Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:00
Introducing
0:00
the first TV brand streaming
0:02
Xbox games, Samsung Neo QLED
0:04
and OLED with gaming hub. With Samsung
0:06
Neo QLED and OLED TV, you
0:08
can stream your favorite game straight to your TV.
0:10
No console required. All you need is compatible
0:13
controller and Internet. Play hundreds of games
0:15
from the biggest streaming services including NVIDIA,
0:17
GeForce NOW, Luna, Nutomix, and Xbox
0:19
app. There's always something to play in something
0:21
fresh to discover with thousands of new and classic
0:23
game titles. Gain access to games in seconds
0:26
without worrying about purchasing new hardware, downloads,
0:28
storage limits, all you need is an Internet connection
0:30
and a compatible controller. There is a special promotion
0:33
going through the end of the year. Purchase a twenty twenty
0:35
two Samsung Neo QLED or OLED,
0:37
and you get your first months of Xbox Game
0:39
Pass Ultimate and an Xbox controller on
0:41
Samsung. Disclosure through twelve thirty one
0:43
while supplies last purchased a twenty twenty two new QLEDQNE5B
0:47
or above. OLED S95 via participating retailers.
0:49
For terms and conditions, including redemption instructions,
0:51
see controller redemption dot com or see retailer. Control
0:53
color may vary. After the game pass trial, you will automatically
0:55
be charged unless limit one. Terms, xbox
0:58
dot com slash description terms.
1:04
Mark Zuckerberg was my college classmate, And
1:07
so I think I'm the one thousand one
1:09
hundredth person on Facebook. That's
1:11
doctor Nina Vassen.
1:12
So I've been on since two thousand and four and
1:15
looking at this research Actually, I'm realizing that
1:17
my classmates and I were in that user group.
1:19
Nina is now a psychiatrist and professor
1:22
at Stanford. Long ago though,
1:24
she was a very early Facebook user,
1:26
which made her an inadvertent participant
1:29
in a recent study published by my
1:31
other guest today. My name
1:32
is Alexey McCarran. I'm an assistant professor
1:35
in Applied Economics at MIT Sloan
1:37
School of Management. Some
1:38
of Alexey's research focused is
1:40
on the economics of media, specifically
1:43
social media. In twenty twenty,
1:46
he published study that looked at whether
1:48
the Russian social media platform, VK,
1:51
fomented anti government protests
1:54
around the country in twenty eleven.
1:57
Participation
1:57
in persons was associated strongly
1:59
with your
1:59
social media use, but the cause of impact
2:02
was unclear, and that's where we
2:04
came in. In that, study, Alexey
2:06
and his coauthors, relied on the
2:08
introduction of VK at its
2:10
founders university as a randomizing
2:13
device. What we noticed
2:15
is that VK was created in two thousand
2:18
six by a student of
2:20
Saint Petersburg State University, and
2:22
the first users of VK War,
2:24
students of that university at that particular time.
2:27
Saint Petersburg State University attracts
2:29
students from all over Russia. more
2:31
from some cities than others. And
2:34
that suggested a natural experiment to
2:36
Alexei and his colleagues. So
2:39
I'm from a very small town in Russia. If
2:41
I go from that particular hometown
2:44
to study at that particular university in my hometown,
2:46
the penetration that social media platform is gonna
2:48
be higher. Why? potentially, I
2:51
can send the link to my friends
2:53
and family who remained in
2:55
my hometown and they will be the second
2:57
layer of adoption across Russia.
3:03
If
3:03
you want figure out the impact that social
3:05
media has on political activity, you
3:08
can't just compare protest rates
3:10
in areas with high versus low
3:12
social media because that
3:14
isn't random. But
3:16
you could compare protests, rates, and
3:18
cities that just happened to
3:20
be exposed to more social media
3:23
by chance. Because people
3:25
there were connected to students who
3:27
knew the founder of VK. In
3:30
that case, uptake could
3:32
actually be random. Alexey
3:34
and his colleagues were right. The
3:36
more people in a given city used
3:38
VK, the more likely it
3:40
was there would be protests in the city,
3:43
and the larger those protests
3:45
were. Social media
3:47
was fueling political behavior.
3:50
That study got alexey thinking. else
3:53
could social media change human
3:55
behavior.
3:58
From
3:58
the Freakonomics Radio Network,
3:59
this is FreakonomicsMD. I'm
4:02
Bob Bujena. Today on the show,
4:05
since its inception, people
4:07
have speculated that social media
4:09
could have psychological effects
4:11
on its users. But
4:13
does it cause mental health problems?
