French presidential election, 2022

[Antonio the Sixth]

Okay, so I tried to overcome an autism and made a map using the (probably incorrect) Data.gouv results. Hopefully we'll soon get accurate ones and they will show the same patterns.

Anyway, back in 2017 I made a map showing the top-two leading candidates by Department. The results were quite fascinating, as there was a lot of geographic variation across the country. Macron, Le Pen, Fillon and Mélenchon all made it through in specific parts of the country, and in a variety of combination (except Fillon-Mélenchon, lol - nowhere put both candidates ahead, understandably):




Well, I made the same map for 2022, but unfortunately it's far less interesing:



Obviously because there are only 3 candidates within runoff distance instead of 4, that means there were only 3 combinations this time (well, except for Wallis & Futuna which hilariously put Pécresse second - we stan our Queen of the wide Pacific). But still, with Mélenchon coming so tantalizingly close to qualifying, you'd expect a lot of Departments to vote for a Macron-Mélenchon or Le Pen-Mélenchon runoff. Unfortunately, that wasn't really the case: geographically speaking, France overwhelmingly voted for a Macron-Le Pen runoff, and Mélenchon actually came in the top 2 in fewer Departments than in 2017. Partly, this is due to Macron doing better - while the gap between Mélenchon and Le Pen went down slightly, the gap between him and Macron was significantly larger (6 points instead of 4.5), which ate into the number of Le Pen-Mélenchon Departments. But in addition, Mélenchon's improvements were largely concentrated in Ile-de-France (as well as DTOMs, where he did indeed qualify for the runoff almost everywhere). Elsewhere, in non-Parisian metropolitan France, he stagnated, allowing Le Pen to overtake him in much of the Southwest where he had come second in 2017. The result is, sadly, a more boring map, with IdF sticking out like a sore thumb and not much else going on.

Oh well. I'll try to find other map concepts that reveal more interesting patterns.

[Antonio the Sixth]

Why did Le Pen do so well in Lille?

She didn't.  That yellow wart near the top with the tiny red dot in the middle is Lille.

Yeah, this would be the equivalent of asking "Why did Bush do so well in Austin?" lmao.

[Antonio the Sixth]

Why did Le Pen do so well in Lille?

She didn't.  That yellow wart near the top with the tiny red dot in the middle is Lille.

Yeah, this would be the equivalent of asking "Why did Bush do so well in Austin?" lmao.

Bush might have won Austin in 2000. He won Travis County by 5 pts and Austin is about 75% of the county by population.

Huh, wow. That really was a different world back then.

[Antonio the Sixth]

A request: I'm going to do some maps of the results at arrondissement and municipal level for the City of Paris and the Petite Couronne and happen to have a strongly principled objection to decision of authorities to include prisoner ballots in with the results from the 1st arrondissement - psephological vandalism in my view. Does anyone have the results for that arrondissement without the numbers from the prisoner polling district?

I think your best bet is to find the specific precinct where all these votes were dumped in (should be easy as it will contain more than half the votes in the 1st overall) and just remove it.

Precinct-level results can be found here (though I can't vouch for the data's accessibility): https://www.data.gouv.fr/fr/datasets/election-presidentielle-des-10-et-24-avril-2022-resultats-definitifs-du-1er-tour/

[Antonio the Sixth]

To get back to the electoral analysis, one thing that struck me, and that made it hard to analyze the results at first (and is the reason why projections were slightly off on election night), is the fact that the swings in Ile-de-France were so different from those in the rest of the country. This is very unusual, as usually while IdF is of course electorally distinctive in some ways (most notably by having a very weak FN), it used to be a swing region for most of the Fifth Republic, and even when it wasn't, its overall swings tended to match those of the rest of Metropolitan France. This time though, that wasn't the case.

