 13.2% which is actually pretty low by the standards of the mining basin. It is not necessarily a protest vote against politicians of the “UMPS” but a protest vote against unemployment, industrial decline, immigration and insecurity (for those voters, the four problems are closely linked to one another). It is no longer a more well-off conservative vote, and it is the least Poujadist type of electorate.

hybrid of the second type – the rural conservative vote – and the third type – the WWC vote – is a kind of distant, isolated rural or exurban vote which is increasingly a protest vote in not-too-affluent “forgotten communities” against isolation from urban cores, their intellectual “tolerant” elites and exorbitant property prices in urban areas (such as Paris or Lyon). It is property prices and white flight which has pushed this populaire (old, low-income, traditionally working-class or working poor electorate) electorate of working-class tradition though not, in many cases, of unionized large industrial working-class tradition. The departments of the Meuse, Haute-Marne, Marne, Aube, Aisne and other parts of Picardie concentrates a good part of this vote, but it can be observed in the Vexin and the Perche (which is more rural than exurban). Unemployment or immigration is not particularly high (in fact, it is likely below average) but they are still touched by economic problems, criminality and the effects of immigration.

These are what I construe as the three general types of electorate, which are in some cases similar to one another but in other regards are rather different. Their difference can be seen in the reaction of these three types to the Sarkozyst tentation.

The first type reacted the most to the Sarkozyst tentation, as can be seen by Le Pen’s heavily loses along the Mediterranean and in Rhône-Alpes. In the Alpes-Maritimes, where Sarkozy won his best result in France in the first round (43.6%), Le Pen’s loses were heavy and concentrated in the most affluent and conservative regions: -14.26 in Cannes, -12 in Antibes, -11.95 in Cagnes-sur-Mer, between -12.4 and -13.3 in Nice, -11.9 in Menton. In the Var, the results are similar: -12.67 in Fréjus and Saint-Raphaël, -10.4 in Draguignan, -9.5 in Hyères, -6.7 and -7.9 in Toulon’s two constituencies or -10.1 in La Seyne. Loses were equally as heavy in Aubagne (-9.7), Gardanne (-8.7), Orange (-10.8), Carpentras (-9.9), Nîmes-centre (-11.7), Vauvert and Aigues-Mortes (-11.2), Béziers (-12), Sète (-10.6) or Montpellier-sud and Lattes (-10.3). For these more affluent, conservative petit bourgeois, Sarkozy’s rhetoric about work (his appeal to la France qui se lève tôt), immigration and insecurity had a distinct appeal. These voters, not all that much into voting FN for the sake of protest but more for specific reasons, saw Sarkozy as somebody who took up Le Pen’s concerns while being less dangerous, less radical, younger, more realistic and more able to deal with those issues.

The type 1-bis, similar to the first type in terms of preoccupations, also reacted favourably to Sarkozy’s appeal. Le Pen’s loses were heavy in places such as Meyzieu (-13.6), Bourgoin/La-Tour-du-Pin (-10.1), Givors (-10.8) or Romans-sur-Isère (-9.5). It is a similar type of suburban middle-class, concerned with the law-and-order thematic and perhaps Sarkozy’s “la valeur travail” meritocratic rhetoric.

The second type, the rural conservative vote, was the other category which responded most favourably to Nicolas Sarkozy. Jean-Marie Le Pen had done very well in Alsace in 2002, but did relatively poorly in the region in 2007 (he lost 9.89% in Alsace, the highest of any region). Losses were heaviest in Strasbourg Nord (-10.9), Strasbourg Sud (-10), Wissembourg (-10.41), Haguenau (-10.7), Illkirch-Graffenstaden (-11.4), Molsheim (-10.8), Mulhouse Est (-10.5), Altkirch-Thann (-10.2) and Hunigue (-10.1).

The third type did not react as favourably, but a sort of split decision occurred. Le Pen’s loses were very pronounced in traditionally conservative working-class areas such as Forbach (-10.8), Saint-Avold (-9.6), Altkirch-Thann (-10.2) and Yssingeaux/Le Puy Est (-8). They were equally pretty heavy in more left-leaning working-class areas such as Firminy (-10.4), Audincourt (-8.3), Moyeuvre-Grande (-8) and Rombas (-8.4). However, where the FN vote is in large part an old white working-class protest vote against immigration, unemployment and economic decline, Le Pen’s loses were rather small. This is the case in the Nord-Pas-de-Calais: -1.6 in Béthune, -0.1 in Bruay, -2.1 in Liévin, -3 in Marchiennes, -1.8 in Calais, -4.4 in Lens, -4 in Douai though amusingly -6 in Hénin-Beaumont.

