Disappearing point - mapping the rout of the Front National and La France Insoumise

France has a well-established tradition of electoral cartography and it’s something the French do very well. Under the Third Republic (1870-1940), long before more recent innovators like Alain Lancelot and François Goguel, newspaper editors employed graphic artists to design maps showing the political balance across the national territory. Nowadays, no history or politics textbook worth the name can forego visual representation of political trends with a few maps.

Now, in an age of ever-improving mapping software and the ability to reproduce these on paper or via webpages (interactive or otherwise), the range and quality is generally much higher than UK readers are used to. The development of software that can map not just France’s departments but also its 35,000 communes means that even the casual user can drill down into more detail than she or he could ever really need. The sheer detail available via the electoral supplements of the French dailies puts the output of the British press not only in the shade but also to shame.

Before looking at the new political maps of France that have emerged over the last few weeks, I should explain some of the terms I shall use here.

The first are commune and department. The commune is the basic unit of local administration in France. There used to be more than 36,000 of them, but recently they were rationalised down to 35,416! They vary enormously in size, from urban centres like Paris, Lyon or Marseille down to the smallest village. The next level up are the departments, created at the time of the French Revolution and still intact more than 200 years later. (Above the departments there are regions, which feature here for ease of geographcial location, but are not used so much to represent electoral trends.)

The departmental map of France is the one most readers will be familiar with and it has tended to be the one political geographers use for mapping politics onto French territory. But, as I have already mentioned above, the development of maps of France’s communes can show some very distinctive, intradepartmental differences that a purely departmental map does not. It’s also very useful when looking at voting patterns for presidential elections which are not subdivided into constituencies. I’ll come back to that point later when I consider the results of the first round of the legislative (general) elections.

Let’s look at some maps

The first is from the ministry of the interior website and shows the result of the first round of the presidential election on the basis of the winner in each department. (The ministry also publishes detailed results per candidate per department and commune here, but you have to be sharp-eyed to spot the link).

Perhaps what first strikes the viewer looking at this map is not the mass of dark blue (Le Pen) and dark pink (Macron) but the handful of royal blue for Fillon and the three metropolitan (mainland) departments where Mélenchon was in first place. We can deal with these quite quickly.

The three departments in the north-west that voted for Fillon are his home department of the Sarthe, to the bottom-right of the three) and its immediate neighbours (anticlockwise), the Orne and Mayenne. Haute-Savoie (down in the Alps) and especially the Lozère, in the southern Massif, Central, both have long and deep traditions of moderate conservatism. In the former case, as we shall see a little later when we dig into the communal map, Fillon also did well in Savoie, although Macron ‘won’ the department.

If you had told me Mélenchon would win in three metropolitan departments, I would have put money on the Seine-Saint-Denis, the suburban department to the north of Pairs and also on Haute-Vienne, where left-wing rural radicalism remains strong and whose capital, Limoges, is known as ‘la ville rouge’. I might not have guessed at the Ariège, but it also has a long tradition of ‘left of the left’ politics and is nicknamed ‘terre de résistance’. Very Mélenchoniste.

Source: Ministère de l'intérieur

But the overwhelming impression, of course, is of two east-west blocks with a few incursions by the pink of Macron into Le Pen’s bleu marine (geddit?) and a string of blue departments in the south-west. Again, much of this is as we would expect, based on Le Pen’s performance in 2012 and intermediate elections since then.

In regional elections in December 2015, Le Pen and her niece Marion Maréchal Le Pen (retired) very nearly took control of the assemblies for the Haut-de-France in the north and Provence-Alpes-Côte d’Azur (PACA) and the FN’s progress in the Grand-Est (Alsace, Lorraine and Champagne) and Burgundy and France-Comté is not new either. Le Pen also confirmed her progress along what is sometimes called the Languedocien Mediterranean littoral and inland along the Tarn and Garonne valleys. What is new is her progress in upper Normandy and parts of the Beauce, the breadbasket of France.

The departments of upper Normandy – Seine-Maritime and the Eure – have been FN targets, or terre de mission, for a while. The former, which includes Dieppe, Rouen and Le Havre, suffers from some of the same problems as the Nord and the Pas-de-Calais, where Le Pen’s party has put down firm roots: deindustrialisation on the one hand and an agricultural crisis on the other have proved fertile terrain for the FN, although, paradoxically, the urban centres remain resistant. It’s the périurbain (suburban and small town France) and rural areas in these that vote FN. Rouen put Macron in first place, Dieppe and Le Havre Mélenchon. And it was a similar picture in the Loire, the department that links the blue in north and south. Its principal population centre, Saint-Etienne, voted Mélenchon.

In total, Le Pen came first in 47 of France’s metropolitan departments. Macron did well in the west and the centre, in the Paris region, and in much of the old Rhône-Alpes region (with the exception of the Loire). His success matches very closely the performance of François Hollande in 2012, except in upper Normandy.

Still, looking at electoral map on a departmental basis is something of a blunt instrument. The map of the winner in each commune gives a more nuanced sense of the political landscape. I have included two below for reasons that I’ll explain in due course.

The first map is from the front cover of the printed edition of Le Monde (25 April 2017) and it shows the winner of the first round of the presidential election by commune: grey for Le Pen, yellow for Macron, blue for Fillon, red for Mélenchon and, astonishingly, light blue for Jean Lassalle, who won in a handful of rural departments in his native Béarn. (The zones marked out by white lines are the regions.) Although this scanned version of the map is not altogether easy to read, I hope it is clear enough.

Source: Le Monde, 25 avril 2017.