Here I've calculated the results for the 11 regions of mainland France other than IdF:
Macron 27.35% (+4.42)
Le Pen 25.47% (+2.25)
Mélenchon 19.86% (+0.63)
Zemmour 7.03%
Pécresse 4.53% (-14.88)
Jadot 4.50%
Lassalle 3.49%
Roussel 2.43%
Dupont-Aignan 2.16%
Hidalgo 1.81%
Poutou 0.80%
Arthaud 0.59%

And here's the Ile-de-France results:
Macron 30.19% (+1.56)
Le Pen 12.97% (+0.40)
Mélenchon 30.24% (+8.49)
Zemmour 7.47%
Pécresse 6.19% (-16.00)
Jadot 5.40%
Lassalle 1.59%
Roussel 1.92%
Dupont-Aignan 1.59%
Hidalgo 1.43%
Poutou 0.60%
Arthaud 0.40%

In IdF, the momentum was almost entirely concentrated around Mélenchon, who amazingly managed to come first in France's most populous region. This wasn't even just a product of tactical voting, either: even the non-Mélenchon left candidates managed to add 0.88 points to their overall scores from 2017. While Macron gained a point or two, his electorate was massively shifted to the right, eating up most of Fillon's voters. Le Pen also barely improved from her already low 2017 score. In the rest of mainland France, meanwhile, the roles are reversed: Mélenchon gained just half a point (less than non-Mélenchon left candidates, who collectively gained 2.24), while Macron and Le Pen gained significantly more. So outside of IdF, the gap between Mélenchon and Le Pen actually grew (from 4 points to 5.62). Le Pen actually won the non-IdF mainland vote in 2017, edging out Macron by 0.3. Macron's larger gains outside IdF allowed him to pull ahead this time, and seems to indicate that he bled far less support to the left overall. It's going to be interesting to see the long-term implications of this process.

Finally, let's not forget about the third big bucket of votes that propped Mélenchon up significantly: the DTOMs. While their electoral impact is limited by low turnout (only 43.22% valid turnout in 2017 and 42.09% in 2022) they nevertheless provided massive swings toward Mélenchon that helped him come so close to Le Pen.

Macron 20.56% (+0.15)
Le Pen 21.30% (-0.60)
Mélenchon 39.99% (+19.22)
Zemmour 4.06%
Pécresse 4.00% (-16.73)
Jadot 2.17%
Lassalle 1.21%
Roussel 0.71%
Dupont-Aignan 2.43%
Hidalgo 1.79%
Poutou 0.77%
Arthaud 1.01%

The 2017 result in DTOMs was an almost 4-way with each candidate within 1.5 point of each other (and Le Pen, of all candidates, hilariously ahead). This time however, Mélenchon really swept through, nearly doubling his score and distancing any competition. Hilariously, if you looked at these results with no context, you might assume he siphoned off mostly 2017 Fillon voters, but of course the reality probably involves some major communicating-vases effect, along with general DTOM randomness. Either way, that helped pad Mélenchon's result and, if he'd done a bit better elsewhere, could have been the key factor bringing him to the second round.

I already posted the full expat results earlier in the thread. All that's left is Corsica, which can easily be found online and which amounts to less than 150k votes, so I won't bother posting the results here.

[Antonio the Sixth]

To get back to the electoral analysis, one thing that struck me, and that made it hard to analyze the results at first (and is the reason why projections were slightly off on election night), is the fact that the swings in Ile-de-France were so different from those in the rest of the country. This is very unusual, as usually while IdF is of course electorally distinctive in some ways (most notably by having a very weak FN), it used to be a swing region for most of the Fifth Republic, and even when it wasn't, its overall swings tended to match those of the rest of Metropolitan France. This time though, that wasn't the case.

Because I have an intellectual horror of admitting anything positive about the flipping Parisians, I have been wondering a fair deal about the extent to which IdF patterns hold  up in the other metro areas of France, or the degree to which this was a specifically Grand Paris phenomenon. For instance, Mélenchon's third best department outside of IdF whas Rhône. Rhône! Lyon! And all of a sudden Auvergne-Rhône-Alpes is one of the most left wing regions in the country. I'm sorry but this just isn't... right...