The hybrid of the second and third types, the “forgotten communities” vote, was the most resistant to Nicolas Sarkozy’s appeal. In these lower middle-class or populaire exurban or rurban isolated communities of working-class or light industrial tradition, Le Pen resisted well. Loses were below the national average in departments such as the Aisne (-3.9), Meuse (-4), Cher (-4.1), Indre (-3.5), Haute-Marne (-5.4), Marne (-5.5), Vosges (-5.3), Seine-Maritime (-4.8) and of course the Pas-de-Calais (-2.4). Therefore, Nicolas Sarkozy’s rhetoric on law and order, work and authority was better received by traditional FN voters who are more affluent (and urban/suburban) than those who are less affluent or rural, whose vote for the FN is driven heavily by a deeply ingrained anti-establishment streak.

An Ifop analysis of those Le Pen 2002 voters who switched to Sarkozy versus those who did not reveals a similar contrast: those who switched to Sarkozy included 35% of cadres or professions intermédiaires (and 44% of ouvriers and 13% of farmers or shopkeepers), while those who remained loyal to the fold were heavily working-class: 59% ouvrier against 23% of cadres or professions intermédiaires (and 14% of farmers or shopkeepers).

The effects of exurbanization and urban sprawl (or périurbanisation) on the FN’s vote since the 1980s is particularly striking. In 1984, the FN’s vote was heavily concentrated in urban areas reaching peaks in urban or inner suburban areas, but weaker in rural areas. In 2007, the FN vote had been almost entirely drained out of the core of major urban areas such as Paris or Lyon, but was stronger in rural or exurban areas. A spatial analysis of the FN’s vote in 1995, 2002 and 2007 (by Loïc Ravenel at the Université de Besançon) is particularly revealing of the effects of urban sprawl. In 1995, Le Pen received his national average in urban cores and won his best results 25km from the urban core before progressively declining (with a final bump in areas 100km or more from the core). In 2002, Le Pen won about 1% less than average in urban cores and won his best results 35km from the urban core before declining (with another, less pronounced, bump 100km away). Finally, in 2007, Le Pen performed about 2% below average in urban cores, and wins his best results 35-45km away from the urban core while the subsequent decline is less pronounced than in past years (and the 100km away bump is far more pronounced). In 1995, Le Pen was at or above national average in an circle 0 to 55km from the urban core. In 2002, Le Pen was at or above national average in a circle encompassing areas 15 to 65km from the core. In 2007, Le Pen was at or above national average in a circle encompassing areas 15 to 90km from the core.

Le Pen’s decrepitude in 2007 was particularly pronounced in urban and suburban areas, except perhaps in Paris where Le Pen lost only 4.8% – perhaps because the FN’s collapse in Paris was already completed in 2002. There is a general pattern of major decline in support for Le Pen between 2002 and 2007 in most urban areas: Lyon, Marseille, Nice, Montpellier, Perpignan, Toulouse, Rennes, Lorient, Brest, Le Havre, Rouen, Amiens, Lille, Metz or Dijon. The FN electorate in these cities, which is already rather small as it is, is probably composed in large part of more affluent middle-classes who responded favourably, like the first type, to Sarkozy’s appeal on the basis of insecurity, authority, immigration or work.

Le Pen’s Zones of Resistance

In metropolitan France, Jean-Marie Le Pen managed to increase his support in two constituencies: Ault (Vimeu, +1.12%) and Abbeville (Ponthieu, +0.8%) in the Somme, both in the Baie de Somme. The Baie de Somme, a particularly important hunting region (waterfowl or gibier d’eau), had been the CPNT’s candidate (Jean Saint-Josse)’s best constituencies in 2002. The CPNT electorate, especially in Picardie and Normandie, is particularly right-wing, and Saint-Josse voters in those regions transferred in large part to Le Pen in the runoff, some of the only non-Le Pen voters to do so (see here). In 2007, with the CPNT’s collapse, Le Pen was due to capture some of this far-right friendly electorate. CPNT’s vote may also explain, in part, Le Pen’s strong resistance in the rural areas of the Centre, Poitou, Charentes and Aquitaine.

In the Limousin, the heart of Chiraquie, Sarkozy badly underperformed the right’s performance in 2002 with Chirac at its helm, as the favourite-son vote of Chiraquie flowed to Royal or Bayrou, but also Le Pen in a far more modest part. In Chirac’s constituency (Ussel), Sarkozy won 12% less than the combined right in 2002, and Le Pen lost only 0.02% between the two elections. In Bernadette Chirac’s canton (Corrèze), it appears as if Le Pen increased his result by atleast 1% between 2002 and 2007.

Jean-Marie Le Pen held his 2002 result in Corsica (15.7% in 2002, 15.3% in 2007). The FN’s presidential performance in Corsica far surpasses its paltry results in legislative or regional elections, largely because some of the more radical Corsican nationalist voters tend to vote for Le Pen in presidential election, largely because of the xenophobic and ethnonationalist undertones of some of the radical nationalists’ rhetoric on the island.

The Bases of Marine Le Pen’s Surge?

