We have long been moving into a world of increased convenience and pleasure, at the cost of strong and local tight-knit communities — which has led to profound isolation and loneliness.
In this essay, I argue that we must find ways to provide children with vibrant
The teen mental illness crisis is not just concentrated in the United States. I explain how and why it hits Gen Z girls across the Anglosphere around the same time and in the same way.
My first of three posts w/
@JonHaidt
, Part 1: The Anglosphere.
@Scientific_Bird
Four things here:
1. These trends are not split up by sex. Our argument is sex-specific.
2. the WHO GHE (Which is the dataset being used here) underestimates known rates of suicide. So, even the U.S. data after 2010 is not correct here. (and it is the case for many nations
Nordic teens are struggling very much like Anglo teens. I have looked at 10 countries, and they all show the same pattern:
Teenage girls are struggling more than all other age groups and much more than they were a decade ago.
@JonHaidt
&
@ToMpy1337
A fascinating perspective by
@DKThomp
A mystery that I have also been working to make sense of.
In addition to Thompson's views, I argue that teens from Anglo nations (which are generally highly individualistic and secular) tend to have weaker in-person social ties and
@JonHaidt
@DKThomp
Also, one way I think about the variation is by Imagining two different kids:
One is born in a low-income area of India, has six siblings, has to walk to school every day, is relied upon in the home, has a mobile phone with spotty access, and their family is deeply religious.
It's time to reclaim life in the real world. Join
@freyaindiaa
and the FreeTheAnxiousGeneration team for the Phone Free Summer challenge. One day a week, starting June 14.
@JonHaidt
Video features the great
@FreeRangeKids
Learn More:
@EricLevitz
An important part of the story, I believe, is we need to go back further. I have data from Australia, which shows the same trends. But as you can see, it's not reverting to an older mean-- the 80s were particularly bad, for different reasons than they are now.
Why are youth self-harm rates declining in Denmark, while self-reported rates of depression and anxiety on the rise? I discuss in my updated post on Nordic youth mental health.
@novonordiskfond
@JonHaidt
@Luke_OBrien
@JonHaidt
@NickHanauer
Hey Luke, are you still having this issue? There shouldn't be a paywall up anymore. If there is, will make sure this gets fixed asap.
Along the way, we have learned that the mental health of liberal girls fell first and fastest in the early 2010s. Our article on the reverse CBT hypothesis.
@JonHaidt
Some argue that the adolescent mental health crisis is not international. In this post,
@JonHaidt
@ToMpy1337
and I show that European girls are doing worse than they were a decade ago—especially those from the least religious & most individualistic nations
The phone-based chilhood is making kids sick and needs to end now. Social media companies ensare teens into a social trap that they cannot get out of. It's time to change course.
The Anxious Generation in one article.
@JonHaidt
@JonHaidt
@CJFerguson1111
@PeteEtchells
@cjsewall9
I also want to push back on the self-harm claim. It seems like we are using a single study of New Jersey Hospitals to speculate that most of the rise is self-harm among girls in the United States or, even, in other countries has nothing to do with actual increases?
There is so
@JonHaidt
@voxdotcom
@EricLevitz
@JonHaidt
To add to this, there are a few key things missed from the article:
1. The WHO GHE suicide estimates have been *underestimating* known rates of adolescent suicide across many nations *since 2010.* This is because they are estimating based on previous years.
When we
It is time for schools to go phone-free. Smartphones undermine attention, learning, and relationships.
It's like we give each child a portable soda dispenser 24 hrs/day, 7 days/wk. Let's give them 7 hrs of healthy childhood back.
@JonHaidt
@JonHaidt
And it's not only self-reported rises in poor mental health. Gen Z girls in the U.S., UK, Canada, Australia, and New Zealand have higher rates of suicide than any prior generation of girls at that age for which we have data for.
@JonHaidt
@cjsewall9
@JonHaidt
We can see the specific changes in rates for ED visits by year here. The hospitalization data does not allow us to see what exactly happened between 2009-2012. But the ED vists do let us see into that.
Guest post from one of my favorite psychologists, Jean Twenge, who is never afraid to dive head-first into a debate.
@jean_twenge
@JonHaidt
Why Academic Pressure Cannot Explain the Mental Illness Crisis
@cjsewall9
@JonHaidt
Oh yes. I actually cite this in the article.
The pediatrics article is covering all kids ages 3-17. But we can get some parsing of the trends by age/sex.
In 2009, there was an estimated 82,887 girls (ages 3-17) hospitalized for mental health conditions. By 2019, it is 123,342.
@StefanFSchubert
@JonHaidt
@SteveStuWill
Yes, we plan to address suicide rates in an upcoming post. Suicide rates do differ across nations. We will address our explanations for this soon.
@timmyhotboots
@JonHaidt
Hey Timmy, I'll be discussing non-Anglo countries in a future post, but check out our page:
This has a selection of google docs that collect studies on this topic across the globe
Factories in the early 20th century faced a collective action problem, risking safety for competitive advantage. Today, social media companies face a similar dilemma.
