I know we neurodiverse, introverted people spend time on social media because we feel more comfortable. The more depression and anxiety I have - or my son has - the more we don't want to go out. Being able to socialize in an online group of other neurodiverse people who "get it" is very important. On the other hand, if you're concerned about your child's social media use, there is no harm in having a conversation with them and sharing your concerns.
I agree with Peter's conclusions, though. School has caused real trauma for people in my family. Periods of high social media use did not seem to change things one way or another.
One thing I would have liked to see in context here would have been more information about what the correlation between reported rates of depression and anxiety are to the rates of actual suicide. I was given to understand that they tend to differ substantially, particularly between the genders, and it would have helped with parsing some of the arguments here to have that information available in the text. I realize it's not true of everyone, but as the mother of young boy, I've discovered that lately I care a lot more about actual suicide rates, than the number of teen girls self-reporting being anxious. In some ways I think the latter skews the way we interact with data along these lines, but I'm not an expert, just a worried mom.
Good point, Eleanor. I discussed the gender difference in Letter D2. I've focused on suicides throughout these D letters because measures of anxiety and depression have varied from study to study and time to time, so it is not possible to chart rates of them over the 70 year period I have been considering with a consistent measure. However, concerning this letter, the research shows little or no effect of social media use on any measure of anxiety or depression used in such studies. Suicides, of course, are too rare to use as a measure in such research. As I pointed out in Letter D5, suicides go up whenever school is in session and down when it is not in session, and this is more true for boys than for girls.
I’ve wondered the same as a mom of 3 young boys. It seems there are a decent amount of impulsive (is that the best word?) suicides. They lack a history of depression and their decision is made around an incident rather than a something they’ve been planning for a while.
Do you think social media might be a catalyst? That is, it makes a good youth-culture better and a harmful youth-culture worse. The effects would cancel each other out, preventing if form showing up in the statistical research. Or do the studies control for that?
Imagine a mixed age play culture vs. a same age culture. Might social media have a completely different effect on those two culture?
Good point about social media being a catalyst. I will try to bring that into a future letter. I will discuss what seem to be the benefits and the harms of social media use for teens and suggest, as you do, that they may cancel one another out in the statistical data.
And as Jonathan Haidt point out; most of digital technology have a net positive effect. He mentions; Skype, Teams, Phone calls, texting, multiplayer videogames
Changing kids life to revolve around posting pictures of themselves, in order to get likes and go viral, is bad. So it's the hyper viral platforms spreading this bad youth culture we need to fight. The hyper viral platforms such as twitter, instagram, facebook, and tiktok. Fighting the unhealthy communities on these platform are one way to decrease anxiety. Increasing free play is another.
I look forward to a time where kids have the skills to enjoy free play on twitter, instagram, facebook. and tiktok. But right now, they're getting trapped in a game of social approval.
Here is Haidts on how social media / hyper viral platforms, harms the youth:
Haidt also point out that social medie isn't like sugar, where you only affect yourself. But it also have a network effect, that is like passive smoking or driving too fast.
He points out, that experimental studies need to take this network effect into account. If you stop using facebook for a month, but all your friends continue to use it. Then you'll still live in a culture influenced by hyper viral platforms.
Jacob, the poins you (and Jon Haidt) make here are reasonable and I think are valid to a considerable degree, but difficult to test with research. I hope to take them into account in a future letter. We do now live in a networked world. It isn't going to work to try to reverse that or take it away from kids. To the degree that it is harmful we need to learn how to minimize that. To the degree that it is helpful, we need to learn how to maximize that. My main disagreement with Jon os that he would have us deprive teens, maybe especially young teen girls, of this tool, whereas I would have us talk with teens about about their uses of social media and discover how it may be helping and harming them, in ways that would make them (and us) aware of dangerous versus safe ways of using it.
This is what I wanted to ask as well. Could social media use by my classmates make my life worse even if I am not on it? Perhaps by increasing judgmental attitudes about appearances among my peers, or by me being left out of social groups or missing new fashions because I am not on social media, by increased bullying behavior, etc. “Culture” is hard to measure but some of these specific hypotheses could be.