4:15
Immediately after Facebook comes
4:18
to a particular college, a terrific
4:20
bump, in the medical health problems
4:22
among students at that college. And
4:24
if Facebook and other social media
4:26
platforms are causing mental health
4:28
problems, Can we do anything
4:30
about it? In medicine, we take
4:32
this oath to do no harm, and that
4:34
was really the idea that we had around
4:36
how do we change the user experience?
4:56
When
4:56
I was in medical school, I myself experienced
4:58
depression and I think that that
5:01
personal experience is the very
5:03
first thing that drew me to mental health.
5:05
That's the psychiatrist doctor Nina
5:07
Dawson again from Stanford. She
5:09
runs the brainstorm lab for mental
5:12
health innovation, and she's also
5:14
chief medical officer of a mental
5:16
health startup called Real.
5:18
She says the COVID nineteen pandemic
5:21
has been a challenge, but
5:23
has also made people pay more
5:25
attention to mental health.
5:27
Where mental health is today is
5:29
actually what I thought it would look like maybe
5:31
a decade from now. A lot of my
5:33
colleagues in the field talk about this silver
5:35
lining that we've experienced because
5:37
of how so many people
5:40
struggled with their mental health during the pandemic,
5:42
people who never had before. it really
5:44
brought to the forefront the importance of
5:47
addressing mental health as a society.
5:49
During the first year of the pandemic, Rates
5:51
of anxiety and depression around the
5:53
world grew by twenty
5:55
five percent according to the World
5:57
Health Organization. One
5:59
major reason for this massive increase
6:01
was stress from social isolation. The
6:05
pandemic shown a spotlight on
6:07
mental health, but it also
6:09
made it harder for people to get
6:11
help. Data from the centers for
6:13
Medicare and Medicaid services indicated
6:15
that on average, Adults and
6:17
children with public insurance experienced
6:20
around a twenty five percent decline
6:22
in mental health care utilization
6:25
between March and October of
6:27
twenty twenty compared to that
6:29
same period in twenty nineteen. And
6:32
even as medical care normalized during
6:34
the pandemic, use of mental
6:36
health services rebounded more
6:39
slowly than other specialties.
6:45
When
6:45
people first start experiencing the symptoms of
6:47
mental illness to when they actually
6:50
start officially seeking help
6:52
within that healthcare system, that
6:54
amount of time is
6:56
eleven years. That's
6:58
a huge delay. And what it shows is that
7:00
people are really only seeking help when they're in
7:02
crisis. Nina
7:03
has thought a lot about how to intervene
7:06
long before someone is
7:08
in crisis. Telehealth services
7:10
are part of the answer, but
7:13
it's also about meeting people
7:15
where they are. which
7:17
these days tends to be on
7:19
social media. We
7:20
have this idealistic sense of like the
7:22
old school doctor who would go knocking on
7:24
doors with their bag and treating people
7:26
in the community, and then we think where are
7:28
people living their lives? Literally
7:31
billions of people are living their
7:33
lives online. Right now,
7:35
worldwide, the average
7:37
number of hours that we are spending
7:40
on social media is
7:42
two point five hours a day.
7:44
What do
7:45
we know about how these platforms can
7:47
affect psychological health?
7:50
time
7:50
spent on social media has
7:52
been shown to really have, I would
7:54
say, dire impact on mental
7:56
health correlations to
7:58
increases in depression, anxiety,
8:02
substance use, suicidal
8:04
thoughts, self harm thoughts,
8:07
and even body image issues and
8:09
body image pathology. What we're
8:11
also seeing is that the
8:13
more time people spend on social
8:15
media the more likely these different
8:17
things are to increase. We don't yet have
8:19
the exact relationship.
8:23
I started my economics
8:25
journey back in the day when I was in
8:27
Russia. I started working on this
8:29
project on the impact of social media on
8:31
anti authority in projects. So
8:34
ever since I started working on, that
8:36
project I was very curious about
8:38
social media, how can
8:40
we study the cause of impact of social media
8:42
and various aspects of our lives? And
8:45
it's very difficult and that's why only
8:47
wrote two papers and not more. That,
8:51
again, is the economist Alexei
8:53
Materin. from MIT. We
8:55
already
8:55
talked about his first paper on social
8:58
media and protests in Russia in
9:00
twenty eleven. It was
9:02
a fascinating study with a clever
9:04
design and a clear
9:06
predecessor to his second bit of work.
