Is Facebook Bad for Your Mental Health?

Nina VASAN: Mark Zuckerberg was my college classmate, and so I think I’m the 1100th person on Facebook.

That’s Dr. Nina Vasan.

VASAN: So, I’ve been on since 2004, and looking at this research, actually I’m realizing that my classmates and I, we’re in that user group.

Nina’s now a psychiatrist and professor at Stanford. Long ago, though, she was a very early Facebook user. Which made her an inadvertent participant in a recent study published by my other guest today.

Alexy MAKARIN: My name is Alexey Makarin. I’m an assistant professor in Applied Economics at M.I.T. Sloan School of Management.

Some of Alexey’s research focuses on the economics of media. Specifically, social media. In 2020, he published a study that looked at whether the Russian social media platform “VK” fomented anti-government protests around the country in 2011.

MAKARIN: Participation in protests was associated strongly with social media use, but the causal impact was still unclear, and that’s where we came in.

In that study, Alexey and his co-authors relied on the introduction of VK at its founder’s university as a randomizing device.

MAKARIN: What we noticed is that VK was created in 2006 by a student of St. Petersburg State University, and the first users of VK were students of that university at that particular time.

St. Petersburg State University attracts students from all over Russia — more from some cities than others. And that suggested a natural experiment to Alexey and his colleagues.

MAKARIN: So, I’m from a very small town in Russia. If I go from that particular hometown to study at that particular university, in my hometown, the penetration that social media platform is going to be higher. Why? Because potentially, I can send a link to my friends and family who remained in my hometown, and they will be the second layer of adoption across Russia.

If you want to figure out the impact that social media has on political activity, you can’t just compare protest rates in areas with high versus low social media, because that isn’t random. But you could compare protest rates in cities that just happened to be exposed to more social media, by chance, because people there were connected to students who knew the founder of VK. In that case, uptake could actually be random.

Alexey and his colleagues were right: the more people in a given city used VK, the more likely it was there would be protests in the city, and the larger those protests were. Social media was fueling political behavior. That study got Alexey thinking: how else could social media change human behavior?

From the Freakonomics Radio Network, this is Freakonomics, M.D. I’m Bapu Jena. Today on the show: since its inception, people have speculated that social media could have psychological effects on its users. But does it cause mental health problems?

MAKARIN: Immediately after Facebook comes to a particular college you observe a bump in the mental health problems among students at that college.

And, if Facebook and other social media platforms are causing mental health problems —can we do anything about it?

VASAN: In medicine we take this oath to do no harm. And that was really the idea that we had around how do we change the user experience.

*      *      *

VASAN: When I was in medical school, I myself experienced depression and I think that that personal experience is the very first thing that drew me to mental health. 

That’s the psychiatrist Dr. Nina Vasan again, from Stanford. She runs the Brainstorm Lab for Mental Health Innovation, and she’s also chief medical officer of a mental health startup called Real. She says the Covid-19 pandemic has been a challenge — but has also made people pay more attention to mental health.

VASAN: Where mental health is today is actually what I thought it would look like maybe a decade from now. A lot of my colleagues in the field talk about this silver lining that we’ve experienced because of how so many people struggled with their mental health during the pandemic, people who never had before, it really brought to the forefront the importance of addressing mental health as a society.

During the first year of the pandemic, rates of anxiety and depression around the world grew by 25 percent, according to the World Health Organization. One major reason for this massive increase was stress from social isolation.

The pandemic shone a spotlight on mental health, but it also made it harder for people to get help. Data from the Centers for Medicare and Medicaid Services indicated that, on average, adults and children with public insurance experienced around a 25 percent decline in mental health care utilization between March and October of 2020, compared to that same period in 2019. And even as medical care normalized during the pandemic, use of mental health services rebounded more slowly than other specialties.

VASAN: When people first start experiencing the symptoms of mental illness to when they actually start officially seeking help within the healthcare system, that amount of time is 11 years. That’s a huge delay. And what it shows is that people are really only seeking help when they’re in crisis.

Nina has thought a lot about how to intervene long before someone is in crisis. Telehealth services are part of the answer. But it’s also about meeting people where they are. Which, these days, tends to be on social media.