Great article and on a critical topic. My worry is that the cat is already out of the bag. Meaning, studies that evaluate whether someone is happier reducing screen time may be missing the broader point that the test and control groups already were big users to begin with. So taking away from the test group may not show much of an improvement or may be influenced by the other factors you note. I have heard that looking at data pre and post 2012 (or whenever the real launching off point for the iphone was) is a better comparison and that since that time, Americans are far more lonely than they were before. Someone else referenced the catalyst effect...I agree with this. If you're happy anyway this makes you more happy and you see happy things. If you are not, this points out all the areas you are missing and how much fun others are having. Regardless, the cat is out of the bag, IMO, and I'm not sure how we ever go back.
The article is astounding! I'm a bit surprised for the lack of evidence with social networks. How can we link that to the fact that Facebook revealed serious connection and design orientation of Instagram with addiction, anxiety, etc?
Peter, the focus of these studies seem to be connecting the use of social media and digital technology as the cause to increased anxiety/depression. Could it be that it is not social media or digital technology per se causing the mental health issues but instead what they are missing out from by using technology (i.e. face to face interactions) that is causing the breakdown in mental health? I am trying to keep up with all the news letters so that is a topic that may have been discussed already. I just feel I have witnessed over the last 25 years as a school based SLP, the decline in free play, along with the increase of academic demands that are not developmentally appropriate (both topics you have covered in depth!) along with a push to 1:1 technology in schools. Kids are now staring at screens all day, then going home to doing more homework on computers, playing video games or interacting on social media, all which drastically reduces the amount of face to face interaction that we all need as social beings! Could it be related to the way kids are getting a dopamine release from technology versus face to face interactions? As with many hot topics, I feel the answer is always more complex! Thank you for your great articles!
Yes, IF social media (or screen time generally) were having an effect on mental health that effect might well be mediated by a decline in other activities ("What they are missing" because of time on screens), but the point is that there is no good evidence to date that time on screens (or on social media) is having an effect on mental health. if there were an effect it would be interesting to speculate on how that effect might be mediated.
I am so grateful the research that you and others are doing on this topic. Has anyone looked to see if additions to the vaccine schedule starting in the year 2000 and moving forward could have had any impact in mental health? What about the introduction of the vaccine Gardisil in 2006? I tend to feel that the rise in mental health issues may be due to a combination of many factors, not just one.
Hi, and thank you for this Substack and all your fantastic work!
The edge of the academic discussion regarding digital seams now to be way beyond "screens are bad" - the leading hypothesis instead being that social media (not digital media in general) is affecting girls 10-14 (and not people in general). This is championed by several, but maybe most prominently by Haidt and Twenge detailing this argument backed up by an overview of a couple of hundred articles: https://jonathanhaidt.substack.com/p/social-media-mental-illness-epidemic. Haidt then asked people to criticise him which detailed the arguments even further, a second round of answers available here: https://jonathanhaidt.substack.com/p/why-some-researchers-think-im-wrong. I can't recite all the details of the conversation, but some important aspects are these:
1) Nowadays a large share of data sets seam to show correlation between social media use and worse mental health among young girls. Most reports reports cited in your Substack are looking at digital media more generally rather than social media, or doesn't tell what's happening to younger girls specifically. (I do recognise that you cite some sources that have this destinction and doesn't find any effect, but Haidt and Twenge have listed many more that have found effects).
2) Secondly, some effects reported as "small" actually aren't that small if you compare with what you'd normally expect from social factors in large data sets.
3) Most importantly, the drastic surge in bad mental health among young girls happened at the same time in many countries. The countries in the Anglosphere showed a drastic increase in suicides and self harm at the same time starting around 2010. Twenge argues that it's hard to find any other thing that happened to all these countries at the same time than the many fast, well recorded tech driven changes in life style between 2010 and 2015 or so: https://substack.com/browse/recommendations/post/138052249
Do you think any of these things would alter your conclusions?
