Share this page!

Enter Twitter Thread URL to Unroll

Needs to be the full URL like: https://twitter.com/threadreaderapp/status/1644127596119195649

How to get URL link on Twitter App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Crémieux Profile picture
@cremieuxrecueil
Jun 13 • 4 tweets • 2 min read Twitter logo Read on Twitter
Accounting for selection into disadvantaged neighborhoods, the effect of neighborhood disadvantage on the likelihood of committing a violent crime is nearly nullified.

The cross-sectional effect and the causal one do not match because neighborhood effects are largely selection. Image
The same is true for property crime.

Your neighborhood's poverty and crime rates do not seem to drive your own risk of committing a crime. Image
The within-person results for violent crime showed a peak odds ratio (1.22) at disadvantage deciles of 6 and 9. For property crime, the peak OR was 1.09 in decile 9 and it was p = 0.05. For other crimes, it was 1.08 in decile 10 and it was marginal (95% CI: 1.01-1.16).
Account for multiple comparisons and the within-person results all disappear.

Neighborhood effects are about selection more than they're about neighborhoods affecting people.

Source: sciencedirect.com/science/articl…

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Crémieux

Crémieux Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @cremieuxrecueil

Crémieux Profile picture
Jun 14
Japanese labor use is horrifically inefficient.

Here's an example: Do you need three people whose job it is to tell singular cars to pass? Do you need five restaurant or grocery store workers standing around?

If you answered "yes", then you may live in Japan or South Korea. Image
This misuse of labor is also why I think there's a popular view that "whatever you can do, an Asian kid can do better." The videos you see of Asian kids specializing at some job to an insane degree are the result of a horrendous misallocation of valuable human capital.
The cultural distance between Japan and the West when it comes to work is staggering. Everyone should read Patrick McKenzie's essay on the topic: kalzumeus.com/2014/11/07/doi…

Here are some choice bits. First, have you heard of salaryman overtime? Image
Read 11 tweets
Crémieux Profile picture
Jun 12
The disproportionate neediness of a few people is why experiments that are plausibly pop-representative like RAND's, Oregon's, and Karnataka's generate null effects on health, and why quasi-experimental evidence on things like Medicaid expansion appears to show benefits. https://www.overcomingbias....
In the general population, care people won't go out of their way to obtain because of cost constraints has a minimal value.

When people are given free insurance and they start going out of their way to obtain more care, it's usually superfluous and some of it's harmful!
Quasi-experimental studies that show benefits to the expansion of programs like Medicaid cannot be generalized except to similarly disadvantaged populations because their effects are driven by affecting people who are disproportionately needy.
Read 4 tweets
Crémieux Profile picture
Jun 12
It should be obvious that environmental and genetic correlations are dependent on the level of heritability.

It can therefore be important to obtain correct heritabilities to understand how environments operate. Image
I chose these parameters because they match the relationship between parental SES and children's IQs.

Those two quantities are typically related at up to 0.35 and the genetic correlation is usually around 0.72. If SES is 50% heritable and IQ is 50% heritable, then all
of the beneficial part of their relationship is due to genes because the resulting environmental correlation is r = -0.01.

If IQ were 20% heritable, the environmental correlation would be broadly good at r = 0.20.

But if it was 90% heritable, it would be very bad at r = -0.56.
Read 6 tweets
Crémieux Profile picture
Jun 4
This behavior seems to be universal, and I believe it may explain the utility of sovereign wealth funds for reducing inequality.

A recent paper on Australia's choice to allow pension withdrawals during COVID provided relevant evidence.

Who withdrew their pension when allowed? Image
Australia needed fiscal stimulus, so they let citizens pull up to $20,000 from their retirement accounts. These accounts, which are funded by a 10.5% charge on wages for all workers, are typically locked down until retirement.

When people were allowed to, they spent the cash. Image
Around one-quarter of people aged >34 chose to withdraw from their accounts when they were eligible. There were two waves of eligibility in which you could take out up to $10,000 both times.

Do you see what amount people tended to take out? Image
Read 12 tweets
Crémieux Profile picture
Jun 2
I actually saw this as a poster and chortled a bit because it was the first time I had ever seen a p-value of 0.967 treated as consistent with finding an effect.

There are a few things to note about supposed cortisol effects. Image
Aizer, Stroud & Buka did another study of the effects of in-utero cortisol exposure.

Luckily for me, I had their dataset, so I could output results for a latent g too! I think I did the controls the same, but I used all the available data instead of a subset. Image
Malanchini et al. found that cortisol measures were poorly related to one another. They also found that their levels varied throughout the day in reasonably consistent ways.

There are aspects of measurement that results can hinge on! Image
Read 6 tweets
Crémieux Profile picture
Jun 2
Psychosis greatly elevates the odds someone commits a homicide.

This is especially true for the first episode of psychosis, in which the homicide rate is 14.45 times greater than after treatment.

We have ample data that shows treatment works. Image
Herttua et al. found that when people were on their antipsychotic meds, they were 40% less likely to be suspected of a violent crime. Image
Sariaslan et al. found that when people were prescribed antipsychotic, their risk of crime perpetration was reduced by 33-50%. Image
Read 7 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

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

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

Send Email!

:(