9:09
Facebook or the Facebook as it was
9:11
initially called was launched
9:12
in two thousand four for Harvard
9:15
students, by Harvard students. Among
9:17
them, Mark Zuckerberg. About
9:19
a month later, they opened the
9:22
platform up to students at Gail,
9:24
Columbia, and Stanford. As
9:26
we all know, it didn't stop
9:28
there. Now nearly two decades
9:30
later, Facebook has around
9:32
three billion active monthly
9:34
users.
9:36
There are
9:39
other platforms of course, but
9:41
Facebook changed the game.
9:43
Some of its effects have been very
9:45
visible. Facebook is
9:47
widely credited as being one force
9:49
behind the Arab spring uprisings
9:51
Indonesia and Egypt in twenty
9:54
eleven. A few years later in
9:56
twenty seventeen, around
9:58
seventy percent of people who attended the
9:59
women's march in Washington
10:02
DC told researchers
10:04
they heard about it on Facebook.
10:06
What it's
10:06
been doing to our health, however, has
10:09
been a little less clear.
10:10
There are a lot of papers
10:13
that document a significant
10:15
association between mental
10:17
health problems and social media
10:19
use. most of the results
10:21
show that people who use
10:23
Facebook a lot or other social media
10:25
platforms, they seem to be experienced
10:27
in worse mental health. But of course, we know
10:29
that this is correlation and there are some
10:31
issues with that. Maybe when I'm
10:33
depressed, I use social media
10:35
more, or it could be that, for
10:37
instance, if I break up with my
10:39
girlfriend, then maybe the next day I'm gonna be
10:41
depressed and at the same time I'm gonna have
10:43
more free time to use to spend on
10:45
social media
10:49
These
10:49
correlation issues make it hard
10:51
to prove that Facebook or any
10:53
social media platform is
10:55
causing poor mental health.
10:57
What we do know is that nearly
10:59
one in five adults in the
11:01
US and half of all adolescents
11:04
live with mental illness. These
11:07
conditions range in severity, but
11:09
most often include major
11:11
depressive episodes and anxiety
11:13
disorders. A lot of factors can
11:15
contribute to someone developing a
11:17
mental health problem, things like
11:19
family history, living with chronic
11:21
illness, or coping
11:23
with traumatic experience. Often,
11:26
it can be hard to prevent or
11:28
mitigate these conditions.
11:30
but But what
11:31
if for some people, it was as
11:33
simple as deactivating Facebook.
11:36
A few years ago, Some researchers at
11:38
Stanford and NYU conducted
11:41
an experiment where they paid a
11:43
group of people randomly to
11:45
do just that.
11:48
And what
11:51
they show is that if you pay
11:53
people to deactivate their Facebook accounts one
11:55
week or four weeks after those
11:58
people, they
11:59
report
11:59
that they feel better. about
12:02
themselves, their general will be in is
12:04
better so they report less
12:06
anxiety, they report less depression. Alexey
12:08
and his coauthors wanted to take
12:10
a different approach, but
12:12
they weren't sure how to tackle this
12:14
thorny question surrounding social
12:16
media and mental health.
12:18
They debated a few different strategies
12:21
until one of them suggested an
12:23
approach that was familiar to
12:25
Alexey. So
12:26
we followed a very similar logic
12:28
to the paper that I just described
12:30
before. We looked at the staggered
12:32
adoption of Facebook across college
12:34
campuses. from two thousand four
12:36
to two thousand six. So when
12:38
Facebook was just created, Facebook did not
12:40
immediately become available
12:42
to all students across
12:44
the US. they slowly
12:46
started being available at Harvard,
12:48
Stanford, and so forth, and then slowly
12:50
they became available to other colleges.
12:52
And that process took about one and a half
12:54
years, up to two years. So
12:57
we compare students at colleges
12:59
that just receive Facebook versus
13:01
students that were at
13:03
colleges at the same time, but
13:05
without Facebook. How
13:06
are you measuring mental health
13:08
in these universities?
13:10
We got access to this
13:12
survey, it's called the National College Health
13:14
Assessment, and it is one of
13:16
the most comprehensive surveys and
13:18
one US car issuance related to their
13:20
health and behavior, and they
13:22
have a lot of variables related to
13:24
mental health. You have questions
13:26
about depression. You have questions about the take up
13:28
of anti depression therapy. You have
13:30
questions about anxiety, you believe me,
13:32
eating disorders. That survey is collected
13:34
every semester across
13:36
many, many colleges. And moreover, it
13:38
exists from two thousand to two
13:40
thousand eight, so that allows us to
13:42
get a sense of student mental health
13:44
at different colleges before and after
13:47
Facebook was created. immediately
13:49
after Facebook comes to a
13:51
particular college, a geographic bump
13:53
in the medical health problems amongst
13:55
students at that college. So you look at
13:57
universities that have similar trends of
13:59
mental health issues
14:00
prior to the introduction of Facebook
14:02
and show that in those sets of
14:04
colleges in which Facebook is introduced.