More to the point though, while the Paris metro area is easy to pull out because it is largely coterminous with the Île de France region, it's harder to do that with the other ones. For instance, Lyon stretches across Rhône, Ain and Isère - with the periurban regions of the latter two being much more Le Pen inclined then the combination of city centre, low income banlieues to the east and swanky sububs to the west that you see in the department itself. And most other big cities would fail to even be the singular determining factors in their own departments.

On which note, the Jean Jaurès foundation had some anaylsis on the Mélenchon vote - or the specifically identifying four vague demographic groups building up his support (young people, ethnic minorities, the Dom-Tom, and a certain type of low-mountain rural region with heavy alternative lifestyle types, a contestatary (often protestant) history and heavy public sector dependence). But also, had this summary of how the Mélenchon vote developed from 2017 in a selection large cities:



Point being, it's obviously unsatisfactory in so far as it slices off the banlieues and the exurbs. But, well, the double digit gains are in a variety of types of place. Or at least, the predictably racist reasons that I'm sure we've all seen about why Mélenchon did so well in Seine-Saint-Denis - well - it clearly doesn't apply in a lilly white city like Rennes; nor does it explain the relatively tepid gains in comparison in somewhere like Marseille

Indeed, if anything this suggests a major swing of the vaguely leftish urban bobo types toward Mélenchon (and presumably, away from Macron). Which is certainly interesting, since those are not the type of former PS, Macron-curious voters that I would have expected to come back to Mélenchon (it's certainly telling that he's down in a lot of the rural Southwest and Massif Central areas - Roussel definitely took votes away from him there, but it's still not enough to explain the underperformance).

So, the bottom line appears to be that the left vote is boboifying even further, somehow. Not a kind of trend that bodes well for its future, and should probably relativize celebrations of Mélenchon's success as indicating a future path for the French left.

[Antonio the Sixth]

I watched the debate live last night and I guess I broadly agree with the pundit consensus. Le Pen did a lot better than in 2017 and avoided any catastrophic mistake that would have risked collapsing her support. She was at her strongest talking about bread-and-butter issues (her proposals are sh*t but she has the advantage of channeling French people's lived experiences in a way Macron as the incumbent just can't). She stumbled badly on Russia and Europe, and later on institutional issues, where Macron completely skewered her. Macron was very aggressive this time, arguably coming across as condescending or at least not shaking off the image of arrogance he's always had hanging around him. But ultimately that allowed him to point out all the flaws and inconsistencies in Le Pen's program, and that should convince some voters that's she's simply too much of a risk.

The debate probably won't change much overall, which is good for Macron obviously. I don't think we should predict a complete Le Pen collapse like in 2017, at least not yet, but neither should it be too close like some pre-first round polls indicated.

[Antonio the Sixth]

While we prepare for the upcoming runoff, here's a bit more about the man who didn't qualify for it. Since a lot of maps of the Mélenchon vote have already been posted, I made this one based on his percentage of registered voters rather than of valid votes, to better capture his overall support. I also tried to experiment a little with continuous color scales. Still not sure how I feel about it (it seems to work better in some maps than others), but it was worth an attempt.



Not a ton of surprise there, but the IdF concentration of Mélenchon's support stands out even more. In the rest of mainland France, Mélenchon actually declined slightly as a percentage of registered voters (from 15.02% to 14.82%). The areas where Mélenchon gained are...interesting. That includes almost the entirety of the Eastern side of the hexagon, especially Alsace and Rhône-Alpes, as well as in a few departments that are home to major emerging metropoles like Haute-Garonne. So overall, another confirmation of what we were talking about with PB regarding the "metropolitan" nature of the left's gains this year. Admittedly the picture would not be as stark with Roussel included, and I should probably make a map comparing Mélenchon'17 to Mélenchon+Roussel'22 next.

[Antonio the Sixth]

Here's another map - one I've been meaning to make since I saw the final results. I noticed you could break the electorate up in to three broad ideological blocs of almost equal size. Of course those categorizations are eminently contestable, but I think they still reveal some interesting things.

