> I am curious why they used a logistic regression.
CNAF is not a competitive employer.
French public service is essentially split in two: the State, and the various "Social Security" bodies which manages most of the welfare system (among them, the Caisse Nationale des Allocations Familiales which manages part of social & family welfare).
The "State" side has a very competitive track for stats/econ/ML employees, namely the national statistics office members (INSEE). As the name does _not_ indicate, they work everywhere in the government, not just at INSEE.
So if you're good at stats and in public service, the odds you end up in the CNAF team instead of the statistiscal office of the Social Ministry is about zero.
(emphasis mine) is yet another instance of adding capital letters by English speakers in what seems to me to be random place. I have yet to see an explanation.
In France, the tax is not on the amount of restate but on the inherited income, ie each kid gets taxed on his share. So if there are n kids, the amount of tax paid is nT(K/n), not T(K). Since T is progressive (convex), nT(K/n) < T(K).
Given that the first 100k are not taxed and that there are three kids, only about 14 millions will be taxed at the 45% rate, not 20.
Note that we are talking about each kid making 6,67M€ (four centuries of minimum wage, three centuries of median wage).
If they earned their money through their labour instead of through their birth, they would actually be taxed more (IR + CSG + CRDS).
As for the grandkids, there are a bazillion ways to reduce the tax bill: life insurance is not inheritance-taxed, you can donate money to your kids before your death (about 300k€ per kid every 15 years before it’s taxed - if there are three kids, three grandkids per kid and everyone does a little tax planning you can give 9 grands kids * 300k€ * say give them money at 0, 15, 30, 45 and 60 yo that’s 13,5M€ originally from the grandma that won’t even count as inheritance, etc.).
Still, the untaxed 100K threshold is quite low, it's a shame that a lifetime of work is so heavily taxed. Especially given how the government is spending it...
My personal theory about this high inheritance tax is that it's real purpose is to keep people in a slave-like state in respect to big corporations and the state itself. Having too many people escape the rat race and thus having time to think for themselves or freedom to go elsewhere is seen as a real danger for those in power.
If you look at the effective inheritance tax rate (ie tax paid on actual money flow from dead people to inheritors), even the top 0.1% of all inheritance is around a 10% rate. See figure 2 in https://www.cae-eco.fr/staticfiles/pdf/cae-note069s.pdf. Note the the nominal tax rate is around 30% - the difference is because a lot of stuff is not taxed, or partially taxed.
I think the issue is that French inheritance taxation has both a very high facial rate and massive amounts of exemptions / loopholes. People that think they can grok the fiscal system in a few minutes look at the high nominal rates , not the loopholes.
> In France, the tax is not on the amount of restate but on the inherited income, ie each kid gets taxed on his share.
Yes but the "progressivity" is harsh. Anything over 500 K per kid is already at 30% of taxation and anything over 1.7m is at 45%.
Also for one grandma, the n for the number of kids tends to not tend towards infinity but more towards something like 2.
So although you're right that it's not 45% on 20 m EUR (actual painting seems to be worth a bit more than that but it's to simply) but it's close. Because with two kids and the 100 K EUR excluded, you're still quickly at insane brackets rates.
For one kid I have the actual rate (I now did the computation) at 43% (instead of 45%) and for two kids I have the actual rate at 41.6%.
I understand that you're technically right to talk about n kids but it's a bit disingenuous to talk about "progressivity" of the tax rate when the effective tax rate is likely going to be 41.6% or 43% instead of the 45% I took as a shortcut.
And at 90 years old sadly the inheritance planning ain't going to be very effective. Even her kids are likely already 65 to 70 y/o, so their own inheritance planning won't allow many donation every 15 years.
I think you will find the CAE note on inheritance taxation quite interesting: https://www.cae-eco.fr/staticfiles/pdf/cae-note069s.pdf. Note the the top 0.1% of individuals, who receive on average 13M€ over their lifetime -
Companies can either care about cover letters, or not. Applicants can etiher write genuine, informative letters or generic useless letters.
There are only two possible stable equilibria: a "low" don't care/generic letters or a "high" care / informative letters. If most applicants write generic letters, there's no point for the company to pay attention to it. If companies don't pay much attention to it, there's no point in wrtiting more than the blandest, lowest effort letter.
Sounds like you're in a field in the "high" equilibrium. Most of the software industry is in the "low" equilibrium.
---
Personally I like cover letters (in the "high" equilibrium of course):
- they basically front load a lot of the interviews, with information that come up in the interviews. I'd rather have as much filtering as possible thourgh the cover letter and CV than in interviews, because it is much more costly for both parties to interview.
- people complain that it overlaps with the content of the CV/resume. The question imho is purely one of presentation: do you want to convey information with full sentences and paragraphs or bullet points? Much like I think powerpoint-style bullet point lists are much worse than actual language to convey rich information, I'd rather have actual text than blurbs in the typical resume. Makes it slower to read and thus requite low number of applicants though.
I’ll add that writing a cover letter is articulating why you are applying and interested in a role. It’s a window into a thought process. It shows someone else how you structure your story, your summarizations, etc. Also, easy way to understand verbosity.
I’m not an SWE by trade so don’t have that experience. But it is useful to practice breaking down your thought process into succinct valuable chunks. In a SWE, it might involve explaining an architectural choice or proposing something new. In sales it might be explaining your pitch approach for a prospect.
My background is actuarial. Highly technical. I spent a lot of time writing why I was doing things. Good then to get a glimpse upfront.
If the company releases every year, you can always buy the lastest tech when you decide to buy. If it only releases every four year, unless you just happen to be in need of a new phone the exact year of the release, you're buying (slightly) older tech.
> Rather than giving everyone a significant reason to switch every few years, the industry seems to think that releasing new models with incremental differences every 9-14 months is the best way forward. It's dumb IMO.
There's a significant reason to switch every few years because (current model vs model of a few years ago) are quite different, even if (current model vs last yeast model) are not.
> The term "liberal arts" for an educational curriculum dates back to classical antiquity in the West, but has changed its meaning considerably, mostly expanding it. The seven subjects in the ancient and medieval meaning came to be divided into the trivium of rhetoric, grammar, and logic, and the quadrivium of astronomy, arithmetic, geometry, and music. The modern sense of the term usually covers all the natural sciences, formal sciences, social sciences, arts, and humanities.
CNAF is not a competitive employer.
French public service is essentially split in two: the State, and the various "Social Security" bodies which manages most of the welfare system (among them, the Caisse Nationale des Allocations Familiales which manages part of social & family welfare).
The "State" side has a very competitive track for stats/econ/ML employees, namely the national statistics office members (INSEE). As the name does _not_ indicate, they work everywhere in the government, not just at INSEE.
So if you're good at stats and in public service, the odds you end up in the CNAF team instead of the statistiscal office of the Social Ministry is about zero.
Hence SAS. Hence logit. And so forth.