14:07
There's an increase in mental health issues
14:09
compared to universities that looked very
14:11
similar before in
14:13
terms of mental health issues. they don't
14:15
experience that increase.
14:17
Yes,
14:17
we observed that after the
14:20
introduction of Facebook, the
14:22
student mental health ecorages with Facebook
14:24
persons relative to students
14:26
at colleges that did not get Facebook.
14:28
When you
14:29
say that mental health worsened
14:31
What specifically did you
14:33
see? The most
14:34
effective conditions are depression and
14:36
anxiety. Most students report that
14:38
they were depressed severely actually,
14:40
the question was stated the following way.
14:43
Have you felt so severely that this
14:45
was difficult for you to function? And how many
14:47
times you felt so in the past year?
14:49
that question, we observed the biggest
14:51
effects, but we also observed some other effects
14:53
for other conditions.
14:57
After the break, What
14:59
else was in the data? I
15:00
didn't actually realize how bad the
15:03
situation is in the United States in terms
15:05
of mental health. We'll also
15:07
look at how social media platforms
15:09
are trying not only to
15:11
mitigate harm, but to
15:13
actually improve users' mental
15:15
health. We created a whole platform
15:18
of little mini tools,
15:20
little activities that people can
15:22
do on the platform in
15:24
real time. I'm Bob Pujena, and this
15:26
is FreakonomicsMD.
15:40
EconomicsMD is
15:42
sponsored by JustWorks. If
15:44
you're a small business leader
15:46
looking for an easier way to onboard
15:48
and manage remote employees, then
15:51
you need to learn about just works.
15:54
Just works makes it easier for you
15:56
to start run and grow a
15:58
business. Let me tell you how.
15:59
Just Works gives you the peace of
16:02
mind and knowing that payroll will
16:04
run smoothly. Your team has
16:06
access to high quality benefits, and
16:08
your business is getting the support it
16:10
needs. Just works makes
16:12
it simple to hire and manage
16:15
remote employees across all
16:17
fifty states. Just works helps
16:19
you comply with state payroll tax
16:21
requirements, keep up with state
16:23
labor laws and access health
16:25
insurance plans in any
16:27
state. You have
16:27
access to a cloud based platform,
16:30
that enables managers and employees
16:32
alike to quickly and securely
16:35
access benefits, payroll, and
16:37
other HR functionality from
16:39
anywhere, anytime. Plus,
16:41
there are no hidden costs. Take
16:43
a look at just works transparent pricing
16:46
by visiting just works dot com
16:48
slash pricing. That's just
16:50
works dot com slash pricing
16:53
or details. Introducing
16:55
the first TV brand streaming
16:57
Xbox games, Samsung Neo QLED
16:59
and OLED with gaming hub. With
17:01
Samsung Neo QLED and OLED TV,
17:03
you can stream your favorite game straight to your
17:05
TV. No console required. All you need
17:07
is a compatible controller and Internet. Play hundreds
17:09
of games from the biggest streaming services including
17:12
NVIDIA, GeForce NOW, Luna,
17:14
Nutomix, and box app. There's always something
17:16
to play in something fresh to discover with
17:18
thousands of new and classic game titles. Gain
17:20
access to games in seconds without worrying about
17:22
purchasing new hardware, down loads or storage limits, all you
17:24
need is an Internet connection and a compatible
17:26
controller. There is a special promotion going
17:28
through the end of the year. Purchase a twenty twenty
17:30
two Samsung Neo QLED or
17:32
OLED you get your first three months of Xbox
17:34
Game Pass Ultimate and an Xbox controller
17:36
on Samsung. Disclosure through twelve thirty
17:38
one while supplies last purchase a twenty twenty two
17:40
new LED QNEB or above or OLED
17:42
S95B you participate in retailers. For terms
17:44
and conditions, including redemption instructions, see controller
17:47
redemption dot com or see retailer. Control color may
17:49
vary. After getting passed trial, you will automatically be
17:51
charged and let's cancel. The mid one terms, x box
17:53
dot com
17:53
slash the friction terms.
18:00
Alexey Macarin's new study is
18:03
called simply social media, and
18:05
mental health, and it was published
18:07
in November of this year in the
18:09
American Economic Review.