Left (Arthaud-Poutou-Mélenchon-Roussel-Hidalgo-Jadot): 31.95%
Liberal Right (Macron-Pécresse): 32.62%
Nationalist Right (Dupont-Aignan-Le Pen-Zemmour): 32.29%

That leaves out the unclassifiable Lassalle's 3.13%, but otherwise splits French voters into almost perfect thirds.

So, what does this 3-way divide look like mapped out? I finally took the time to find out:


For France, a genuinely shocking degree of geographic bias. The far-right bloc came out ahead in a solid majority of departments, and if France had an "Electoral College" it would be solidly ahead of it too. It of course wins overwhelmingly in the traditionally right-wing Mediterranean France, and in France's vast postindustrial Northeast, but also in large swathes of the country that don't have such a strong political identity, like the Centre region and even much of the traditionally left-wing Southwest. This is only partially explained by the nationalist bloc's weakness in IdF: even if you look at the non-IdF continental France only, the bloc only wins with a plurality of 34.66%, not that far from the liberals' 31.88%.

However, it is clear that the nationalist bloc is much weaker across France's globalized metropoles, and Departments anchored by one such tends to be strongholds for either the left or liberal bloc. On the liberal side, the Grand Ouest also really stands out as its last redoubt, along with the traditionally posh West side of IdF. Other areas of support includes also upscale areas like Haute-Savoie and Rhône, as well as a little redoubt of what might have once been Hollande Country in the Massif Central and Bayrou Country in the Pyrénées-Atlantiques. As well as the expats and Pacific territories, where support for continued unionism with France usually translates into support for the center-right establishment.

As for the left, it (and this really cannot be stressed enough) dominated Ile-de-France, winning 39.99% in the region. Also in most non-Pacific DTOMs, where Mélenchon's scores were pretty staggering. Aside from that, its areas of support in non-Idf continental France tend to be anchored around metropoles like Nantes, Bordeaux, Toulouse and Grenoble. A few redoubts of rural leftism survived, like Ariège, Lot, Haute-Vienne and Hautes-Alpes, but those are just the remnants of once-vast swathes of leftist support across the South.

[Antonio the Sixth]

https://twitter.com/EuropeElects/status/1517467333454876674?s=20&t=XHaWp9qRB...


Also a 53/47 odaxa poll, which is no change from their last numbers, seven days ago. Their projected abstention is much lower than anyone else.

Final Yougov. No nice voter movement chart like last time, but the transfer percentages are similar:

https://twitter.com/YouGov/status/1517458617095569408?s=20&t=rqO1oqGxIDE9Ikv...

Macron is basically two touchdowns ahead with a two minute warning. They just need to not fumble the ball twice over the course of 4 downs.

Good job finding a metaphor no French person would understand.

[Antonio the Sixth]

Took me long enough to update it, but here's the new bar chart summarizing the breakdown of the electorate in every Presidential first round:



The combined left is almost back to its level from 1969. Hurray, I guess?

And with Le Pen and Zemmour combined, the far right is higher than ever. Oh well. At least with the collapse of Pécresse there isn't much of anything else to the right of Macron.

[Antonio the Sixth]

Here's another map - one I've been meaning to make since I saw the final results. I noticed you could break the electorate up in to three broad ideological blocs of almost equal size. Of course those categorizations are eminently contestable, but I think they still reveal some interesting things.

Left (Arthaud-Poutou-Mélenchon-Roussel-Hidalgo-Jadot): 31.95%
Liberal Right (Macron-Pécresse): 32.62%
Nationalist Right (Dupont-Aignan-Le Pen-Zemmour): 32.29%

I was looking at these a bit too, and the department that most closely reflects the national average (less than 1% for each block) is... Savoie.