Jean-Marie Le Pen’s collapse in 2007 did not lead to the eradication of the FN as a potent political force. Following the FN’s weak showings in 2008 and 2009, many had started presuming that Nicolas Sarkozy might be able to do to the FN what Mitterrand had done to the PCF: kill the party as a major political actor by eating up its electorate. Following the 2010 regionals, and especially the 2011 cantonals, there is no chance of that happening. Marine Le Pen, the daughter and successor of the patriarch, regularly polls at 15-18% in the run-up to next year’s presidential election and she poses an underlying threat to Nicolas Sarkozy’s candidacy. Marine has been able, pretty spectacularly, to lift a dying (and bankrupt) party from the brink of political extinction and return to its former splendour. What can her father’s collapse in 2007 teach us about the rejuvenation of the FN?

Her father’s stronger resistance in populaire regions such as the NPDC seems to have laid the fertile base for Marine’s restructuring of the FN along the bases of a solidly working-class electorate as exemplified by her Hénin-Beaumont stronghold. Now more than ever, the FN seems to be deeply rooted as the “premier parti ouvrier” (largest party with working-class voters). In 2010, with Marine Le Pen as the FN’s regional candidate, the FN outperformed its 2007 performance in all but one constituency of the NPDC (Valérie Létard’s constituency in Valenciennes).

The third type seems to have remained ever so solidly frontiste and the 2010 regional elections showed the elimination of Sarkozy’s gains from the FN in conservative working-class areas such as the Moselle coal basin. The first type seems to have returned to its traditional far-right roots as well: in 2010, Jean-Marie Le Pen very much outran his 2007 performance in PACA especially in the Var (+6.5), Alpes-Maritimes (+8.5) and Bouches-du-Rhône (+6.7). The corruption cases surrounding Sarkozy’s government and discontent surrounding the government’s criminality record seems to be the main causes for the first type’s sudden reversal between 2007 and 2010-2011.

Type 1-bis has not returned in droves, but it has returned in good part: the FN gained ground in suburban Lyon and Saint-Etienne in 2010. The FN, in the 2010 regional elections, showed surprising vitality in urban and suburban areas (Oise, Aube, Loiret, Seine-et-Marne, Val-d’Oise, Yvelines but also Toulouse, Bordeaux or Lille) where Le Pen had suffered the most in 2007. One element at play here might be unhappy traditional right-wing voters whose vote for the FN is a conservative protest against Sarkozy (we had seen strong FN performances in some upscale areas of Paris in 2010), another element might be the result of the reversal of the type 1-bis electorate back to the FN.

The second type still seems the most reluctant to return to its FN tradition, with the FN still registering disappointing results in the Bas-Rhin in 2010. Interestingly, the FN’s 2010 results, compared to its 2007 performance, was not particularly strong in the areas where it had resisted well in 2007: the Somme, Ardennes, Meuse, Haute-Marne, northern Aisne and the Bray in Seine-Maritime. Other weak FN performances in 2010 vis-a-vis 2007 in departments such as the Ardennes, Haute-Saône, Côte-d’Or seem to be based on a strong local appeal of a favourite-son right-wing candidate (Warsmann, Joyandet, Sauvadet etc), while others might be based on local factors – Sébastien Jumel, the PCF-FG’s top candidate in Haute-Normandie, carried a strong personal appeal in the Bray and Dieppois, which might explain the FN’s relative weakness there in 2010.

The FN’s collapse in the 2007 presidential and legislative election did not mark the beginning of the end for the FN, in fact it only marked a spectacular but ultimately short-lived trough which the party has come out from looking rather strong. However, the differences in Nicolas Sarkozy’s appeal to the FN electorate reveals fascinating details about the different types of FN voters and their reasons for voting FN. Jean-Marie Le Pen’s strong resistance with the third type and the hybrid type revealed, by 2007, that the FN had become solidly encysted in lower-income regions with a populaire tradition, something which explains why Marine Le Pen has structured the FN’s revival around the third type and the hybrid type.

The changing face of the French left (1995-2007)

In the 1995 presidential election, PS candidate Lionel Jospin won 47.36% of the vote in the runoff. In the 2007 presidential election, PS candidate Ségolène Royal won 46.94% of the vote in the runoff. A difference of barely 0.42% between the two results, even if the two elections were a full twelve years apart. The similarity of the results won by the left’s candidate in both runoffs, twelve years apart, makes these two elections particularly interesting for comparison. 2007 is the most recent presidential election, and presidential elections are the best starting points for comparisons because they are the “real elections” where people vote on issues and candidates, not on their usual hatred of the incumbent government. 1995 is, before 2007, the last election in which the runoff was “normal” – that is, a regular right-left contest.

Given that the two candidates in 1995 and 2007 won basically the same percentage nationally, surely their two maps are very similar? Things couldn’t be more different. Look at a basic map of the 1995 and 2007 runoffs and it is shocking how different the maps are considering the national picture is one of similarity.