Elias, thank you for this comment. I'll try to deal with some of your points in a future letter where I'll respond to a number of questions and comments. Of course, as you mentioned, I did point out that many of the reviews I found looked specifically at social media effects and failed to find any of meaningful size. Your comment about effects being specific to girls of age 10-14 led me to look back at the recent, very large study by Silje Steinsbeckk and colleagues, whose subjects were 10, 12, 14, and 16 years of age at the start of the study and where the variable in question was use of social media. They looked at the data separately for each age group, separately for boys and girls, and found no effects for any of these groups. The great majority of research studies simply fail to support the view that social media is having a large harmful effect on girls. I have no doubt that some girls are negatively affected and some are positively affected, but overall there are no social media effects that can explain the large increases in suicide among both girls and boys between 2008 and 2019.
Hello again, and thank you for replying. A privilege to be able to interact with you.
I agree that the work by Steinsbekk is relevant and important here. But Haidt and his colleagues found quite a lot of studies pointing in another direction in their overview resulting in this long Substack on the subject: https://jonathanhaidt.substack.com/p/social-media-mental-illness-epidemic. When it comes to correlational studies, they found 66 studies out of which 55 showed significant correlation and 11 showed no correlation. They claim that .. "In fact, there is a revealing pattern found across many studies and literature reviews: Those that look at all screen-based activities (including television) for all kids (including boys) generally find only small correlations (usually less than r = .10), but as you zoom in on social media for girls the correlations rise, sometimes to r = .20, which is quite substantial, as I’ll show in a moment.".
There are also longitudinal studies, where the author's conclude that the time span between change in social media usage and some measure of mental wellness is the key: "7 studies used a week or less (5 of them were daily), and only 1 of the 7 found an effect. But 33 studies used a month or more (20 were annual) and of these, 24 found a significant effect. So a simple dose-response model in which social media is like poison (where cutting consumption on Monday makes you feel better on Tuesday) does not seem to be supported. But 73% of the studies that looked for causal effects a month or more in the future found them."
One important reason why some other meta studies have reported little or no effects seem to be that very few have really narrowed down on social media and depression and anxiety among young girls. This seems very important in at least two aspects: One is that there is over reporting of bad effects of digital media in general, which you have rightfully pointed out. The other is that the effect of social media on girl's depression and anxiety is mixed up with other things and doesn't get the focus it should.
A quite new report, which is of another kind but I really think needs consideration (also if threre are reasons to debunk it) is the one from Sapiens labs using a data set from a large number of countries. It's called "Age of First Smartphone/Tablet
Sorry for bombarding you with this, I've already taken a lot of your time but thought these things were relevant for the continuous discussion. Let me end with saying that I fully agree that we need to give our kids more opportunity for free play. My work with adolescents here in Sweden (and reading academic studies mentioned here) has nevertheless made me believe that watching beauty influencers on instagram doesn't have any of the free play qualities we're looking for. Games and other forms of digital media may have more of it.
Looking forward to your future posts here and elsewhere.
Regarding the Sapien Labs cross-sectional study: I recall that Haidt was excited by the sheer size of this cross-sectional study both in number of participants and participant’s country of residence. I second Elias’s desire to have Peter’s objective and highly trained eye on this report. In particular, I would love to have his feedback on the study’s methods and whether the dataset’s scale overrules smaller cross-sectional studies that lead to different conclusions.
Peter, thank you so much for reviewing the literature and being such an engaging expert!
Am I the only one who wants to sit in a room with Peter Gray and Jonathan Haidt and a big bucket of popcorn?
“Now we ban them from the outdoors, so the only regular way they can get together without adult interference is through social media.”
So, so important. The main adult interference should be interfering with them til they go outside and play!
I know we neurodiverse, introverted people spend time on social media because we feel more comfortable. The more depression and anxiety I have - or my son has - the more we don't want to go out. Being able to socialize in an online group of other neurodiverse people who "get it" is very important. On the other hand, if you're concerned about your child's social media use, there is no harm in having a conversation with them and sharing your concerns.
I agree with Peter's conclusions, though. School has caused real trauma for people in my family. Periods of high social media use did not seem to change things one way or another.