As in, I know the focus in terms of the left wing vote is rightly on the metro areas, but I have my own biases in terms of being interested in what the neighbours are doing and it is quite striking the degree to which the two Savoies have gone from being rock solid Conservative heartlands - Haute-Savoie was one of Sarkozy's top 5 departments - to actually being both close to the national average and giving relatively ok scores to the left, even in the Haute. All the more so that, Geneva suburbs aside, neither department has any noticeable urban areas.

Chambéry isn't noticeable?!? Excuse me??

Like, if you mean that it's not a capital-M Metropole in the Guilluy sense, then fair, but an urban area of 200k inhabitants isn't nothing, and it certainly has a major influence on Savoie's politics. Without it, Savoie would certainly not so closely match France as a whole - which I agree makes it fascinating. I was actually telling Nathan just yesterday about how Savoie a fascinating microcosm. It has a trendy, economically dynamic and boboifying capital in Chambéry, various stipes of suburbs around it, a lot of very upscale resort areas around the Alps, old industrial and postindustrial centers in the Maurienne, and a lot of quaint rural towns. It has a bit of everything and that's probably why it matches France so well. Of course, anything that could be said about Savoie was said almost a decade ago in this post in Gaël's sadly defunct blog. Definitely worth a read.

[Antonio the Sixth]

Here's another map - one I've been meaning to make since I saw the final results. I noticed you could break the electorate up in to three broad ideological blocs of almost equal size. Of course those categorizations are eminently contestable, but I think they still reveal some interesting things.

Left (Arthaud-Poutou-Mélenchon-Roussel-Hidalgo-Jadot): 31.95%
Liberal Right (Macron-Pécresse): 32.62%
Nationalist Right (Dupont-Aignan-Le Pen-Zemmour): 32.29%

That leaves out the unclassifiable Lassalle's 3.13%, but otherwise splits French voters into almost perfect thirds.

So, what does this 3-way divide look like mapped out? I finally took the time to find out:


For France, a genuinely shocking degree of geographic bias. The far-right bloc came out ahead in a solid majority of departments, and if France had an "Electoral College" it would be solidly ahead of it too. It of course wins overwhelmingly in the traditionally right-wing Mediterranean France, and in France's vast postindustrial Northeast, but also in large swathes of the country that don't have such a strong political identity, like the Centre region and even much of the traditionally left-wing Southwest. This is only partially explained by the nationalist bloc's weakness in IdF: even if you look at the non-IdF continental France only, the bloc only wins with a plurality of 34.66%, not that far from the liberals' 31.88%.

However, it is clear that the nationalist bloc is much weaker across France's globalized metropoles, and Departments anchored by one such tends to be strongholds for either the left or liberal bloc. On the liberal side, the Grand Ouest also really stands out as its last redoubt, along with the traditionally posh West side of IdF. Other areas of support includes also upscale areas like Haute-Savoie and Rhône, as well as a little redoubt of what might have once been Hollande Country in the Massif Central and Bayrou Country in the Pyrénées-Atlantiques. As well as the expats and Pacific territories, where support for continued unionism with France usually translates into support for the center-right establishment.

As for the left, it (and this really cannot be stressed enough) dominated Ile-de-France, winning 39.99% in the region. Also in most non-Pacific DTOMs, where Mélenchon's scores were pretty staggering. Aside from that, its areas of support in non-Idf continental France tend to be anchored around metropoles like Nantes, Bordeaux, Toulouse and Grenoble. A few redoubts of rural leftism survived, like Ariège, Lot, Haute-Vienne and Hautes-Alpes, but those are just the remnants of once-vast swathes of leftist support across the South.

Following up on this, I decided to make a map of the top 2 blocs in each department. In other words, this tells you which bloc finished last.