As the 2012 election approaches, I figured it would be interesting to look at the changing face of the French left in terms of its electoral clientele and the type of voter it has lost in twelve years and the type of voter it has gained in that period. The map below compares the runoff performance of Jospin and Royal by constituency. A constituency shaded in red indicates that it voted more heavily for Royal than Jospin, of course a deeper shade of red indicates that Royal performed far better than Jospin while a lighter shade of red indicates that Royal outperformed Jospin marginally. Conversely, a constituency shaded in blue indicates that it voted for heavily for Jospin than Royal, and again a deeper shade of blue indicates that Jospin did far better than Royal had done. Because overall Jospin did some 0.38% better than Royal (in metropolitan France), the constituencies which are shaded in light blue (cyan) indicate that while Jospin did better than Royal, the margin between his performance in 1995 and her performance in 2007 was smaller than -0.38% – meaning that overall that constituency did not swing towards Royal but trended (swing below national average) towards Royal.

Note: this article uses exit poll data from 1988, 1995, 2002 and 2007 from Ipsos – because they’re the most easily accessible, and because they tend to be quite accurate pollsters. For the 2010 regional elections, data from OpinionWay is used.

The two most shocking aspects of this map are its close correlation with the traditional map of the FN vote and its concentration east of the Le Havre-Valence-Perpignan axis and, on the other hand, the emergence of three major red blocks: Île-de-France and the Parisian basin; the Massif Central and Limousin in the centre; and Brittany, Maine, Anjou and Poitou in the west (Béarn and the Basque County are a smaller but just as significant fourth block of red). I think the first comment about the shockingly close correlation of the map of the left’s decline since 1995 with that of the FN strength east of the old Le Havre-Valence-Perpignan axis is the most important one and the one which merits the most explanations.

The regions east of the Le Havre-Valence-Perpignan axis are the most industrialized areas of France. This is, of course, a pretty reductionist analysis but, in general, the areas west of that axis tend to be less economically marked by heavy industry and more marked, at least historically by agriculture and today by tertiary service-oriented industries. The regions east of the axis certainly include some very rural areas, but most of the large industrial centres of France are here: the coal mines of the Nord-Pas-de-Calais, the petrochemical industry around Le Havre, the working-class hinterland around Rouen and the Seine valley, the coal mines of the Lorrain basin, the steelworks of upper Lorraine, the large petrochemical and shipping installations outside Marseille or the isolated cités cheminotes along the Paris-Lyon railway. These are the most industrial areas, and by consequence the most working-class areas.

Once upon a time, the French left – like most of the European left – was the uncontested party of the working-class and dominated the working-class vote with some 70% of the vote. The tough reality of power for the left, among other factors, has weakened its hold on the working-class vote. From the highs of the post-war years (estimated at 70%), the left has seen its support dwindle pretty drastically with working-class voters (ouvriers) to the point where their voting is no longer markedly different with that of the wider electorate, or only marginally biased in the left’s favour by less than 5%. The main benefactor of the slow decline of the left’s support amongst workers was the FN, whose emergence as a potent political actor beginning in 1983 (Dreux by-election) corresponds to the electorate’s rebuke of the left in the midst of the early-1980s recession. While many serious analyses have indicated that the FN actually gained more amongst the 3 in ten workers who were traditionally right-wing than amongst historically Communist or left-wing working-class voters, the FN still drew at least some of its new support in the mid-1980s from working-class voters who had voted loyally for the PS or PCF in the post-war era.

As time went on, the left’s decripitude with the ouvriers was progressively accentuated. Conversely, while the FN’s presidential vote was stable at 15-16% between 1988 and 2002, there was a pretty dramatic realignment of forces within the FN electorate: the FN progressively lost strength with shopkeepers and the lower middle-classes while gaining quite dramatically with ouvriers. The trend was confirmed in 1995: in the first round, Jospin won 20% of ouvriers against 27% for Jean-Marie Le Pen, 17% for Robert Hue (PCF) and 14% for Chirac. In 1988, Mitterrand had received the support of 40% of ouvriers against 21% for Le Pen, 15% for Lajoinie (PCF) and a paltry 9% for Chirac. Yet, there exists the phenomenon of gaucho-lepenisme – traditionally left-wing voters who vote for Le Pen in the first round but then return to their left-wing roots in the runoff against the traditional right (23% of Le Pen’s first round voters in 1995 voted for Jospin in the runoff). Jospin still won 65% of ouvriers against 35% for Chirac, and a look at his results by constituency or cantons confirms that. The left-wing slant of the vote ouvrier had declined, but it remained, with teachers (67% Jospin) the most solidly left-wing constituency.

The left in power between 1997 and 2002 certainly did not strengthen the left with its old core electorate. In 2002, Jospin won only 15% of ouvriers in that fateful election which shook the left to its core. Le Pen polled 30% with those voters, making them by far his best socio-professional category.