One thing I would have liked to see in context here would have been more information about what the correlation between reported rates of depression and anxiety are to the rates of actual suicide. I was given to understand that they tend to differ substantially, particularly between the genders, and it would have helped with parsing some of the arguments here to have that information available in the text. I realize it's not true of everyone, but as the mother of young boy, I've discovered that lately I care a lot more about actual suicide rates, than the number of teen girls self-reporting being anxious. In some ways I think the latter skews the way we interact with data along these lines, but I'm not an expert, just a worried mom.
Good point, Eleanor. I discussed the gender difference in Letter D2. I've focused on suicides throughout these D letters because measures of anxiety and depression have varied from study to study and time to time, so it is not possible to chart rates of them over the 70 year period I have been considering with a consistent measure. However, concerning this letter, the research shows little or no effect of social media use on any measure of anxiety or depression used in such studies. Suicides, of course, are too rare to use as a measure in such research. As I pointed out in Letter D5, suicides go up whenever school is in session and down when it is not in session, and this is more true for boys than for girls.
Thanks for adding the clarity! I've been reading all your letters and finding them very helpful, but in a lot of ways it's a very overwhelming topic 😅
Yes, it is an overwhelming topic!
I’ve wondered the same as a mom of 3 young boys. It seems there are a decent amount of impulsive (is that the best word?) suicides. They lack a history of depression and their decision is made around an incident rather than a something they’ve been planning for a while.
Do you think social media might be a catalyst? That is, it makes a good youth-culture better and a harmful youth-culture worse. The effects would cancel each other out, preventing if form showing up in the statistical research. Or do the studies control for that?
Imagine a mixed age play culture vs. a same age culture. Might social media have a completely different effect on those two culture?
Good point about social media being a catalyst. I will try to bring that into a future letter. I will discuss what seem to be the benefits and the harms of social media use for teens and suggest, as you do, that they may cancel one another out in the statistical data.
And as Jonathan Haidt point out; most of digital technology have a net positive effect. He mentions; Skype, Teams, Phone calls, texting, multiplayer videogames
Changing kids life to revolve around posting pictures of themselves, in order to get likes and go viral, is bad. So it's the hyper viral platforms spreading this bad youth culture we need to fight. The hyper viral platforms such as twitter, instagram, facebook, and tiktok. Fighting the unhealthy communities on these platform are one way to decrease anxiety. Increasing free play is another.
I look forward to a time where kids have the skills to enjoy free play on twitter, instagram, facebook. and tiktok. But right now, they're getting trapped in a game of social approval.
Here is Haidts on how social media / hyper viral platforms, harms the youth:
Click timestamp-link to Podcast. https://youtu.be/-4AAST_AdSg?si=GJJ3xTXA4VMOOKS9&t=451
Or google: Jonathan Haidt Debates Robby Soave on Social Media
And start at 8:00
Haidt also point out that social medie isn't like sugar, where you only affect yourself. But it also have a network effect, that is like passive smoking or driving too fast.
He points out, that experimental studies need to take this network effect into account. If you stop using facebook for a month, but all your friends continue to use it. Then you'll still live in a culture influenced by hyper viral platforms.
Jacob, the poins you (and Jon Haidt) make here are reasonable and I think are valid to a considerable degree, but difficult to test with research. I hope to take them into account in a future letter. We do now live in a networked world. It isn't going to work to try to reverse that or take it away from kids. To the degree that it is harmful we need to learn how to minimize that. To the degree that it is helpful, we need to learn how to maximize that. My main disagreement with Jon os that he would have us deprive teens, maybe especially young teen girls, of this tool, whereas I would have us talk with teens about about their uses of social media and discover how it may be helping and harming them, in ways that would make them (and us) aware of dangerous versus safe ways of using it.
This is what I wanted to ask as well. Could social media use by my classmates make my life worse even if I am not on it? Perhaps by increasing judgmental attitudes about appearances among my peers, or by me being left out of social groups or missing new fashions because I am not on social media, by increased bullying behavior, etc. “Culture” is hard to measure but some of these specific hypotheses could be.