@JonHaidt
@HumaneTech_
& I propose solutions the collective action traps we face today
Collective action at work. Jewish day schools around the country are taking the lead, bringing their communities together, to set new and better norms around tech use at schools and in their communities.
@KarlPettersso10
@GunnarBj
@KarlPettersso10
@GunnarBj
I'm actually working on a piece for Scandinavia next. Here's the google doc I'm working off of. I'd be thrilled if you want to take a look and add anything you think I am missing:
@post_liberal
@JonHaidt
Yes! There are discrepancies in Denmark between reports of anxiety and depression and self-harm rates (the self-harm study you cite goes up to 2016, so I do not know what has happened since).
I address this in the post. Let me know if you have thoughts about the discrepancy!
@cwarny
But to nuance this: not all of us have adapted well to a world of ubiquitous sugar (or other mass changes in our food industry)... I also think there are meaningful distinction to be made between kids and adults.
An important part of the story, I believe, is we need to go back further. I have data from Australia, which shows the same trends. But as you can see, it's not reverting to an older mean-- the 80s were particularly bad--for different reasons than they are now.
@EricLevitz
One theory that might explain *part* of the teen mental health crisis is that we're reverting to the long-run mean, after a period of exceptionally low levels of adolescent mental distress (tho the suicide rate among girls specifically is at all time-high)
@cwarny
@balajis
I agree that it would be better to keep it voluntary, but I think you underestimate broad compliance when people face what is perceived as existential threats.
@cwarny
@cjsewall9
@JonHaidt
You are absolutely right. This is something that I want to improve on, and will try to think of ways to best do this in the future, with the skillsets I currently have (and what I hope to develop). Thanks for this feedback!
@JayMan471
@Scientific_Bird
It’s true that it is the most objective measure but that does not mean that other measures should be discounted.
If we dismiss other measures because they are imprecise we miss an enormous amount of information about sub-populations.
Self harm, like suicide, is also a
@Brunobhjf
@JonHaidt
Wow, really interesting. We would love your help finding studies and research from South America. I have had difficulty with this, and it would be of enormous help.
If you request comment access, you could add in your thoughts / comments in the doc. Great to connect w/ you
@sergiotpinto
@paulnovosad
@mikekofoed
I really appreciate you reading through the post.
I have looked at England and Australia (just for suicide rates) which shows similar underestimates:
"In January 2018, Mr. Zuckerberg received a report estimating that four million children under the age of 13 were on Instagram, according to a lawsuit filed in federal court by 33 states."
@EricLevitz
@JonHaidt
@EricLevitz
Hey Eric,
Overall, I really appreciated this essay.
But I want to note that I address many of these arguments around diagnostic changes in this post.
I also show major issues with the WHO GHE data is underestimating rates of suicide among teens across many
@cjsewall9
@JonHaidt
The issue is that the CDC does not provide yearly data for the 10-14 year-olds for hospitalizations. The closest I can get for you is 2013 and 2017, which is a rise of 56.8%.
I have more data point for ED visits. So from 2012 to 2019 for 10-14, the % increase was 91.3%.
Just
@JonHaidt
And one of the reason people disagree on the international scope of the problem is that they are relying on a dataset that systematically underestimates rates of youth mental illness in countries that have robust national mental health data.
@MaxCRoser
@jmhorp
Hi all,
I want to first apologize for the confusion in my article. Max is right that I had muddied the waters by combining his graph with my arguments about the GBD.
I have made updates to the article that I hope make it is accurate as possible.
But I do not understand:
1)
"We are wired to believe our senses, especially when they converge. Illusions, historically in the realm of curiosities, may soon become deeply woven into normal life."
@sergiotpinto
@paulnovosad
@mikekofoed
Hi Sergio,
I am deleting the previous tweet because it was not fully accurate.
The trends in the OWID graphs you have above are *not* relying on the GBD. OWID relies on WHO estimates, which rely in part on the GBD (But only for countries without good death registration data).
@StefanFSchubert
@JonHaidt
@SteveStuWill
Agreed, I wanted to be consistent with my other graphs by including the relative increase. I do not make the case that self-harm rates in Sweden have increased in a way that supports our hypothesis. But the rise during COVID, to me, seems evident.
@cjsewall9
@JonHaidt
Just to respond to the first point here. My approach has been to average a few years before 2010. The CDC does not provide rates for the 10-14 year-olds in 2009. So I averaged 2007-08 together, to try to find a relatively stable base rate that can look at pre- and post- 2010.
@rubenarslan
I actually have the simple mean comparisons. You can see that in this google sheet:
I also provide a lot more information in the supplement for the post:
I chose to do 3+ because that has been a general standard in published
@post_liberal
Additionally, there are multiple plausible explanations for declines. I invite you to add your thoughts and comments in the respective google doc I have been curating: