But there's an even better way in. This table (.csv format) has all the protected designation products listed out by commune. And of course we have election results by commune. So it should be pretty easy. data.gouv.fr/fr/datasets/ai…
Stick it together with that in Excel.
Results are in. Top AOC for Macron is from the commune of Le Clat (11140) a tiny village in Aude département, in Occitanie, in French Catalonia. 43 registered voters. 25 voted for Macron, 1 was a blank vote, 0 for Le Pen.
Oops, got my numbers wrong there. only 31 registered voters. 25 voted for Macron. 1 Blank. 0 for Le Pen. Anyway. The rest is right. But thinking about it, maybe I need to up the quality. Require a top tier AOC. I'll do Le Pen's now and think again.
Flammerécourt (52110) in Grand-Est. Quality sparkling wine. 57 registered voters. 48 votes. 1 for Macron (absolute LOL at this village trying to figure out the splitter 🤣). 1 spoiled. 46 for Le Pen.
France is varied eh?
This will need a rethink.
Right. We go again. This time a map. Much less efficient, but built in vibe checks make it less likely I'll mess up, and it should help me refine the question. And this time I'm just looking at AOC wines. But where is Champagne?!?! data.gouv.fr/fr/datasets/de…
Okay. This is another way in. Get each AOC Wine, add up all the votes for its communes. That way the two winners are ones I can actually buy. And who cares about Champagne. I'll figure out why that's missing some other day.
Looking good. Especially for Le Pen. Crikey. Some BIG names in there. Saint-Estèphe, Chablis, Châteauneuf-du-Pape, Pauillac. It's as we'd guess. The Loire for Macron. The Rhône for Le Pen.
I'm going to have to pick ones I can purchase (so not Clairette). So I'll go red. Le Pen's two are gonna be Cotes Du Rhone Villages Plan De Dieu and a Saint Estephe (£££s eek). Macron's --- oh dear,...
I am not to be able to justify that purchase I don't think. Understanding this "President of the Rich" thing now TBH.
Yeah that's not happening Macron.
A lot of prime Burgundy is Macron territory, I didn't realise. I guess it's rich people. The Loire being Macron was expected. I'd have though Bordeaux would be too, but Le Pen is strong in some parts. I think I'd best put this on a map.
People asking after the Crémants. Crémant de Loire is your Macron choice. Crémant de Limoux (the original sparkling wine) is your Le Pen choice.
Well there we are. Yet again with apologies to Champagne, I'll try and fix you another day. I'm using the standard colours, Yellow = Macron, Le Pen = Dark Blue. Here around Angers is dominant Macron wine country. Chablis goes for Le Pen. Big choices.
The lower Rhône is all Le Pen, but Clairette de Die just a bit higher up and off the river a bit is strong Macron. As for Bordeaux, well.
The former bulk wine areas in Languedoc are very Le Pen. Perhaps a desire for protection against cheaper Spanish bulk wine. It's an anti-globalisation sentiment that I can understand. French wine growers want French to drink French wine like in the past. theguardian.com/lifeandstyle/2…
The decline in French alcohol consumption is just astonishing. The average French person drinks less than a quarter as much wine today as they did in the 1960s. And look at that cidre decline.
Did Le Pen promise to bring school wine back in her manifesto? I have my reservations, but I think it's important for cultures to find common ground at this time so it may be worth investigation?
Well, well, well. UK drinking trends are nothing like French ones. Astonishing. jrf.org.uk/sites/default/…
Armed with my tables and maps from yesterday, I shall now go to @latitude_wine and try and buy their most Macron and their most Le Pen wine.
It's very important to ship early, get feedback, digest that feedback, and make improvement. Agile or something. And what I've learned from the first two iterations of this is that I need to extend my methodology to the first round too. And also to French Rhum.
I also still need to think through how I map overlapping AOCs. And how I do the labelling when there's so much repetition. Do you like wine @undertheraedar? I might need to call in a favour.
@undertheraedar But first the data analysis. We're now beyond the Excel and QGIS stage. This needs reproducible and fast analysis (so I can do lots of rounds of analysis). For most people that means Python or R. For me that means C# .NET in Visual Studio. It's what I'm used to and it's fast.
@undertheraedar One of the reasons I love @datagouvfr -- the discussions are fantastic. Here, someone has asked the same question I had about Champagne, and the data owner has replied with an explanation and a promise to work on the solution. Really good stuff.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
The bus use graph in the FT bus article is fantastic. Contains the absolute minimum information to make the correct point clearly and honestly. ft.com/content/cce174…
Here is the same data by me and @neilmcclure100 from our 2017 submission to the House of Commons Transport Select Committee Bus Market Inquiry. It addresses most criticisms of @jburnmurdoch's graph today, but it would be so bad to put in a newspaper. docs.google.com/document/d/1IZ…
There's just too much data on it. Trends before deregulation? Not useful for a newspaper. Absolute values not percentage change? That's a lot of numbers no-one can parse. Precise table number for the source? You want that on your mobile phone do you? Really?! You don't.
My impression of Bristol is that there aren't a huge number of empty homes. But what do locals think?
I ask because of how many of the stories I've heard recently, on the radio, on here, in writing, have been about Bristol. Huge competition to rent and buy.
I looked up the annual population survey household count estimates for Bristol (noisy data because of the small sample at LA scale). And the dwellings count. Subtract one from the other and you get "surplus dwellings". I've written why this is a bad idea. tomforth.co.uk/emptyhomesforg…
Some context on the "doesn't France have great elections data?" findings many have made in recent days. They have excellent data in lots of other ways too. And this is a big part of why. Very very very different from the UK experience of recent years.
The way that France achieves excellent data is very different from how the UK does. And surprisingly controversial to discuss within England.
Best example of where the French approach is worse is public transport timetable data. We have it for everywhere in Great Britain in a single place in a single format. France has a patchwork of local data releases in different formats with some missing. transport.data.gouv.fr/stats?locale=en
I always quite like it that (by my proprietary methodology that I can't prove in public isn't rubbish) one of the industries that is strongest in London compared to the rest of the UK is fashion. Not financial services. Fashion.
There are some specialisms within financial services in London too I hasten to add. But other places do them too, often just at a diffent scale. But there are some bits of the fashion industry (not the manufacturing) that are very London in a way that few other industries are.
Your Burberry coat still gets made in Keighley and Castleford. But other bits of it, dare I say the "intangible" bits of it largely get done in London, almost uniquely within the UK. burberrycareers.com/content/Manufa…
Shout out to @mathieugallard et al. who it feels like have called the French election right all along.
Thanks in advance to the French data journalists awaiting the official results overnight to analyse it all. Looking forward to reading all about it tomorrow.
And best wishes to the vast majority of decent politicians and strategists in France who've got a few years now to try and figure out what can appeal to the vast majority of French who voted for Le Pen who aren't that racist -- and sell that before Macron's final term is up.
Fun fact, The Economist Intelligence Unit usually scores France as a flawed democracy while the UK is a full democracy. One reason why France scores worse is lower turnout in elections. Last UK general election, turnout was 67%. Keep an eye out for France's turnout today. 😏
The EIU use (or at least they used it last time I checked the methodology in detail) turnout in national parliament elections, which are much lower in France, which is why the UK scores higher.
The EIU don't use turnout for the House of Lords elections of course. 🙃