Great article and on a critical topic. My worry is that the cat is already out of the bag. Meaning, studies that evaluate whether someone is happier reducing screen time may be missing the broader point that the test and control groups already were big users to begin with. So taking away from the test group may not show much of an improvement or may be influenced by the other factors you note. I have heard that looking at data pre and post 2012 (or whenever the real launching off point for the iphone was) is a better comparison and that since that time, Americans are far more lonely than they were before. Someone else referenced the catalyst effect...I agree with this. If you're happy anyway this makes you more happy and you see happy things. If you are not, this points out all the areas you are missing and how much fun others are having. Regardless, the cat is out of the bag, IMO, and I'm not sure how we ever go back.
The article is astounding! I'm a bit surprised for the lack of evidence with social networks. How can we link that to the fact that Facebook revealed serious connection and design orientation of Instagram with addiction, anxiety, etc?
Good article, Peter. I prefer your conclusion to the introduction funnily enough.
I just read (the below link) and even if it were clear there’s an issue with smart phones, the solution is the same.
Don’t restrict freedoms, teach safety, and promote healthier behaviours.
I’d be interested in your comments/thoughts?
https://open.substack.com/pub/jonathanhaidt/p/anglo-teen-suicide?r=r5oro&utm_medium=ios&utm_campaign=post
Peter, the focus of these studies seem to be connecting the use of social media and digital technology as the cause to increased anxiety/depression. Could it be that it is not social media or digital technology per se causing the mental health issues but instead what they are missing out from by using technology (i.e. face to face interactions) that is causing the breakdown in mental health? I am trying to keep up with all the news letters so that is a topic that may have been discussed already. I just feel I have witnessed over the last 25 years as a school based SLP, the decline in free play, along with the increase of academic demands that are not developmentally appropriate (both topics you have covered in depth!) along with a push to 1:1 technology in schools. Kids are now staring at screens all day, then going home to doing more homework on computers, playing video games or interacting on social media, all which drastically reduces the amount of face to face interaction that we all need as social beings! Could it be related to the way kids are getting a dopamine release from technology versus face to face interactions? As with many hot topics, I feel the answer is always more complex! Thank you for your great articles!
Yes, IF social media (or screen time generally) were having an effect on mental health that effect might well be mediated by a decline in other activities ("What they are missing" because of time on screens), but the point is that there is no good evidence to date that time on screens (or on social media) is having an effect on mental health. if there were an effect it would be interesting to speculate on how that effect might be mediated.
I am so grateful the research that you and others are doing on this topic. Has anyone looked to see if additions to the vaccine schedule starting in the year 2000 and moving forward could have had any impact in mental health? What about the introduction of the vaccine Gardisil in 2006? I tend to feel that the rise in mental health issues may be due to a combination of many factors, not just one.
Hi, and thank you for this Substack and all your fantastic work!
The edge of the academic discussion regarding digital seams now to be way beyond "screens are bad" - the leading hypothesis instead being that social media (not digital media in general) is affecting girls 10-14 (and not people in general). This is championed by several, but maybe most prominently by Haidt and Twenge detailing this argument backed up by an overview of a couple of hundred articles: https://jonathanhaidt.substack.com/p/social-media-mental-illness-epidemic. Haidt then asked people to criticise him which detailed the arguments even further, a second round of answers available here: https://jonathanhaidt.substack.com/p/why-some-researchers-think-im-wrong. I can't recite all the details of the conversation, but some important aspects are these:
1) Nowadays a large share of data sets seam to show correlation between social media use and worse mental health among young girls. Most reports reports cited in your Substack are looking at digital media more generally rather than social media, or doesn't tell what's happening to younger girls specifically. (I do recognise that you cite some sources that have this destinction and doesn't find any effect, but Haidt and Twenge have listed many more that have found effects).
2) Secondly, some effects reported as "small" actually aren't that small if you compare with what you'd normally expect from social factors in large data sets.
3) Most importantly, the drastic surge in bad mental health among young girls happened at the same time in many countries. The countries in the Anglosphere showed a drastic increase in suicides and self harm at the same time starting around 2010. Twenge argues that it's hard to find any other thing that happened to all these countries at the same time than the many fast, well recorded tech driven changes in life style between 2010 and 2015 or so: https://substack.com/browse/recommendations/post/138052249
Do you think any of these things would alter your conclusions?
Looking forward to your next post. All the best!
Elias, thank you for this comment. I'll try to deal with some of your points in a future letter where I'll respond to a number of questions and comments. Of course, as you mentioned, I did point out that many of the reviews I found looked specifically at social media effects and failed to find any of meaningful size. Your comment about effects being specific to girls of age 10-14 led me to look back at the recent, very large study by Silje Steinsbeckk and colleagues, whose subjects were 10, 12, 14, and 16 years of age at the start of the study and where the variable in question was use of social media. They looked at the data separately for each age group, separately for boys and girls, and found no effects for any of these groups. The great majority of research studies simply fail to support the view that social media is having a large harmful effect on girls. I have no doubt that some girls are negatively affected and some are positively affected, but overall there are no social media effects that can explain the large increases in suicide among both girls and boys between 2008 and 2019.
Hello again, and thank you for replying. A privilege to be able to interact with you.
I agree that the work by Steinsbekk is relevant and important here. But Haidt and his colleagues found quite a lot of studies pointing in another direction in their overview resulting in this long Substack on the subject: https://jonathanhaidt.substack.com/p/social-media-mental-illness-epidemic. When it comes to correlational studies, they found 66 studies out of which 55 showed significant correlation and 11 showed no correlation. They claim that .. "In fact, there is a revealing pattern found across many studies and literature reviews: Those that look at all screen-based activities (including television) for all kids (including boys) generally find only small correlations (usually less than r = .10), but as you zoom in on social media for girls the correlations rise, sometimes to r = .20, which is quite substantial, as I’ll show in a moment.".
There are also longitudinal studies, where the author's conclude that the time span between change in social media usage and some measure of mental wellness is the key: "7 studies used a week or less (5 of them were daily), and only 1 of the 7 found an effect. But 33 studies used a month or more (20 were annual) and of these, 24 found a significant effect. So a simple dose-response model in which social media is like poison (where cutting consumption on Monday makes you feel better on Tuesday) does not seem to be supported. But 73% of the studies that looked for causal effects a month or more in the future found them."
One important reason why some other meta studies have reported little or no effects seem to be that very few have really narrowed down on social media and depression and anxiety among young girls. This seems very important in at least two aspects: One is that there is over reporting of bad effects of digital media in general, which you have rightfully pointed out. The other is that the effect of social media on girl's depression and anxiety is mixed up with other things and doesn't get the focus it should.
A quite new report, which is of another kind but I really think needs consideration (also if threre are reasons to debunk it) is the one from Sapiens labs using a data set from a large number of countries. It's called "Age of First Smartphone/Tablet
and Mental Wellbeing Outcomes". https://sapienlabs.org/wp-content/uploads/2023/05/Sapien-Labs-Age-of-First-Smartphone-and-Mental-Wellbeing-Outcomes.pdf
Sorry for bombarding you with this, I've already taken a lot of your time but thought these things were relevant for the continuous discussion. Let me end with saying that I fully agree that we need to give our kids more opportunity for free play. My work with adolescents here in Sweden (and reading academic studies mentioned here) has nevertheless made me believe that watching beauty influencers on instagram doesn't have any of the free play qualities we're looking for. Games and other forms of digital media may have more of it.
Looking forward to your future posts here and elsewhere.
Regarding the Sapien Labs cross-sectional study: I recall that Haidt was excited by the sheer size of this cross-sectional study both in number of participants and participant’s country of residence. I second Elias’s desire to have Peter’s objective and highly trained eye on this report. In particular, I would love to have his feedback on the study’s methods and whether the dataset’s scale overrules smaller cross-sectional studies that lead to different conclusions.
Peter, thank you so much for reviewing the literature and being such an engaging expert!