The ‘by commune’ map reveals that behind the departmental map, something else was going on, especially in PACA, where there is some blue (this was Sarkoland back in 2007 and 2012), but also red. And further west, in what is now known as Occitanie, there is a red hinterland behind the dark grey, Lepéniste littoral. Indeed, with the exceptions of the Ile-de-France and a smattering of Breton communes, Mélenchonisme was largely a southern phenomenon (and that's a left-wing tradition that goes way back). The map also shows some important incursions by the FN along the lower Garonne valley and on both banks of the Gironde estuary that we haven’t seen before.

I have also included below the map published by Le Figaro on the same day. It has the advantage of showing the department names but also France’s major cities and underlines the weakness of Le Pen’s performance in the urban centres, on the whole. It also, by choice of colour scheme – purple for Le Pen rather than dark grey – gives a rather more alarming impression of her performance. The matter of colour selection is more or less universal for all the other candidates: orange or yellow for Macron, red for Mélenchon, blue for Fillon and pink for Hamon. Le Pen remains problematic. Her father was usually represented with brown and she sometimes is too. (Brown shirts, see?)

Source: Le Figaro, 25 avril 2017

Speaking of the other candidates, Le Monde also produced maps of how each of the five main candidates performed in the first round.

Source: Le Monde, 25 avril 2017

Le Figaro published a similar graphic by commune.

Source: Le Figaro, 25 avril 2017

And yet, for all the detail, the maps of support by department or commune don’t really convey the tale of the election properly either, because to look at Le Figaro’s graphic, if we removed the scores next to the maps, you would imagine that Le Pen was in the lead because they reflect percentage of the vote, not population size.

This is an aspect of electoral mapping that has exercised Hervé Le Bras, a demographic historian. Over the last few years he has brought the techniques used in mapping demographics to bear upon analysing the FN vote in France. What he has produced looks very different from the conventional administrative and political map. It takes a bit of getting used to, but it very effectively demonstrates the périurbain and rural implantation of the FN in comparison to Macron. I have scanned the hard-copy version from Le Monde (27 April 2017) below, but it can also be found here. (You’ll need to pay to access the full article.)

Source: Le Monde, 27 avril 2017

What Le Bras and his team have done is to present Macron in shades of purple and Le Pen in her more traditional brown according to how they scored in each commune or urban centre according to its population and their share of the vote above their national average. So, the light purple shows where Macron scored between 0 and 5% above his 24.01% average, the mid shade 5-10% and the deep one more than 10% and where Le Pen scored less than her national average. The same goes in reverse for Le Pen.  The areas marked in darker green represent communes where they BOTH scored above their national average, while the light green shows those where they both scored below their average. Le Bras is keen to underline here ‘les deux France’ in opposition to one another – Macron and Le Pen. So, the lighter green area to the north and north-west of Paris represents the heavily populated suburban area of Seine-Saint-Denis, where Mélenchon did particularly well and where neither Le Pen nor Macron made their average.

What shows up less well on the reproduced image here are the boundaries of France’s urban agglomérations (groups of large towns and cities). Nevertheless, Le Bras’ projection does the job and the user can very quickly see that, apart from the north, south-east and south-western Mediterranean coast, Le Pen scores poorly in urban France. And even in the north, around Lille the left-wing traditions persist, while the same is true in the cities of the Grand-Est. And despite Le Pen’s success along the Mediterranean littoral, she scored poorly in Nîmes and Montpellier, the capitals of the Gard and the Hérault.

Whether one’s sentiment at Le Pen finishing second to Macron was one of disappointment or relief, her score was still a record for the FN and the logical next step is to take the first round performance of the presidential candidates and extrapolate these to the legislative elections. This is easy enough to do, because French constituencies follow municipal and departmental boundaries and can be mapped precisely. It’s not an exact science of course, and no-one could have predicted Le Pen’s meltdown in the last week of campaigning for the presidentials nor the FN’s loss of direction in the run-up to the general elections, but looking at the figures published by Le Monde in its 26 April edition underlines just what a trainwreck Le Pen has presided over since then.

Patrick Roger, one of the senior political correspondents at Le Monde, was at pains to underline that the map below was in no way a projection of what would happen in the legislative election. Nevertheless, it makes for fascinating reading. The vote of the 23 April put Macron ahead in 230 of France’s 566 metropolitan and overseas constituencies, Le Pen in 216, Mélenchon in 67 and François Fillon in 52. The results would also have led to 238 four-way run-offs and 225 three-ways… Of the 216 constituencies where she was in first place, moreover, Le Pen had scored between 25 and 30% in 105 and over 30% in another 83. On that basis, Roger hazarded, an FN group in the National Assembly of one hundred deputies was by no means out of the question. (A very interesting article regarding two-, three- and four-way runoffs can be found here.)

What a difference six weeks make.

Source: Le Monde, 26 avril 2017

Hindsight is, of course, a wonderful thing and we now know that Le Pen couldn’t hold it together. Coming second and scoring less than 22% in the first round was bad enough. Only getting to 34% in the second was worse and the map below, from Le Figaro, shows how that played out between Le Pen and Macron.

Source: Le Figaro, 8 mai 2017

The contrast between the map Le Monde published on 26 April and the one that followed the real first round on 11 June is stark. The choice of yellow for La République en marche doesn’t help with the visualisation of the map, but at least it’s consistent. (The figures in each department correspond to the titles of each constituency: Corrèze 1, Corrèze 2, Pas-de-Calais 11 and so on. Imaginative, non?)

Source: Le Monde, 13 juin 2017

We now know that in the face of En Marche! the extremes have collapsed. It is their voters who failed to turn out on 11 June and Le Pen and Mélenchon know it. Between rounds Le Pen has appeared on national radio and TV, exhorting those who voted for her on 23 April to come out again on 18 June. Reste à voir.