18:11
Before the break, he told us
18:12
that their findings suggest a
18:14
bump in mental health problems
18:16
among students at a college
18:18
after Facebook is introduced
18:21
there. What that bump translates
18:23
to is a worsening
18:23
of what Alexey and his coauthors
18:26
called their index of poor mental
18:28
health by about eight
18:30
percent. Facebook access also
18:32
led to a nine percent increase
18:34
in depression among students
18:36
on college campuses. and a twelve
18:39
percent increase in generalized
18:41
anxiety disorders. Alexey
18:43
says the mental health effect of using
18:46
Facebook is about one fifth the size
18:48
of how it feels when you lose
18:50
your job. So
18:52
how does this new work which
18:54
suggests Facebook causes mental
18:56
health problems compared
18:57
to prior work, which only showed an
18:59
association between the two. Is
19:01
actually strikingly similar?
19:03
Why
19:06
is it that
19:07
social media has
19:10
this causal effect on mental health
19:12
issues? What's the mechanism?
19:13
we believe that Facebook users
19:16
engage in unfavorable social comparisons, and
19:18
that's what's driving the negative impact of
19:20
Facebook and mental health. And the
19:22
fact that we started Facebook at
19:24
its inception, it automatically rules out
19:26
a lot of stories that you might have about,
19:28
for instance, the like button, the news feeds,
19:30
the political news on social media.
19:33
All of that is nonexistent. Right?
19:35
So the fact that we observed that
19:37
Facebook had a negative impact all the
19:39
way in two thousand four, two thousand five, two thousand
19:41
six that suggests that sudden
19:43
in the nature of social media
19:46
leads to worse mental health
19:48
outcomes. And in our case, we believe that
19:50
it's due to unfavorable social
19:52
comparison? I mean, that's an interesting point to me because
19:54
you're looking at a period when Facebook was
19:56
really in its early stages. So
19:58
some of the features
19:59
are different now. It was really all
20:02
about
20:02
peer networks and social
20:05
comparisons back then. If you find
20:07
that on average, there's an x percent increase in
20:09
mental health problems in a
20:11
university that adopted Facebook.
20:13
Do you then scale that up by the fact that it
20:15
could be only two percent of people?
20:17
that
20:17
use Facebook or
20:19
not. Because that would tell us something about what
20:21
is the effect of a single
20:23
individual using Facebook who otherwise
20:25
wouldn't have used it. Yeah. That's
20:26
a great question. We don't scale up. We
20:28
believe that actually the use of
20:31
Facebook and college campus at the time was
20:33
almost universal and the adoption was
20:35
very fast. the common estimate that we
20:37
found is that about eighty five
20:39
percent of college
20:41
students had a Facebook account and we're using
20:43
it on a database. Wow.
20:45
Okay. So it's quite high. When
20:47
you set out to do the study, were
20:49
you expecting to find
20:51
an
20:51
adverse impact of Facebook introduction
20:54
on mental health. So what definitely surprised
20:55
me is how clean in the end this natural
20:58
experiment was. I did not expect it to
21:00
be as clean. because
21:02
when you work with the substation of
21:04
status, it's not like you're operating in the
21:06
control environment. When I started writing
21:08
the background, Section four or
21:10
paper, I didn't actually realize how bad
21:12
the situation is in the United States in
21:14
terms of mental health trends among
21:16
the the pictures that
21:18
you see in the data are
21:20
stark. So for instance, the percentage
21:22
of individuals from eighteen to twenty
21:24
three for report of experience in
21:26
a major Depressive episode in the
21:28
past year increased by eighty three percent
21:30
so almost doubled between twenty
21:32
eighteen and the fact
21:34
that this trend is sort of
21:36
coinciding with the spread of social media
21:38
is really begging the question whether
21:40
social media is to blame. It strikes me that
21:42
what you're doing, it does fit
21:44
into a broader important economic
21:46
question. Is this societal question of
21:48
what is
21:48
it that drives people's happiness or
21:51
economics we talk about utility? in
21:53
very standard economic
21:55
models of utility. We
21:57
don't typically factor in the
22:00
idea that our well-being may be a
22:02
function of what we perceive
22:04
the well-being or accomplishments of
22:07
others to be. And there is economic
22:09
literature out there, which would suggest that
22:11
in some cases, our own utility
22:13
is lower when the utility of others
22:16
is higher. one of the things that's coming out of social comparison is
22:18
that you're comparing accomplishments
22:21
relative to that of what you observe from
22:23
your peers. And
22:25
if you benchmark yourself in
22:27
that way, it's maybe not a surprise that
22:29
something like Facebook could actually lead
22:32
to declines in well-being. But what's even
22:34
more striking is that we're talking about
22:37
a situation where these individuals might actually
22:39
experience a clinical manifestation that
22:41
we call depression. So that's much more severe just
22:43
saying, okay, I'm unhappy because
22:45
Alex say just, you know, put on Twitter
22:47
that he published a paper in the American economic
22:50
review. one of the ways in which we
22:52
document that indeed
22:53
social comparison seems to be a channel
22:55
is that after the introduction of Facebook,
22:58
students suddenly believe
23:00
that the other students consume
23:02
much more alcohol than they did before.
