Post

More from @lymanstoneky

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Dec 2

https://twitter.com/DrEmmaZang/status/1995655529608167790

reasonably confident this is a yale demog prof publicly getting a fairly common quant demo problem wrong very publicly

i could be wrong, but the probability nerds in the comments and the people like me doing simulations all seem to agree that it's 50/50

https://twitter.com/DrEmmaZang/status/1995655529608167790

okay, i see what @DrEmmaZang is arguing, but i think this is not a case of some kind of clever problem design but just a lexical problem.

the problem clearly asks what the ratio will "eventually" become. that is, towards what number is it converging. and it is converging to 0.5, asymptotically. hence the 0.5 answer everyone is giving. the question is literally asking the asymptote.

the correct answer to "what will eventually happen to the fraction of girls" is "it will trend towards 0.5"

now, at any given time, it can be above or below 0.5. @DrEmmaZang seems to believe (and FWIW Grok agrees) that it will always asymptotically converge from above, so any "real" society with these rules will be >0.5

but i trivially falsified this. across a bunch of simulations of n=500,000, much bigger than any "primitive" society we might imagine from the question prompt, i had tons of cases where the realized proportion was <0.5. i think the average of the simulations was probably around 0.5002 or something-- but even at numbers much bigger than is plausible for the question text, the simple fact is that you can't even guarantee convergence from above. so the answer "the share of girls will be somewhere asymptotically above 0.5" is not correct; it's easy to generate simulations where this isn't the case.

FWIW, i've literally seen a version of this problem (tho for boys instead of girls) in demo homework, and the correct answer was indeed 0.5

so I think what's going on here is 1) @DrEmmaZang misread the question and didn't notice it's actually asking about the asymptote ("eventually...") and 2) given the "primitive society" part the notion that we should assume large numbers apply isn't even correct to begin with

the correct answer is clearly "it will generally be about 0.5 girls with an asymptote at 0.5." the fact that the expected value at any specific finite number may be 0.500001 is irrelevant since, for any finite number in a primitive society, the variance will be comparatively enormous.

for reference, here's 30 simulations of 100k families. you can see that there are plenty under 0.5. for the 30 100k simulations the average is actually 0.500034, which is below the expected approximation of 0.508. nor was it even converging anywhere close to 0.508 actually.

Read 10 tweets

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Nov 13

Me and @bobbyfijan have argued that to get more families in America, you need family-friendly housing.

Today at @FamStudies , I show further evidence: first, from a new study showing how house size shapes fertility; second, in the YIMBY case study of the Chicago Loop.

A new study uses data on movers and fertility to estimate how housing costs and home sizes influence fertility. The takeaway is: they both matter!

This is what we've argued at IFS: YIMBYs tend to be laser-focused on boosting supply to reduce cost, while ignoring the size issue.

What's striking is the new study shows that although "YIMBY for family-friendly units" actually reduces prices by less than "YIMBY for small apartments," it actually increases fertility by twice as much.

Size matters when it comes to babymaking.

Read 12 tweets

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Sep 30

A few years ago I was chewing on a graphs like these ones.

Apartment-dwelling is rising over time. But the evidence suggests that apartment-life is not great for family formation. It's hard to add SFH given land constraints, commute times, etc. So what to do about housing? 1/🧵

The first and most obvious step is just: remove any obstacles that do exist for more dense, young-family-friendly SFH. We wrote a big report on that topic at @FamStudies back in March. We tackled affordability, how to get more dense starter-home neighborhoods, crime, etc.

But as I was chewing on this topic back in early 2024, I had a chance to meet @bobbyfijan at an event organized by Steve Teles supported by @Arnold_Ventures about housing. We realized that we had a common interest: solving the "family apartment problem."

Read 14 tweets

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Aug 6

I AM ONCE AGAIN BEGGING THE CDC TO FIX THE OBVIOUS ISSUE HERE

I *think* the official vital statistics data is based on actual gestational mothers, but I'd like CDC to clarify that in public.

In the meantime, birth certificates need to be listing legal, genetic, and gestational parentage ON. SEPARATE. LINES.

Children have a right to know their full parentage, we should compel it via the force of law to be listed on their birth documents.

Read 15 tweets

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Jul 21

Are you online? Then you've probably seen the takes: rich men should just marry a pretty, submissive Applebee's waitress. There's a whole genre of tweet that seems to fantasize about highly available food service workers.

I decided, at my peril, to take it seriously and test it

Who is right? The online Waffle House Fantasists, or @CartoonsHateHer 's pro-girlboss takes?

In today's post at @FamStudies , I argue.... kind of neither!

To start with, credit where it's due: the pro-girlboss take from @CartoonsHateHer stands on a solid foundation of decades of work on assortative mating, which I replicate. The richer you are, the more you assortatively you mate!

Read 16 tweets

Lyman Stone 石來民 🦬🦬🦬

@lymanstoneky

Jul 16

SO this actually isn't what's going on.

What's going on is way more interesting!

UNIQLO is vastly further up the retail foodchain in the US than it is in Asia.

KFC and McDonalds are way fancier in Asia than America.

Why?

Because the export versions are always the best versions.

Exporting intrinsically creates costs: transport, transactions, often tariffs. As a result, exporting is rather challenging for most firms, which is why most firms do not export products.

Firms that do export products are entering a larger, more competitive space, so have to compete harder.

And to justify the cost of export, they end up having to move upmarket vs. their home market product. It's rare that the export-version is worse than the domestic-version.

It really is true that foreign McDonald's is better!

You can get respectable Macanese egg tarts at KFC-Hong Kong!

Read 7 tweets

Share this page!

Enter URL or ID to Unroll

Lyman Stone 石來民 🦬🦬🦬

Try unrolling a thread yourself!

More from @lymanstoneky

Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬

Lyman Stone 石來民 🦬🦬🦬

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!