Actually a bit more variety here than in the map for top 2 candidates. We see a substantial part of France would have placed the left bloc in at least second place. Fascinatingly, this part of France is highly concentrated in (aside from IdF, which we've already talked about at length) Southern France. This North-South divide is also a bit of an oddity in France: in a way, you could almost see it as a return to the maps from the 60s and 70s, when the left's strength was concentrated in a C-shaped arc stretching from the Massif Cental to the Southern Alps. Since these days however, the left collapsed on the Mediterranean Coast, while gaining in areas like the Grand Ouest. We still see some of the effects of that in Brittany, which largely went Lib > Left > Nat. But the most fascinating thing is this relatively resilient Left bloc around the Mediterranean arc. Of course, in this areas the nationalist right came far ahead, as you could see in the previous map. Still, the fact that the left came in second here is interesting. By contrast, it is striking how little you can see the old left strongholds in the North and East. Coming third in Seine-Maritime, Meurthe-et-Moselle, Ardennes, and even Nievre and Pas-de-Calais(!!!) is absolutely pathetic, and a stunning indictment of the French left's decline in it old post-industrial bases. This is a whole new political geography of France.

[Antonio the Sixth]

A simple average of the final polls of 10 polling firms that conducted polls this final week has Macron at 55.3%. For whatever it's worth, it's a little lower (54.5%) among the 3 pollsters that were most accurate in the first round. The trendline on Wikipedia, meanwhile, points to something slightly above 56%.

Guess we'll find out tomorrow.

[Antonio the Sixth]

Thank you Gaël. Fantastic maps, as always. I only wish I had your patience to deal with constituency-level maps.

A simple average of the final polls of 10 polling firms that conducted polls this final week has Macron at 55.3%. For whatever it's worth, it's a little lower (54.5%) among the 3 pollsters that were most accurate in the first round. The trendline on Wikipedia, meanwhile, points to something slightly above 56%.

Guess we'll find out tomorrow.

There's also the roughly 4 point underperformance of Le Pen in  2017 compared to the polls. Though I have a feeling they won't be out by much.

There's really no way of knowing if the polls will be off by that much, and if so, in which direction. I don't think there's any consistent pattern to find in the past few electoral cycles, so I won't even hazard a guess. We'll have to find out in 6 hours or so.

[Antonio the Sixth]

A simple average of the final polls of 10 polling firms that conducted polls this final week has Macron at 55.3%. For whatever it's worth, it's a little lower (54.5%) among the 3 pollsters that were most accurate in the first round. The trendline on Wikipedia, meanwhile, points to something slightly above 56%.

Guess we'll find out tomorrow.

In your opinion, how much of a chance does LePen have?

I guess the 10% or so that models are converging on seems like a decent conservative estimate, if only because of the big unknown around abstention and blank votes. The former looks like it won't be too high, thankfully, so maybe the chances of a bad surprise are falling further down.

[Antonio the Sixth]

https://twitter.com/EuropeElects/status/1518251021549686784?s=20&t=_l-3_yUje...

https://twitter.com/EuropeElects/status/1518252920621383680?s=20&t=_l-3_yUje...

Looks like stronger turnout in the Southwest and weaker in the Northeast, which seems good for Macron. Lots of noise elsewhere though, and of course the drop in IdF is rather noticeable.

A map of the evolution of turnout from the first round would probably be more helpful.

[Antonio the Sixth]

What percentage would Le Pen have to hit to have won France outside of Paris and its regional area (is that the IDF?).

Back-of-the evelope math suggests that if you take the 2017 results and apply uniform national swing to get Macron at 55%, you get something like 70% in IdF. IdF is about 15% of the vote, so that's about 10% for Macron, 5% for Le Pen. Which means the rest of France would be something like 45% Macron, 40% Le Pen, or renormalized, 53% Macron 47% Le Pen. So if you get Macron down to 52% nationally, UNS suggests that Le Pen might win the rest of France. Still highly speculative, of course.

And will Le Pen have a shot to win the white French vote?

There's no such thing as "the white French vote" statistically speaking, because France doesn't recognize any ethnic categorizations. Anyway even if we did, the answer is almost certainly no.