23:04
But the actual usage of alcohol did
23:06
not increase at all. So students have
23:08
some needed there about how to
23:11
interpret the Facebook posts of
23:13
others, and they believe that everybody is having a
23:15
fantastic time, while in fact,
23:17
nothing changed. the sphere of these
23:19
alternatives to be a big component.
23:21
We also find that in the short to medium
23:23
run, the negative effects of Facebook
23:25
on mental health increase with the length
23:27
of exposure to the platform. And
23:30
actually, students reported
23:32
that their academic performance worsened
23:34
as a result of their mental health
23:36
issues. So
23:36
it's not just a time allocation issue. It's
23:38
that the time you spend on Facebook means
23:40
you're not spending a time studying, but it
23:42
also creates mental health problems
23:45
that make it harder to
23:47
perform academically as well as you were prior
23:49
to the introduction. We got
23:51
very lucky because in the survey
23:54
that we have, there is a specific
23:56
question that asks whether
23:58
your academic performance worsened as a
24:00
result of mental health such as
24:02
anxiety, depression. so far, when we analyzed
24:04
that question, we found that indeed
24:06
Facebook had this downstream impact
24:08
on Acadian
24:09
performance. If
24:11
you
24:12
were to say, you know,
24:15
here's why this study is important.
24:17
Here's what you should make of
24:19
it. and here's what you should with this information.
24:21
How would you articulate
24:22
that? This study is important because
24:24
we think that it's the most comprehensive
24:27
causal evidence on the impact of Facebook on mental
24:30
health. It's also important because we are able
24:32
to zero in on a
24:34
subset of people
24:36
who are particularly vulnerable and who
24:38
experience the most significant deterioration
24:40
of mental health in the past.
24:42
Fifteen years for college students
24:44
and young adults, and what can people do with this information
24:46
is that, first of all, they can be aware of it. So
24:48
they could potentially alter their
24:51
own usage of social
24:53
media and the usage of
24:55
social media by their kids or
24:57
friends. And second, we hope that social
24:59
media platforms sort of figure out a way find
25:01
a balance between maximizing their
25:03
ad revenue and potentially mitigating
25:05
the harms that social
25:07
media does.
25:10
We talk
25:12
about
25:12
why social media has made
25:14
mental health worse. THAT'S SOMETHING THAT
25:16
WE'RE ACTIVELY WORKING WITH SOCIAL MEDIA
25:19
COMPANIES AROUND HOW TO ADDRESS
25:21
THAT. THAT'S
25:21
THE SCIENTISTS DR. NINA
25:24
VASSON AGAIN. In
25:24
twenty eighteen, her lab at
25:27
Stanford called Brainstorm
25:28
started working with the social
25:31
media platform Pinterest.
25:32
they were
25:33
looking to understand different
25:35
patterns and what was driving people to their
25:38
platform. And what they actually saw was
25:40
that the fourth most common
25:43
category of search terms was
25:45
related to mental health, and it was things
25:47
like depression, stress,
25:49
anxiety. When people think of
25:51
Pinterest, they think about things like wedding
25:53
planning or interior
25:55
design or recipes. Pinterest
25:57
has not been a place that historically
25:59
people really
25:59
talked about I wanna go address my mental
26:02
health on this platform. What Pinterest
26:04
realized was, this is
26:06
something that people are coming to
26:08
us for. how can we
26:10
create a space for people to
26:12
be able to feel safe
26:14
and understood?
26:15
Around that same time, a fourteen year old girl
26:17
in England named Molly Russell
26:20
died
26:20
by suicide. Afterward,
26:22
it was discovered viewed more
26:24
than two thousand posts related
26:26
to suicide and self harm
26:29
on social media platforms, one
26:31
of which was Pinterest. This
26:33
was alarming and led to a lot
26:35
of questions around what are
26:37
kids engaging in, how do we keep them
26:40
safe, what are the interventions that can then be done? And
26:42
also, what is the role of social media
26:44
companies to keep users safe when things
26:46
like this happen? There are three big
26:48
things that we did.
26:50
The first
26:53
was what we call micro therapeutics.
26:56
We created a whole platform
26:58
of little mini tools little
27:00
activities that people can do anywhere from
27:02
one to five minutes that
27:04
take the same types of therapies
27:07
that we might do with our patients in
27:09
clinic. but instead that they can
27:11
do on the platform in real
27:13
time, not only things like mindfulness
27:15
and gratitude, but even leveraging
27:18
cognitive behavioral therapy and helping
27:20
people understand their thoughts and
27:22
thought patterns where they
27:24
can start thinking in a more positive
27:27
way. Now, when people go on the
27:29
Pinterest platform and they type in
27:31
depression or stress or anxiety,
27:33
they get an entire pop up
27:35
box. that's separate from their Pinterest
27:37
page, and it's very private.
27:40
The second thing that we did is
27:42
called compassionate search. if you
27:44
go into Pinterest and you start to type
27:46
in something related to suicide,
27:48
you're not now gonna get pushed
27:50
additional content into your inbox like
27:52
you did before. Some things will auto
27:55
fill, but things that we have been able
27:57
to flag as potentially
27:59
dangerous will not auto
28:01
fill So we're not further making things potentially
28:03
harmful for people. And then the third
28:05
thing was working with the algorithm
28:08
engineers specifically around
28:10
how to decrease the amount of self harm
28:12
content on the platform and
28:14
by educating the engineers around
28:17
what is self harm, what is self harm
28:20
content. After six months or so, there
28:22
was a
28:22
significant decrease in self
28:25
harm on
28:27
the platform.
28:33
in
28:33
medicine, we take this oath to do no harm.
28:35
And
28:35
that was really the idea that we
28:37
had around how do we change the user
28:40
experience everything from the way people
28:42
searched, the way that results get popped up
28:44
to them so that
28:45
the platforms are doing no harm to their
28:47
users.
28:50
Not long after Pinterest implemented
28:53
these changes, they noted
28:55
an eighty eight percent drop in
28:57
reports of self harm content by
28:59
users users. The company
29:00
said it was able to remove
29:02
this kind of content three
29:04
times faster than before. But
29:06
is it actually improving users
29:09
mental health? We don't know.
29:12
other
29:12
platforms have also tried to be proactive
29:14
about mental health. Meta,
29:16
the company that owns both Instagram
29:19
and Facebook, allows
29:21
users to flag worrisome posts
29:23
they've seen from friends.
29:25
In twenty seventeen,
29:28
Facebook develop its own algorithms to
29:30
identify users at risk of self
29:32
harm. And they have an
29:34
emotional health page containing
29:36
resources on conditions like
29:39
anxiety and depression. And
29:41
yet just last year, The
29:43
Wall Street Journal published
29:45
a series in which they claim
29:47
that Facebook has long studied and
29:49
known about the harm it
29:51
causes, particularly to mental
29:53
health. And that it has
29:55
continued to publicly deny and downplay
29:57
this information. In twenty
29:59
twenty
29:59
one, more than four billion
30:02
people people around the world
30:04
used some form of social
30:06
media. That's half of the entire population
30:08
of the planet. We've had
30:10
clues about its effects on our
30:12
mental health, but now we're starting
30:14
to get some definitive answers. What
30:16
we do
30:17
with that information is a
30:19
looming question. Nina Vassen thinks a
30:22
good place to start is with
30:23
data. When we look at
30:25
an area like mental health,
30:27
We have enormous data that we
30:29
use in order to do everything
30:31
from assessing someone's symptoms,
30:34
how depressed are they? How anxious are
30:36
they? But we don't really share this
30:38
information with patients. And I always
30:40
try to give an analogy to
30:42
something like diabetes. Where someone
30:44
with diabetes they are very aware
30:46
most often of their blood sugar. They're
30:48
measuring that regularly. They're
30:50
making food choices based on what their blood
30:52
sugar levels are. In mental health, we have
30:54
similar metrics around things like
30:56
depression anxiety scores, sleep
30:58
scores, but patients aren't actively
31:00
using these numbers. And metrics
31:02
What has not historically been a part of the conversation
31:04
with patients is for them
31:06
to understand what these
31:08
numbers are and to engage in
31:10
these numbers themselves. That's
31:12
it for today's
31:14
show. If you or
31:16
someone you know is struggling with
31:18
a mental health problem, It's
31:20
really important to seek help.
31:22
Talk to your primary care doctor
31:24
about finding appropriate services
31:27
or you can call the national suicide and
31:29
crisis lifeline. The number
31:31
to dial is 988
31:33
Trained workers are available
31:36
talk twenty four hours a day,
31:38
seven days a week, and all calls
31:40
are toll free and confidential.