In 2007, Le Pen’s strength with these voters was weakened, though with 23% he still narrowly won them over Royal (21%) and Sarkozy (21%). A word could be said about François Bayrou’s success (16%, up from 2% in 2002) with these same voters, proof that despite his Christian democratic map, Bayrou’s anti-system candidacy did have an impact on this traditionally anti-system electorate (nearly 80% against the EU constitution in 2005). Really, in 2007 the new factor was Sarkozy’s vitality with these voters who had historically been the most “anti-right wing” voting bloc there could be. Nicolas Sarkozy’s gains with Le Pen’s 2002 voters – some 38% of those who had chosen Le Pen on April 21, 2002 chose Sarkozy by the first round – had actually not been most pronounced with those working-class Le Pen voters but rather with the more professional and traditionally conservative portion of Le Pen’s former electorate (those in PACA, the southwest or Alsace). Le Pen’s resistance had been strongest with working-class voters and especially exurban or rurban lower middle-class voters. Nicolas Sarkozy as the candidate of the working-class might have surprised in 2002, when Sarkozy was considered too liberal (in the French sense). He was still a typical balladurien, with a more liberal, internationalist and elitist approach rather than the more nationalist, populist and statist chiraquien style which had prevailed in 1995. But Sarkozy is a wily politician and he knows how to tailor his message to the electorate. In 2007, the liberal Budget Minister of Balladur was replaced by the populistic-nationalistic Interior Minister who struck a chord with a poorer, less educated and more working-class electorate with the themes of controlled immigration, national identity, meritocracy and la France qui se lève tôt (the France which wakes up early). Regardless of what one personal opinion is of Sarkozy and the avered results of this rhetoric, those themes worked for Sarkozy and his strong showing with ouvriers by the first round confirms that. In the runoff, while Royal still won ouvriers with 54% against 46% for Sarkozy, Sarkozy’s showing with this core left-wing electorate had been 11% superior to Chirac’s showing in 1995.

A look at the map confirms what the exit polls read. Some of the right’s heaviest gains between 1995 and 2007 came in traditionally left-leaning (or even more mixed) working-class regions. Sarkozy did about 9% better than Chirac in the core constituencies of the Nord-Pas-de-Calais coal basin. In other constituencies, the same results: +6.9% in Longwy, +8.5% in Rombas, +7.6% in Forbach, +13.9% in Cernay, +8.2% in Montbéliard, +8.6% in Marignane, +7.4% in Istres, +7% in northeastern Marseille, +7.9% in Gonfreville-L’Orcher, +4-5% in Roubaix and Wattrelos or +6.3% in Tourcoing. In other industrial or heavily working-class departments of the north, such as the Oise, Somme, Ardennes and Aisne the right’s gains were just as equally impressive. The bluest areas on the above maps, at least in the east of the country, correlate strongly with a map of ouvriers. Gains were less pronounced, even in the east, in rural areas which are not as marked by a strong presence of ouvriers.

The other area which has shifted strongly to the right are those coastal Mediterranean regions or Provencal back country which have, in recent years, seen major demographic changes, most notably the influx of conservative retirees replacing more left-leaning locals, oftentimes working-class in background. These communities along the Mediterranean riviera and the Provencal back country also include other categories where the left has lost steam, somewhat, since 1995: artisans, shopkeepers and small business owners or employees. In these areas, Sarkozy scored other impressive gains: +5% in Narbonne, +6.8% in Sète, +7.3% in Nimes-2/Vauvert/Saint-Gilles, +7% in Orange and Carpentras, +6% in Brignoles.

The blue regions, which have swung to the right between 1995 and 2007, correlate strongly with an FN map. Not only east of the Le Havre-Valence-Perpignan axis, which is the reductionist view of the FN’s map, but also in other FN strongholds, notably the Garonne river valley for example and its small business owners/artisans and pieds-noirs. 

In contrast, the northwestern half of the country sticks out for its sharp trend to the left. One of the major themes in French electoral geography since the turn of the century has been the sharp shift to the left in regions such as Brittany, the Pays-de-la-Loire, Lower Normandy and Poitou-Charentes. In 1965 and 1974, some of these regions – especially Brittany and the Pays de la Loire were some of the most markedly right-wing regions with the left struggling to even break 30% in some of the deepest rural constituencies of Brittany or the inner west. There are many explanations to this shift. The most important one, in my eyes, is the declining importance of religiosity as a variable. The inner west and Brittany, alongside the southern Massif Central and Alsace, were and remain the most Catholic regions of the country (Catholic being the code word for ‘clerical’ or ‘religious’ as opposed to ‘anti-clerical’). As the left moderated over the course of the post-war era, as the boogeyman of the left being godless communists turned out wrong and as the society moved from a rural society to a urban society; the left gained in strength (the background of local grassroots activism by Christian left organizations such as the JAC or JOC also played a key role). The declining force of the right, compared to 1965 or 1974, in the inner west and Brittany was visible – though not in an extremely pronounced fashion – by the late 1980s and 1995. This trend to the left, like the working-class’ trend away from the left, only intensified between 1995 and 2007. In 2004, the left’s victory in the local elections in Brittany and the Pays-de-la-Loire was if not a shock a groundbreaking change. The other major factor in this trend was urbanization, which I touched on in my previous point. From agricultural regions, the inner west and especially Brittany have transformed into pretty urbanized modern societies. Urban and suburban growth between the 1999 and 2008 censuses was extremely pronounced in the periphery of the region’s large urban cores: Rennes, Nantes, Angers, Brest, Caen, Niort, Poitiers, Vannes, Saint-Brieuc, Le Mans and even La-Roche-sur-Yon. Those who make these regions booming are not old retirees like in the south, but rather middle-aged families who are averagely well-off, work in mid-level jobs (typically) in tertiary industries in the large urban centre.

Although some regions such as Cholet, the Vendéean bocage, eastern Ille-et-Vilaine and the Vannetais gallo were hotbeds of royalism and chouannerie up until the turn of the last century, Catholic regions in France are countries of moderate political orientation: strongly pro-European and generally more progressive on issues such as social policy or immigration. These are the strongholds of the centre, and François Bayrou had done very well in the first round in 2007. When the French right under Giscard or Chirac represented the Orleanist view of the right, these regions felt more at home. But these regions did not necessarily feel right at home in Sarkozy’s Bonapartist view of the right and the more right-wing populist policies of his government and before that his more controversial policy proposals on national identity alienated the more moderate centrist voters who had in the past felt comfortable with Chirac (in his later more moderate version).

Some of the left’s biggest gains came in areas which were traditionally rural and Catholic, but affected by suburbanization in recent years. The numbers on the above map speak for themselves: +8% in Landerneau (Albert de Mun’s old constituency in the 1900s), +5.4% in the Mer d’Iroise region of Léon, +4.5% in Ploërmel, +5.1% in Vitré, +4% in Redon, +3.5% on average in the greater Rennes, +5.3% in Nantes’ wine country, +3.7% in Ancenis, +6% in Angers-Ouest, +4.2% in Avranches and perhaps most shockingly +10.1% in Mortagne/Montaigu – Philippe de Villiers’ heartland and the real, deep ultra-conservative core of the bocage.

In the Deux-Sèvres, which has shifted left on its own as well, the left’s showing in 2007 was perhaps inflated by a strong favourite-daughter effect for Ségolène Royal. She outperformed Jospin by 6 to 8% in her department’s four constituencies, but interestingly the regions where she outran Jospin the most were the northern constituencies of Thouars and Parthenay (+8% and +7.6%) which cover the more right-wing and Vendéean-style north of the department rather than her own constituency (Saint-Maixent, +6.2%) which is more naturally left-leaning.

The constituencies in the west where the swing towards the left was most pronounced were the ones which were most right-wing. Those who had been the lone holdouts of the left when the right was dominant swung, but not with such impressive margins. The Côtes-d’Armor, northwestern Morbihan, Saint-Nazaire, Fontenay-le-Comte or Cherbourg – all older areas of significant left-wing strength – had smaller swings. In the Maine-et-Loire and the Sarthe, it is even more amusing. In the Maine-et-Loire, the old chouan Choletais had the biggest swing to the left while the Baugeois, historically left-wing, swung to the right. In the Sarthe, the swing towards the right was strongest in the east of the department (Saint-Calais) – historically the department’s left-wing region.

The same effect of declining religious practice and alienation with Sarkozy’s populist style can be seen in other Catholic regions: Lozère and the southern Massif Central and especially the Pyrénées-Atlantiques. Voters in François Bayrou’s home department swung particularly heavily towards the left, with the most pronounced swings in Bayrou’s Bearnese highlands east of Pau and the Basque Country (+10.5% for the left in Oloron). But certainly not the same story in Alsace, a region where Royal did extremely poorly in – winning only one commune in the whole region! Jospin had done fairly well in Alsace in 1995, which is not as homogeneous in its political orientation as one might be led to believe. More influenced by Muslim immigration – particularly heavy in Mulhouse and Strasbourg – rural voters in Alsace, Catholic and Protestant, have been more tempted by the FN and the Sarkozy-style UMP than voters in the inner west or southern Massif Central.