VASAN: We have this idealistic sense of the old school doctor who would go knocking on doors with their bag, treating people in the community. And then we think, “Where are people living their lives?” Literally, billions of people are living their lives online. Right now, worldwide, the average number of hours that we are spending on social media is 2.5 hours a day.

Bapu JENA: What do we know about how these platforms can affect psychological health?

VASAN: Time spent on social media has been shown to really have, I would say, dire impact on mental health. Correlations to increases in depression, anxiety, substance use, suicidal thoughts, self-harm thoughts, and even body image issues, and body image pathology. What we’re also seeing is that the more time people spend on social media, the more likely these different things are to increase. We don’t yet have the exact relationship.

MAKARIN: I started my economics journey back in the day when I was in Russia. I started working on this project on the impact of social media, on anti-authoritarian protests. So ever since I started working on that project, I was very curious about social media, how can we study the causal impact of social media on various aspects of our lives. And it’s very difficult, and that’s why I only wrote two papers and not more.

That, again, is the economist Alexey Makarin from M.I.T. We already talked about his first paper on social media and protests in Russia in 2011. It was a fascinating study with a clever design, and a clear predecessor to his second bit of work.

Facebook, or THE Facebook, as it was initially called, was launched in 2004 for Harvard students by Harvard students, among them Mark Zuckerberg. About a month later, they opened the platform up to students at Yale, Columbia, and Stanford. As we all know, it didn’t stop there. Now, nearly two decades later, Facebook has around 3 billion active monthly users.

There are other platforms of course, but Facebook changed the game. Some of its effects have been very visible: Facebook is widely credited as being one force behind the Arab Spring uprisings in Tunisia and Egypt in 2011. A few years later, in 2017, around 70 percent of people who attended the Women’s March in Washington, D.C. told researchers they heard about it on Facebook.

What it’s been doing to our health, however, has been a little less clear.

MAKARIN: There are a lot of papers that document a significant association between mental health problems and social media use. Most of the results show that people who use Facebook a lot or other social media platforms, they seem to be experiencing worse mental health. But of course, we know that this is correlation and there are some issues with that. Maybe when I’m depressed, I use social media more. Or it could be that, for instance, if I break up with my girlfriend, then maybe the next day I’m going to be depressed and at the same time, I’m going to have more free time to use, to spend on social media.

These correlation issues make it hard to prove that Facebook, or any social media platform, is causing poor mental health. What we do know is that nearly one in five adults in the U.S. — and half of all adolescents — live with mental illness. These conditions range in severity, but most often include major depressive episodes and anxiety disorders. A lot of factors can contribute to someone developing a mental health problem—things like family history, living with chronic illness, or coping with a traumatic experience. Often it can be hard to prevent or mitigate these conditions. But what if, for some people, it was as simple as deactivating Facebook? A few years ago, some researchers at Stanford and N.Y.U. conducted an experiment where they paid a group of people, randomly, to do just that.

MAKARIN: And what they showed is that if you pay people to deactivate their Facebook accounts, one week or four weeks after, those people report that they feel better about themselves, their general wellbeing is better. So, they report less anxiety, they report less depression.

Alexey and his co-authors wanted to take a different approach, but they weren’t sure how to tackle this thorny question surrounding social media and mental health. They debated a few different strategies until one of them suggested an approach that was familiar to Alexey.

MAKARIN: So, we followed a very similar logic to the paper that I described before. We looked at the staggered adoption of Facebook across college campuses from 2004 through 2006. When Facebook was just created, Facebook did not immediately become available to all students across the U.S. They slowly started being available at Harvard, Stanford and so forth, and then slowly they became available to other colleges. And that process took about one and a half years, up to two years. So we compare students at colleges that just received Facebook, versus students that were at colleges at the same time but without Facebook.

JENA: How are you measuring mental health in these universities?

MAKARIN: We got access to this survey. It’s called the National College Health Assessment. And it is one of the most comprehensive surveys among U.S. college students related to their health and behavior, and they have a lot of variables related to mental health. You have questions about depression, you have questions about the take up of anti-depression therapy, you have questions about anxiety, bulimia, eating disorders. That survey is collected every semester across many, many colleges. Moreover, it exists from 2000 through 2008. So that allows us to get a sense of student mental health at different colleges before and after Facebook was created. Immediately after Facebook comes to a particular college, you observe a bump in the mental health problems among students at that college.

JENA: So, you look at universities that have similar trends of mental health issues prior to the introduction of Facebook and show that in those sets of colleges, in which Facebook is introduced, there’s an increase in mental health issues compared to universities that looked very similar before in terms of mental health issues, they don’t experience that increase.

MAKARIN: Yes. We observe that after the introduction of Facebook, the student mental health at colleges with Facebook worsens relative to students at colleges that did not get Facebook.

JENA: When you say that mental health worsened, what specifically did you see?

MAKARIN: The most affected conditions are depression and anxiety. More students report that they were depressed, severely. Actually, the question was stated the following way: “Have you felt so severely that it was difficult for you to function and how many times you felt so in the past year?” For that question, we observe the biggest effects, but we also observe some other effects for other conditions.

After the break: What else was in the data?

MAKARIN: I didn’t actually realize how bad the situation is in the United States in terms of mental health

We’ll also look at how social media platforms are trying not only to mitigate harm, but to actually improve users’ mental health.

VASAN: We created a whole platform of, little mini tools, little activities that people can do on the platform in real time.

I’m Bapu Jena and this is Freakonomics, M.D.

 *      *      *

Alexey Makarin’s new study is called simply “Social Media and Mental Health,” and it was published in November of this year in the American Economic Review. Before the break, he told us that their findings suggest a “bump” in mental health problems among students at a college after Facebook is introduced there. What that “bump” translates to is a worsening of what Alexey and his co-authors called their “index of poor mental health” by about 8 percent. Facebook access also led to a nine percent increase in depression among students on college campuses, and a 12 percent increase in generalized anxiety disorders. Alexey says the mental health effect of using Facebook is about one fifth the size of how it feels when you lose your job. So, how does this new work, which suggests Facebook causes mental health problems, compare to prior work, which only showed an association between the two?

MAKARIN: It’s actually strikingly similar.

JENA: Why is it that social media has this causal effect on mental health issues? What’s the mechanism?

MAKARIN: We believe that Facebook users, engage in unfavorable social comparisons, and that’s what’s driving the negative impact of Facebook on mental health. And the fact that we study Facebook at its inception, it automatically rules out a lot of stories that you might have about, for instance, the like button, the newsfeed, the political news on social media. All of that is non-existent, right? So the fact that we observe that Facebook had a negative impact all the way in 2004, 2005, 2006, that suggests that something in the nature of social media leads to worse mental health outcomes. And in our case, we believe that it’s due to unfavorable social comparison.

JENA: I mean, that’s an interesting point to me because you’re looking at a period when Facebook was really in its early stages. So some of the features are different now. It was really all about peer networks and social comparisons back then. If you find that on average there’s an X percent increase in mental health problems in a university that adopted Facebook, do you then scale that up by the fact that it could be only 2 percent of people that use Facebook or not? Because that would tell us something about what is, the effect of a single individual using Facebook who otherwise wouldn’t have used it.

MAKARIN: Yeah, that’s a great question. We don’t scale up. We believe that actually the use of Facebook on college campus at the time was almost universal, and the adoption was very fast. The common estimate that we found is that about 85 percent of college students had a Facebook account and were using it on a daily basis at the time.

JENA: Wow. So, it’s quite high. When you set out to do the study, were you expecting to find an adverse impact of Facebook’s introduction on mental health?

MAKARIN: So, what definitely surprised me is how clean in the end, this natural experiment was. I did not expect it to be as clean because when you work with these observational studies, it’s not like you are operating in a controlled environment. When I started writing the background section for our paper, I didn’t actually realize how bad the situation is in the United States in terms of, mental health trends among young adults. The pictures that you see in the data are stark. So, for instance, the percentage of, individuals from 18 to 23, who reported experiencing a major depressive episode in the past year increased by 83 percent. So almost doubled between 2008 and 2018. And, the fact that this trend is sort of coinciding with the spread of social media is really begging the question whether social media is to blame.

JENA: It strikes me that what you’re doing, it does fit into a broader important economic question. It’s a societal question of what is it that drives people’s happiness? Or in economics we talk about utility. You know, in very standard economic models of utility, we don’t typically factor in the idea that our wellbeing may be a function of what we perceive the wellbeing or accomplishments of others to be. And there is economic literature out there, which would suggest that in some cases, our own utility is lower when the utility of others is higher. One of the things that’s coming out of social comparison is that you’re comparing accomplishments relative to that of what you observe from your peers. And if you benchmark yourself in that way, it’s maybe not a surprise that something like Facebook could actually lead to declines in wellbeing, but what’s even more striking is that we’re talking about a situation where these individuals might actually experience a clinical manifestation that we call depression. So that’s much more severe than just saying, “Okay, I’m unhappy because Alexey just, you know, put on Twitter that he published a paper in the American Economic Review.”

MAKARIN: One of the ways in which we document that indeed social comparison seems to be a channel is that after the introduction of Facebook, students suddenly believe that the other students consume much more alcohol than they did before. But the actual usage of alcohol did not increase at all. So, students have some naivete about how to interpret the Facebook posts of others. And they believe that everybody’s having a fantastic time while in fact, nothing changed. This fear of missing out seems to be a big component. We also find that in the short to medium run, the negative effects of Facebook on mental health increase with the length of exposure to the platform. And actually, students reported that their academic performance worsened as a result of their mental health issues.

JENA: So, it’s not just a time allocation issue, it’s that the time you spend on Facebook means you’re not spending that time studying, but it also creates mental health problems that make it harder to perform academically as well as you were prior to the introduction.

MAKARIN: We got very lucky because in the survey that we have, there is a specific question that asks whether your academic performance worsened as a result of mental health, such as anxiety, depression, and so forth. When we analyzed that question, we found that indeed Facebook had this downstream impact on academic performance.

JENA: If you were to say, you know, “Here’s why this study is important. Here’s what you should make of it, and here’s what you should do with this information.” How would you articulate that?

MAKARIN: This study is important because we think that it’s the most comprehensive causal evidence on the impact of Facebook on mental health. It’s also important because we are able to zero in on a subset of people who are particularly vulnerable and who experienced the most significant deterioration of mental health in the past 15 years, who are college students and young adults. And what can people do with this information is that, first of all, they can be aware of it, so they can potentially alter their own usage of social media and the usage of social media by their kids or friends. And second, we hope that social media platforms figure out a way to find a balance between maximizing their ad revenue and potentially mitigating the harms that social media does

VASAN: We talk about why social media has made mental health worse, and that’s something that we’re actively working with social media companies, around how to address that.

That’s the psychiatrist Dr. Nina Vasan again. In 2018, her lab at Stanford, called Brainstorm, started working with the social media platform Pinterest.

VASAN: They were looking to understand different patterns and what was driving people to their platform. And what they actually saw was that the fourth most common category of search terms was related to mental health. And it was things like, depression, stress, anxiety. When people think of Pinterest, they think about things like wedding planning, or interior design, or recipes. Pinterest has not been a place that historically people really talked about, “I want to go address my mental health on this platform.” What Pinterest realized was, “This is something that people are coming to us for. How can we create a space for people to be able to feel, safe and understood?”

Around that same time, a 14-year-old girl in England named Molly Russell died by suicide. Afterward, it was discovered she had viewed more than 2,000 posts related to suicide and self-harm on social media platforms, one of which was Pinterest.

VASAN: This was alarming and led to a lot of questions around what are kids engaging in? How do we keep them safe? What are the interventions that can then be done? And also, what is the role of social media companies to keep users safe when things like this happen? There are three big things that we did. The first was what we call micro therapeutics. We created a whole platform of little mini tools, little activities that people can do anywhere from one to five minutes, that take the same types of therapies that we might do with our patients in clinic, but instead that they can do on the platform in real time. Not only things like mindfulness and gratitude, but even leveraging cognitive behavioral therapy and helping people understand their thoughts and thought patterns, where they can start thinking in a more positive way. Now when people go on the Pinterest platform and they type in “depression” or “stress” or “anxiety,” they get an entire pop-up box that’s separate from their Pinterest page and it’s very private. 

The second thing that we did is called Compassionate Search. If you go into Pinterest and you start to type in something related to suicide, you’re not now going to get pushed additional content into your inbox like you did before. Some things will autofill, but things that we have been able to flag as potentially dangerous will not autofill, so we’re not further making things potentially harmful for people. And then the third thing was working with the algorithm engineers specifically around how to decrease the amount of self-harm content on the platform. And by educating the engineers around, what is self-harm, what is self-harm content, after six months or so, there was a significant decrease in self-harm content on the platform. In medicine, we take this oath to do no harm. And that was really the idea that we had around, how do we change the user experience? Everything from the way people searched, the way that results get popped up to them, so that the platforms are doing no harm to their users.

Not long after Pinterest implemented these changes, they noted an 88 percent drop in reports of self-harm content by users. The company said it was able to remove this kind of content three times faster than before. But is it actually improving users’ mental health? We don’t know.

Other platforms have also tried to be proactive about mental health. Meta, the company that owns both Instagram and Facebook, allows users to flag worrisome posts they’ve seen from friends. In 2017, Facebook developed its own algorithms to identify users at risk of self-harm. And, they have an “emotional health” page containing resources on conditions like anxiety and depression. And yet, just last year the Wall Street Journal published a series in which they claim that Facebook has long studied and known about the harm it causes, particularly to mental health — and that it has continued to publicly deny and downplay this information.

In 2021, more than 4 billion people around the world used some form of social media. That’s half of the entire population of the planet. We’ve had clues about its effects on our mental health, but now we’re starting to get some definitive answers. What we do with that information is a looming question. Nina Vasan thinks a good place to start is with data.

VASAN: When we look at an area like mental health, we have enormous data that we use in order to do everything from assessing someone’s symptoms, how depressed are they, how anxious are they? But we don’t really share this information with patients. And I always try to give an analogy to something like diabetes, where someone with diabetes, they are very aware most often of their blood sugar, they’re measuring that regularly. They’re making food choices based on what their blood sugar levels are. In mental health, we have similar metrics around things like depression, anxiety scores, sleep scores, but patients aren’t actively using these numbers and metrics. What has not historically been a part of the conversation with patients is for them to understand what these numbers are and to engage in these numbers themselves. 

That’s it for today’s show. If you or someone you know is struggling with a mental health problem, it’s really important to seek help. Talk to your primary care doctor about finding appropriate services, or you can call the National Suicide and Crisis Lifeline. The number to dial is 988. Trained workers are available to talk 24 hours a day, 7 days a week, and all calls are toll-free and confidential.

I want to thank my guests, Alexey Makarin and Dr. Nina Vasan, and thanks to you, of course, for listening. Let us know what you thought about today’s episode. How does social media make you feel? Good, bad, or a little of both? Have you ever tried to cut back on it to improve how you feel? Send us an e-mail or a voice memo, we’re at bapu@freakonomics.com.

Here’s an idea to leave you with about a different way that social platforms can affect our health. Sexually transmitted diseases, or STDs, are a huge public health problem, particularly among adolescents and young adults. Over the last decade, a number of new apps have made it easier for people to … meet other people. Could the staggered uptake of these apps across cities tell us something about their effect on STD rates?

Think about it, and in the meantime, coming up next week:

Next week on the show: They say what’s past is prologue. So, what can the past tell us about the present — and maybe even the future — of our health care system?

THOMASSON: It really shows us how a significant shock can play out even decades later

And what could the lasting effects be of another, more recent shock?

MARTIN: We’ve seen a lot more people applying and matriculating into healthcare programs, more people applying to medical school, nursing school.

That’s coming up next week on Freakonomics, M.D. Thanks again for listening.

 *      *      *

Freakonomics, M.D. is part of the Freakonomics Radio Network, which also includes Freakonomics Radio, No Stupid Questions, and People I (Mostly) Admire. All our shows are produced by Stitcher and Renbud Radio. You can find us on Twitter at @drbapupod. This episode was produced by Julie Kanfer and mixed by Eleanor Osborne, with help from Jasmin Klinger. Lyric Bowditch is our production associate. Our executive team is Neal Carruth, Gabriel Roth, and Stephen Dubner. Original music composed by Luis Guerra. If you like this show, or any other show in the Freakonomics Radio Network, please recommend it to your family and friends. That’s the best way to support the podcasts you love. As always, thanks for listening.

 MAKARIN: Because I’m not a native speaker, I’m always kind of very self-conscious that I’m phrasing things kind of not in fantastic manner sometimes.

 JENA: Alexey, if they can make me sound good, they can make anybody sound good, so don’t worry.