31:42
I want to thank my guests, Alexia
31:45
Macquarie, and doctor Nina Boston. And thanks to
31:47
you, of course, for listening. Let
31:49
us know what you
31:49
thought about today's episode. How does
31:52
social media make you feel
31:54
good? bad or a little of both. Have
31:57
you ever tried to cut back on it to
31:59
improve how you feel? feel Send us
32:01
an email or a voice memo
32:03
We are at babu at freakonomix dot
32:06
com. That's
32:06
BAPU at
32:08
freakonomix dot com. Here's
32:10
an idea to leave you with about a
32:13
different way social platforms can affect our
32:15
health. Sexual
32:16
transmitted diseases or STD's are
32:20
a huge public health problem, particularly
32:22
among adolescents and young adults.
32:25
Over the last decade, a number
32:27
of new apps have
32:29
made it easier for people to
32:31
meet other people.
32:32
Could the staggered uptake of these apps
32:35
across cities tell us something
32:37
about their effect on STD rates?
32:39
Think about
32:39
it. And in the meantime, coming up
32:42
next week, they say
32:43
what's passed is prologue.
32:46
So what can the past tell us about
32:48
the present and maybe even the
32:50
future of our healthcare system?
32:52
It really shows
32:53
us how a significant
32:55
shock can play out even decades later.
32:58
And what
32:58
could the lasting effects be
33:00
of another more recent
33:03
shock. We've
33:03
seen a lot more people applying
33:05
and matriculating into healthcare programs.
33:07
More people applying to medical school, nursing
33:10
school. that's coming up next week
33:12
on FreakonomicsMD. Thanks again
33:14
for
33:15
listening.
33:16
FreakonomicsMD is part
33:18
of the Freakonomics Radio Network, which
33:21
also includes Freakonomics Radio,
33:23
no stupid questions, and
33:25
people I mostly admire.
33:27
All our shows are produced by Stitcher
33:29
and Renbud Radio. You can
33:31
find us on Twitter at doctor
33:34
BahooPod. This episode
33:36
was produced by Julie Canfor and
33:38
mixed by Eleanor Osbourne with
33:40
help from Jasmine Klinger.
33:42
Learick Boudic is our production associate.
33:44
Our executive team is
33:46
Neil Karuth, Gabriel Roth, and
33:48
Steven Dubner, original music
33:51
composed by Louise Gara. If you
33:53
like this show or any
33:55
other show in the Freakonomics Radio
33:57
Network, please recommend it to your
33:59
family and
33:59
friends. That's the best
34:02
way to
34:02
support the podcasts you love.
34:04
As always, thanks for listening.
34:08
Because I'm not a
34:11
native
34:11
speaker, I'm always kind of very self conscious
34:14
that I'm phrasing things not in a
34:16
fantastic manner.
34:18
Sometimes, Alexay. If they can make me sound good, they can make
34:20
anybody
34:20
sound good, so don't worry.
34:25
The Freakonomics
34:27
Radio Network, the hidden
34:29
side of everything.
34:32
Stitcher.
34:38
Introducing
34:40
the first TV brand streaming Xbox
34:42
games. Samsung NIO QLED
34:44
and OLED with Gaming Hub. With a
34:46
Samsung NIO QLED and OLED TV,
34:48
you can stream your favorite games straight to your
34:50
TV. No console required. All you need is compatible controller
34:52
and Internet. Play hundreds of games from
34:55
the biggest streaming services including Nvidia, GeForce
34:57
NOW, Luna, Nutomix, and
34:59
Xbox app. There's always something to play in something fresh to
35:01
discover with thousands of new and classic game
35:03
titles. Gain access to games in seconds without
35:05
worrying about purchasing new hardware, downloads,
35:07
or storage limits All you need is
35:09
an Internet connection and a compatible controller. There is a special promotion going through the end of the
35:12
year. Purchase a twenty twenty two Samsung
35:14
deal QLED or OLED, and you get
35:16
your first three
35:18
months of box game pass ultimate and an Xbox controller on
35:20
Samsung. Disclosure through twelve thirty one while
35:22
supplies last purchased a twenty twenty two new QLEDQNE
35:26
be or above or OLED S95 be your participating retailers. For terms and
35:28
conditions, including redemption instructions, see control redemptions dot
35:30
com or see retailer. Control the color
35:32
may vary after game pass you
35:35
will automatically be charged and let's cancel. The mid one terms, x
35:37
box dot com slash subscription terms.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More