Okay, let me reply to this thread, because I think it's actually very helpful to explain what is wrong with the argument in the paper I was criticizing 🧵

https://twitter.com/lymanstoneky/status/1892273975977283948

The point @lymanstoneky makes is that, since the longitudinal analysis in the paper only looks at the 2018-2023 period and immigrant flows during that period were dominated by low-crime propensity groups, it's plausible that immigration during that period didn't increase crime.

Not only does it show no such thing, but it couldn't possibly show that, because that's obviously false. Immigration has a large impact on crime in Germany and this paper is just another example of politically-motivated academic malpractice. Let's take it apart 🧵

https://twitter.com/derJamesJackson/status/1891792571682234805

The key finding in the paper is that, when you control for the time-invariant characteristics of districts and year-specific effects that affect all district equally, a change in the proportion of foreigners in a district isn't associated with a change in the crime rate.

This reminds me of the Reykjavik summit, which was supposed to be about making relatively small progress on arms control, but where Reagan and Gorbachev came very close to agreeing to the total elimination of nuclear weapons, to the horror of their respective advisers. Putin and Xi are no Gorbachev though, so I don't think it will play out in the same way 😅 The Gorbachev/Reagan dynamic at Reykjavik is fascinating, because they were both sincerely convinced that nuclear weapons were a moral abomination, but their advisers and allies clearly didn't share that conviction and freaked out upon realizing that they were serious about it.

In other news, it looks like the chickens are finally coming home to roost for Macron, as Le Pen seems to have decided to pull the plug on the government. Not only do we have no plan B to form another government, but we also don't know how we're going to vote a budget 🤷‍♂️ As I've been saying since last June, I'm really not sure that Macron will be able to finish his term. I know that everyone thinks I'm insane when I say that, although that's probably less true now than a few months ago, but I really think people underestimate the probability.

Okay, you're going to say that I say that because I'm a reactionary who hate protesters in general (which again is kind of true though besides the point here), but pro-Palestine protesters have to be some of the most retarded people on earth.

https://twitter.com/TimcastNews/status/1816193966985077106

It's true that the deck is stacked against the pro-Palestine side of this thing, but at the end of the day it's impossible to avoid the conclusion that, if they weren't so fucking stupid and unhinged, things would go a lot better for them.

We're at the point in France where, in most large cities, immigrants are responsible for between 1/3 and 1/2 of reported crime or even more than that, but there are people seriously explaining that Science has demonstrated immigration has no causal effect on crime because a paper using a shift-share instrument in Italy on data that are 20 years old found no statistically significant effect 🙃 We are truly reaching level of midwittery previously thought unimaginable. It's really hilarious that those people don't understand that, when some groups are overrepresented in reported crime by a factor of e. g. 5, if a paper using sophisticated statistical jujutsu finds no effect the rational thing is not to believe the paper.

https://x.com/NoahCarl90/status/1813215305029615850

Les instituts de sondage font des projections assez différentes, mais elles mettent toutes le RN beaucoup plus bas que mon modèle. Qu'est-ce qui explique cela ? Je voudrais explorer cette question, car je pense que ça éclaire les enjeux de la modélisation du second tour 🧵

https://twitter.com/phl43/status/1807784385468309598

Quand je modélise la probabilité de victoire au second tour en 2022 des candidats RN en fonction de leur avance ou retard au premier tour, lisez le tweet cité plus haut et le fil dans lequel j'explique la méthodologie pour comprendre ce que je fais exactement, ça donne ça.

Avec les résultats du premier tour, le modèle donne ça :
- RN et LR-Ciotti : 274 (263 - 284)
- NFP : 126 (120 - 132)
- Ensemble : 102 (96 - 109)
- Divers Droite : 20 (17 - 23)
- LR et UDI : 19 (16 - 23)
- Divers Gauche : 15 (12 - 16)

Autrement dit on se dirige vers le chaos 🤷‍♂️

https://twitter.com/phl43/status/1804405828926988433

Encore une fois, c'est à prendre avec des pincettes, puisque c'est basé sur l'association entre les écart au premier tour entre les candidats qualifiés pour le second et la victoire au second en 2022, mais cette association pourrait être très différente cette fois-ci.

We're at the point where Serge Klarsfeld, the most famous French Nazi hunter (who also wrote a very important book about the role of France in the Holocaust), explains that if he has to choose between the far-left and Le Pen's party he'll vote for the latter "without hesitation".

https://twitter.com/Frontieresmedia/status/1802063786729259515

It's really hard to convey to a non-French person how revolutionary this is. If you had told that to someone 30 years ago, he would have thought you had completely lost your mind and laughed to your face.

So many Arabic speakers have reported that the translation was grossly inaccurate without anyone even trying to deny it that, at this point, I no longer have any doubt that it's true.

You can learn a lot about people's intellectual honesty or lack thereof by looking at who shares this video and doesn't take it down even after the issues with the translation have been pointed out to them. I've read all the community notes and checked dozens of replies to this tweet and others sharing the video, so I know some claim ISIS used the word to talk about sex slaves, but even if that's true that's not a defense of the translation since nobody is talking about pregnancy.