[Antonio the Sixth]

58% was basically where Macron's polling against Le Pen was before The NarrowingTM, so ultimately we're more or less back where we started. A clear win, not really close or even competitive, but still very, very far from last time's 66%. When you factor in the record abstention and likely record blank vote as well, Macron's mandate will be historically pathetic. And conversely, Le Pen reaching above the 40% threshold is a historic results that means there's a real chance of the far-right governing France in the near future. It's a scary time to be French, and it's going to remain scary for the coming five years. Let's just hope that in five years the left will find a way to be relevant again.

[Antonio the Sixth]

Worth noting that this is the first re-election of a sitting French President for twenty years.

Macron has under his belt two of the most most one sided victories, with Chirac in 2002 being of course something exceptional. He's extremely close to Pompidou's margin.

He's only the second president, with De Gaulle, to be reelected while controlling a parliamentary majority. And since De Gaulle's first election was indirect, he's the only such president to be popularly elected twice.

Such an undeserved record for such an unimpressive politician.

[Antonio the Sixth]

Melenchon speaking now and making his pitch against Macron and for himself in the legislative elections. Calls for a popular union of parties...built around his previous presidential campaign. Also wants to win a majority and made PM, but he's obviously smoking Hopium here cause the combined left didn't win a majority of the seats in round 1, so even harder to see it happen in round 2.

He is right that people need to turn out and vote in the legislatives, though. For 20 years French people have been idiots to skip the vote that actually matters because the presidential election looks flashier. Now there's fatalism around Macron being reelected and there not being a left option on the ballot, but there WILL be a left option on the ballot in June, and if history is any guide, a lot of left-wing voters will ignore it. We have to hope that Mélenchon can convince them not to.

[Antonio the Sixth]

Is the expectation that Macron's coalition wins the most seats in the National Assembly but loses its majority?

Is it possible these following scenario? A hard scandal hits EM, another scandal hits the right until June, the left wins the majority of the Assembly and Mélenchon becomes prime minister? In this case, would he be the real leader of the country, like Chirac (1986-1988) and Jospin (1997-2002)?

It is absolutely possible - in theory based purely on first-round results it would even seem inevitable - but history shows it's unlikely. Traditionally people who voted for losing candidates idiotically don't even bother voting in the legislatives. We'll see if a sustained campaign by Mélenchon and Le Pen changes that. I'd have to see it to believe it, but hope springs eternal.

[Antonio the Sixth]

Zemmour speaking now. Absolutely pummeled Le Pen then immediately called on a union of the "national camp". Good luck with that, bucko.

[Antonio the Sixth]

With 50% of the vote counted (coming largely from the most rural half of France, as big cities always report last):

Le Pen 27.5%
Macron 26.7%
Mélenchon 18.4%
Zemmour 6.7%
Pécresse 4.7%
Jadot 4%
Lassalle 4% (this is how you know it's a very rural sample)

At 50% of the vote counted, this time, we got around 52% for Macron.

Over the course of the rest of the count for the first round, we had Macron +1.2, Le Pen -4.3, Mélenchon+3.6, Zemmour+0.4, Pécresse+0.1, Jadot+0.6. Extrapolating a bit, it looks like Macron would be set to gain around 4 points, which would get him to 56%. The projections have him significantly higher around 58%, though, so we'll see.

[Antonio the Sixth]

So is this mostly the north of France outside of Paris where the votes have been counted?

Mostly rural France nationwide.

On that point, two more southern departments are joining Lozère:

Alpes-de-Haute-Provence: 51.45% Le Pen, 48.55% Macron. 67% voted for a candidate, and did not abstain or vote blanc.
Cantal: 56.1% Macron, 43.9% Le Pen. 69.43% voted for a candidate.

We're seeing Macron around 13-14 points below his 2017 result throughout the mostly rural and Southern departments that have fully reported so far. So either the projections are significantly off (unlikely) or, more likely, he's lost more ground than average in those places compared to more urban areas. In other words, the cleavage between metropolitan and peripheral France has gotten even deeper.

I guess I should be happy to have another data point for my postdoctoral research, but I'd like nothing more than to see it proven wrong.