Over a decade ago, researchers started a trial to see if they could prevent peanut allergies
They gave a few hundred kids peanuts from ages one to five and told parents of another group to have their kids avoid the stuff
Peanut consumption reduced peanut allergy rates by a lot:
The initial trial and assessment (ages 1-5) was called LEAP. The follow-up at six years of age was called LEAP-On, and it involved asking kids to abstain from peanuts for a year to see if allergy rate differences persisted.
As you can see above, they did.
By the long-term follow-up, we still see a roughly 70% reduction in rates of peanut allergy.
So this annoying condition (for both sufferers and those who have to accommodate them) can be mostly eliminated by mere exposure.
Unless they're starting off allergic, exposing your kids to peanuts seems like a good idea.
And to preempt a concern: No, the results were not driven by mortality among the exposed.
In 1942, the U.S. government forcibly removed more than 110,000 ethnically Japanese people from their homes and sent them to internment camps in remote parts of the country.
People are resilient, but losing everything is hard.
How did victims' lives turn out?🧵
First, we need background.
Japanese citizens began arriving to the U.S. in the latter part of the 19th century.
The scale of migration was substantial. By 1942, 40% of Hawaii was Japanese (Hawaii wasn't a state until 1959).
This influx of immigrants quickly became a political problem.
1886-1911, more than 400,000 Japanese set out to American lands. Citizens called for an end, resulting in the Gentleman's Agreement of 1907:
The U.S. wouldn't harass its Japanese and Japan would restrict emigration.
2024 is the hottest year on record, and it's been hotter than 2023 in part because of a global ban on shipping fuels containing sulfur dioxide.
Problem: SO2 causes acid rain, but it cools the globe. How can we just stay cool?
A new company might have found the solution.
🧵
Acid rain has been on the decline for many years, but in order to finally put the problem to rest, it'll be crucial to knock out sulfur dioxide emissions from shipping.
Globally, those emissions have been concentrated in these boxed-in regions where ships go to-and-fro.
When the International Maritime Organization 2020 regulation went into effect, roughly 80% of sulfur dioxide emissions from international shipping went away overnight.
FDR was the Assistant Secretary of the Navy from 1913 to 1919.
During his tenure, he supported a gay entrapment operation where sailors would allow themselves to be propositioned by gay sailors so the Navy could identify them and kick them out🧵
After enough intel was gathered, twenty potentially homosexual men were rounded up and brought aboard the USS Boxer to be interrogated.
Afterwards, fourteen of them were charged with gay sex acts with the sailors that entrapped them.
When the Senate investigated FDR over this, they were furious.
They wanted him barred from holding public office because he was responsible for a bunch of innocent young sailors being sodomized.
Since he ultimately went on to become the president, we know he got off scot-free.
Here are the current best estimate of income inequality in the U.S.
🧵
If you look at the top 10%, their share of income seems to be flat after taxes and transfers.
The original Piketty-Saez estimate is based on failing to count for taxes and transfers and including capital gains, which are not part of national income, among other issues.
If we look at the top 1%, we see something similar.
Note that "Piketty-Saez" is called "Fiscal Income" here.
It shows the meta-analytic estimate of the effect of getting an additional year of education on people's IQ scores.
It almost-certainly depicts a major overestimate🧵
"Control Prior Intelligence" refers to a design to overcome selection into higher education.
The big reason why you cannot regress years of education on IQs to understand the IQ-boosting effect of another year of education is that kids with higher IQs tend to get more schooling.
So, to get to the effect of schooling, this design has you control for a measure of earlier IQs.
But here's the problem with this: if that early measure of IQs is measured with error, then you're not controlling enough.