There is a huge, solidly red, blob of red right smack in the middle of the map in the Limousin and Massif Central. This is the extended domain of the Chiraquie, Jacques Chirac’s particularly strong electoral base outpouring from his fiefdom in Corrèze. Chirac had a strong favourite-son vote in his constituency but even beyond his department into surrounding departments, and his favourite-son vote tended to break old partisan boundaries: his constituency was the most right-wing in Corrèze on its own but the department and the Limousin is traditionally a base for the left. With Chirac gone, the explosion of his core of support was inevitable and perhaps all the more impressive in its form because of the antipathy between Chirac and Sarkozy, apparently shared by Chirac’s favourite-son electorate. All major candidates besides Sarkozy and even Le Pen did better or as well than in 2002 in the Chiraquie. In the runoff, Royal narrowly won Chirac’s constituency and registered a huge 16.2% swing towards the left. The left gained 15% in Tulle and 12% in Brive. Beyond there, in the Catholic plateaus of the Cantal, Lozère and Aveyron, a dispersion of the Chirac vote and the right’s difficulty with Christian democratic voters mixed to create major swings towards the left: +9.5% in Saint-Flour, +5.8% in Millau and Rodez, +5.8% in eastern Lozère and +5.5% in western Lozère. Some other pretty sharp trends in the Creuse (+8.3% in Aubusson), the Puy-de-Dôme (+8.3% in the Giscard constituency, +5% in Issoire and Riom) and Dordogne (+5% in the Périgord Nord).

The final significant shift towards the left between 1995 and 2007 was that in urban cores. France often talks about Americanization, and regardless of whether it is true in practice, there is a clear Americanization of voting patterns in Europe which is a bit unlike any other EU country. Just as the ouvriers have shifted away from the left towards the FN or the right, the white working-class in America has shifted away from the Democrats towards the GOP. Similarly, just as more liberal affluent suburban or urban voters in America break from the GOP and prefer the Democrats in recent years, similar types of voters have shifted towards the left in France in recent years. The evolution of an urban, young-ish, well educated, generally affluent and professional electorate (the cadres intermédiaires and professions libérales/cadres supérieurs) towards the left is a reversed carbon-copy of the evolution of an older, less educated, poorer and blue-collar electorate away from the left. Traditionally, up until the 1980s and mid-1990s, the CSP+ electorate leaned pretty sharply towards the left. In 1995, Chirac won 65% with professions libérales/cadres supérieurs and 55% with the cadres intermédiaires. In 2007, Sarkozy won the former with only 52% (+13% for the left) and lost the latter with 49% (+6% for the left). The upper middle-class was 60% for Chirac, but only 52% for Chirac. The high income-earners were about 63% for Chirac but only 57% for Sarkozy. In reverse, the lower middle-class had given 51% to Chirac but gave 53% to Sarkozy. Low income-earners, only 38% or so for Chirac gave 44% to Sarkozy. In the first round, Sarkozy did only 4% better than Chirac+Madelin+Boutin with those with higher education, but 8% better with those with less than the BAC (high school diploma).

The map shows this stark evolution well, and no region shows it better than the Île-de-France. There are other factors at play in this specific region: Chirac was mayor of Paris and had another favourite-son vote in Paris, and departments such as the Seine-Saint-Denis and Val-de-Marne have large and growing immigrant communities with which Sarkozy did particularly (unsurprisingly) badly. But Paris itself and especially its inner ring of suburbs have large and growing populations of young professionals, a lot of whom increasingly move to the suburbs for cheaper property prices. Within Paris itself and other neighbouring cities such as Montreuil, gentrification or boboïsation has been at work changing the makeup of old working-class hinterlands in eastern Paris into urban, trendy neighborhoods with increasingly large young and multicultural populations.

All constituencies in the Petite Couronne, even Sarkozy’s own Neuilly-sur-Seine, swung to the left in 2007. The largest swings were unsurprisingly concentrated in Paris, where Chirac had always outperformed a generic right-winger, especially in 1995. In some cases, the swings are impressive: +7.7% in the four core arrondissements, +16.5% in Paris-18th arrondissement (which includes Montmartre), +15.7% in Paris-10, +15% in Paris-18 and 19, +14.5% in Paris-11 and 20, +7.1% in Paris-5 and 6, +11.1% in Paris-11 and 12, +9.6% in Paris-13, +9.9% in Paris-14 and so forth. Swings were smaller in the old bourgeois west end, especially the core-wealthy arrondissements 7, 8 and 16. Outside Paris, the swings were generally higher in those places which have seen significant boboïsation or are otherwise home to large populations of younger, generally well-off and highly educated voters. In the most significant examples, we find +10.1% in Montreuil, +6.8% in Pantin, +6% in Fontenay-sous-Bois and Vincennes, +3.7% in Orsay, +4.3% in Versailles Nord, +8.7% in Epinay, +7% in Colombes (south) and +3.4% in Cergy. Generally, the further you get from the downtown core and the more you get into not-as-bobo parts of the Parisian basin, the swings become minimal or they become swings in the other way (note the “red belt” of swings concentrated around the core in the Grande Couronne departments).