A French journalist asked an IDF spokesman whether they still believed there was a Hamas command center under al-Shifa Hospital, pointing out that the evidence presented after the IDF seized it didn't match the original claims made by Israel.

He replied that it was "unacceptable" for "a democracy like yours to ask another democracy a question" and added that such a question was "extremely weird or even nauseating" (if you see what I mean).

The journalist was almost apologetic and tried to argued that it was normal practice, but the IDF spokesman was having none of it. For some reason I'm supposed to give those people the benefit of the doubt though 🤷‍♂️ In case you have forgotten, let me remind you what the IDF originally claimed was under al-Shifa Hospital.

https://twitter.com/IDF/status/1718010359397634252

I have already retweeted this paper, but I'm doing it again because I think it's fascinating. It suggests that most of the differences in economic policy between mature developed economies that people constantly argue about are second-order relative to demography.

https://twitter.com/MoreBirths/status/1727059812746199301

To be clear, this does not mean that economic policy doesn't matter and couldn't make a substantial difference, because the range of variation in economic policy among mature developed economies is limited.

I'll probably write about this because it keeps coming up and there is a lot to say about the "offer" made at Camp David, starting with the fact that technically a comprehensive offer was never actually made, but for now here's what this amazing "offer" looked like in practice 🙃

https://twitter.com/daddykool/status/1716286595194917178

First, if people are going to make this argument, then they should say that and not constantly repeat the talking point that Barak made a super-generous offer at Camp David. This is a lie.

https://twitter.com/DamonLinker/status/1716472769745916069

The mandatory reference to Munich and reductio at Chamberlain in foreign policy discussions is not just stupid because a comparison with the crisis over Czechoslovakia in 1938 almost never makes sense, but also because people have a very dumb understanding of that crisis. It's not that blaming European statesmen at the time is not appropriate, there is plenty of blame to go around in that episode, but people are always very selective in who they blame and thinking about the crisis primarily in moral terms prevents them from truly understanding it.

This stuff only saves lives *given current incentives*, but it affects incentives by making it more rational for people to get on those boats. If you want to save lives, as opposed to virtue signal, you should either lift restrictions on immigration or do what Australia does.

https://twitter.com/GermanyDiplo/status/1707783245171978250

Of course, the German Foreign Office isn't going to advocate lifting restrictions on immigration if only because it knows that's politically impossible, but it also doesn't want to push for Australia's solution for ideological reasons, so instead it virtue signals on Twitter 🤷‍♂️

In today's episode of "selection is everywhere", many people claim based on observational studies that naturalization improves the integration of immigrants, but this study based on a random lottery finds no effect on either economic and non-economic dimensions of integration 🧵





The study finds no effect whether the effect is defined as intention-to-treat (ITT) or local average treatment effect (LATE), which are both potentially of interest depending on the question you're interested in.

Heureusement la notion de « contenu haineux » n’est pas du tout vague et donc cette régulation ne présente aucun risque d’abus 🙃

https://twitter.com/anonymecitoyen/status/1678317997386084353

Ces gens sont des dangers publics, non pas tant parce qu’ils sont malfaisants, mais parce que ce sont littéralement des abrutis.

"The far-right blames the riots on immigration, but only 10% of those arrested are foreigners while the other 90% are children or grandchildren of immigrants and were born in France, so clearly this has nothing to do with immigration. Look at me and how smart I am." Sometimes I understand better why articles in The Economist are anonymous. economist.com/europe/2023/07…

One can in fact criticize colonialism without ascribing collective guilt to Europeans and making up a parallel universe in which a backward Europe miraculously conquered most of the world despite the fact that it was much more advanced and became rich from the spoils. Someone just asked me "if the colonies were such a burden for France, then how come you fought so hard to keep them?" and, as per usual, the answer is that it was because people believed in stupid memes 🤷‍♂️

J'avais écrit ce fil après la mort de Samuel Paty, mais ce que j'y expliquais s'applique tout autant aux émeutes. Quand ce sera fini, on va nous parler d'éducation, de "plan banlieue", etc. mais tant qu'on ne réduira pas les flux d'immigration ce sera comme pisser dans un violon.

https://twitter.com/phl43/status/1318115587424505856

Ce qui se passe est le résultat prévisible de l'augmentation de la concentration de plus en plus grande de gens issus de l'immigration africaine et maghrébine non qualifiée dans les zones urbaines françaises et ça n'ira qu'en empirant tant que cette immigration continuera 🤷‍♂️

Cette étude est une vaste blague et les soi-disant défenseurs de la "science" qui la reprennent ne valent pas mieux que les mecs qui diffusaient des études bidons en faveur de l'hydroxychloroquine.

https://twitter.com/Le_Figaro/status/1669804790833119235

Ce serait une énorme blague même s'ils avaient appliqué l'idée de base correctement, mais honnêtement je ne comprends même pas ce qu'ils ont fait et je crois que ça n'a mathématiquement aucun sens.