You will tell me that perhaps the Parisian basin could be an exception or better yet is thrown off by the abnormally high vote for Chirac in Paris in 1995. The same pattern is seen with perfect and remarkable stability throughout France. Notice the isolated spots of ‘red’ constituencies even in deep blue areas (or, in some cases, light blue indicating a mere trend). In Lyon, where Chirac had also done very well in 1995 (59%), there were large swings in the downtown core. +8.9%, for example, in the 2nd constituency which is the most bobo constituency. In Marseille, which maintains some starker contrasts between deprivation and affluence, the white working-class northeast saw a big 7% swing towards the right. But in the more trendy areas downtown, there was a 6.6% swing towards the left. In varying strengths, the same swings towards the left are repeated in other urban areas – particularly the more educated and well-off areas or neighborhoods and not as much poorer working-class areas. We see +3.2% in Grenoble’s northeast, but -3.9% in Échirolles in Grenoble’s red (communist) belt. In Dijon, the poorer and more left-wing Chenôve/southern Dijon constituency swung 3.7% towards the right, but in the more well-off (and more right-wing) northwestern Dijon/Fontaine-lès-Dijon, the swing is 1.5% towards the left. In other cities, the same stories: +4.4% in Strasbourg-centre, +3.6% in Nancy (east, north and south), +3.1% in Lille (south) and +3% in Lille (centre), +5.9% in Rouen, +3.6% in western Caen (in contrast to -0.4% in the more populaire east), +3.7% in Rennes (sud), +6.3% in Limoges,  +4.3% in Poitiers (south), +7 and 8% in Toulouse, +5.7% in Montpellier (north-centre), +5.6% in Saint-Etienne (south) and finally in the impressive category: +5.2% in Nantes-Orvault, +8.8% in Nantes (centre) and +11.2% in Bordeaux (centre) which is Alain Juppé’s old constituency.

You will rightfully tell me that 2007 is a bit old now, given what has changed since then. Where are we left off today? The most significant shift since 2007 is that Sarkozy (and the UMP) have lost the ouvriers and his spectacular inroads from 2007 now seem a long way away.

Era % PS 95-R2 % PS 07-R2 % Left R10-R2 % Right R10-R2 % PS 12-R2 (poll) % FN 02-R1 % FN 12-R1 (poll)
France 47% 47% 54% 36% 60% 17.2% 17%
Ouvriers 65% 54% 60% 20% 76% 30% 32%
Ouv/Fra +18% +7% +6% -16% +16% +13% +15%
CPIS* 35% 48% 58% 37% 64% 30% 6%
CPIS/Fra -12% +1% +2% +1% +4% +13% -11%
PI* 45% 51% 63% 29% 61% 14% 16%
CPL/Fra -2% +4% +9% -7% +1% -3% -1%

* The most recent poll which gives crosstabs was Ifop on October 20, with Hollande at 60% nationally.
* Cadres supérieurs, professions libérales or Cadres et professions intellectuelles supérieures
* Professions intermédiaires or cadres moyens

The above chart is based on exit polls, and, for 2012, on actual polling, so it is perhaps not the most accurate picture but it paints a pretty clear overall picture.

Nicolas Sarkozy’s winning coalition in 2007 had been possible because, in part, of his success with ouvriers with whom he poll 46% whereas Chirac had garnered just 35% with them 12 years prior. His gains with lower-income voters in eastern France had compensated for his weaker showing with middle-income voters in western and urban France, where his 52% with the CPIS category was quite tepid compared to the margins Chirac had posted with them in 1995. Since then, the government’s more right-wing policies on matters such as immigration and particular incidents such as the Roma expulsion affair tacked the government and the UMP to the right and did little to please more centrist, moderate voters which CPIS voters can be broadly seen as politically. As a result, CPIS voters have only moved further and further to the left. But the government’s tack to the right appears increasingly desperate and has had little success in wooing over FN voters or lower-income voters such as ouvriers. A poor economy, unpopular fiscal and social policies, an elitist style (bling-bling) and corruption scandals have worked in tandem to make Sarkozy’s strong showings with these voters in 2007 seem like a very distant dream for the right. The exit polls are pretty stark on this point: the UMP polled only 17% with ouvriers in the first round of the regional elections when the UMP polled 27% nationally. In the runoff, the right won only 20% with these voters – tied with the FN. Actual polls for next year’s election shows Marine Le Pen reaching her father’s 2002 levels with ouvriers and Sarkozy collapsing to lows rarely seen even in the days of left-wing dominance of ouvriers – as low as 9% in some polls!

To tie in this story with that of 2012, the fundamental thing here is that Nicolas Sarkozy has lost the ouvriers and has been further isolated with cadres and other middle-income voters. I think that is the fundamental dynamic at work behind the polls.

This article is certainly not thorough. I have made no comment about the fact that ouvriers and lower-income voters form a big part of non-voters, I made only passing references to the FN’s strengths with ouvriers and I completely ignored the Greens’ potential challenge to the PS for the control of CPIS and middle-income voters. A lot more could be said about all these topics, but I think that I’ve covered what I wanted to cover and hit the main points in the exploration of the changing face of the French left between 1995 and 2007.