Catching up with ‘Coddling’ part two: Trigger warnings, screen time v. social media, COVID-19 and the continuing decline of Gen Z's mental health

EDITOR’S NOTE: This is the second part of a multi-part series updating developments since the publication of “The Coddling of the American Mind: How Good Intentions and Bad Ideas are Setting up a Generation for Failure” (2018). See part one here. Future articles can be found here. Special thanks to Jonathan Haidt for his feedback and edits.

If you care about the decline in mental health among young people, which Jonathan Haidt and I discuss in “The Coddling of the American Mind” (COTAM), be sure to check out this month’s cover story for The Atlantic magazine, “The Anxious Child and the Crisis of Parenting,” by Kate Julian — a deep dive into what is causing the spike in anxiety among children. It’s full of important information and insight focusing on what we called “paranoid parenting” in the book. One term Julian discusses is especially crucial to our work: accommodation. 

“Accommodation” here refers to the accommodating behaviors that parents engage in to decrease their children’s anxiety, but which may make the long-term problem worse. For example, letting your child sleep in your bed because she is afraid of the dark might provide short-term relief to your child, but it prevents her from learning how to overcome her fear of the dark and can lead her to believe that she is powerless to overcome that fear.

The problem of parental accommodation is tightly related to two key concepts we explored in COTAM: antifragility and safetyism. Antifragility is a term coined by Nassim Taleb and refers to the fact that many systems (such as the immune system) need stressors in order to grow strong. By accommodating your child’s anxieties, we actually empower those anxieties, which can stunt or stop the natural coping responses that children must develop in order to overcome them. 

Safetyism refers to a culture or belief system in which safety is prioritized above all other values, which means that people are unwilling to make tradeoffs demanded by other practical or moral concerns. Safety is transformed into a sacred value, which means no tradeoffs can be made, no risks can be taken. This is generally accompanied by the tendency to equate emotional discomfort with physical danger. 

As we wrote in COTAM: 

When children are raised in a culture of safetyism, which teaches them to stay “emotionally safe” while protecting them from every imaginable danger, it may set up a feedback loop: kids become more fragile and less resilient, which signals to adults that they need more protection, which then makes them even more fragile and less resilient.

The culture of safetyism, as we wrote in the book, “deprives young people of the experiences that their antifragile minds need, thereby making them more fragile, anxious, and prone to seeing themselves as victims,” and is based on a fundamental misunderstanding of trauma, recovery, and human nature.

Accommodation & Safetyism on campus

Dealing with the effects of safetyism and the accommodations that arise from it are central themes in the work we do at the Foundation for Individual Rights in Education (FIRE) defending free speech, academic freedom, and due process rights on college campuses.

Speech codes, for example, are frequently described as necessary to protect students from intolerant speech, especially students the institution perceives as vulnerable. In a recent FIRE survey of 198 colleges, we found 77% use automatic filters to restrict speech on social media, and in our most recent survey of campus speech codes, nearly 9 out of 10 of the 471 schools we surveyed maintained at least one policy that restricted some form of expression protected by the First Amendment. These codes do nothing to eliminate intolerance, which cannot be effectively confronted when it is not expressed. Instead, the codes enshrine what is often an ongoing First Amendment violation as a policy that serves to increase group polarization and radicalization by ensuring the intolerant will only speak to those who agree with them.

Or consider the recent case at Truman State University in Missouri, where a group of students and administrators denied recognition to a student group dedicated to animal rights and veganism. The reason? Because club members might face “emotional” harm when they encountered “hostile” opposition to veganism. The club was recognized after FIRE intervened. 

The special case of trigger warnings

One of the many things that surprised us after the publication of our original 2015 Coddling article was the fervor with which people critiqued our position on trigger warnings — almost to the exclusion of critiquing any other concerns we raised. Trigger warnings were just one of several trends we criticized, but, as I’ve discovered on many campuses, the belief that they help is an article of faith.

Telling someone that you need to warn them about something they’re about to read can be akin to playing scary music in a horror film — it primes the listener for something scary or threatening.

After a talk I gave at Carleton College in 2016, during which I discussed my concerns about trigger warnings, I was approached by a very upset student who believed trigger warnings were a good and compassionate practice. I want to be very clear about my position then and now. We weren’t offhandedly dismissing the trigger warnings as “snowflakey” (“snowflake” is a term Haidt and I very much dislike). We were saying something different and much more concrete:

1. There was no evidence that trigger warnings help those suffering from PTSD;
2. There are evidence-based, psychological reasons to believe trigger warnings may in fact be harmful; and
3. Trigger warnings put professors in a difficult position and could chill speech about difficult topics in the classroom.

In 2016, we didn’t have sufficient research to back up our hunches, but now that there have been at least five studies on the subject, it does seem clear that point number one is true: Trigger warnings provide no benefit.

As to point number two, the only study to specifically focus on people with PTSD indicates that trigger warnings may increase anxiety. This should not be a huge surprise, as telling someone that you need to warn them about something they’re about to read can be akin to playing scary music in a horror film — it primes the listener for something scary or threatening.

One of the first of these recent studies indicated that those who believed that words are harmful in and of themselves were more likely to become anxious when presented with distressing passages of text. This finding has not been replicated, but in the same study, researchers from Harvard found that the presence of trigger warnings could increase the belief that trauma would be permanently psychologically damaging. 

Just like parents making accommodations for their anxious children may make anxiety worse, the same pattern may be repeating in higher education.

This is of particular concern because while the experience of trauma is relatively common, PTSD is relatively rare. Most people recover from traumatic experiences without developing PTSD. And most of those who do suffer from PTSD eventually recover. If trigger warnings serve to increase students’ tendency to see threats where no threats exist, this could actually increase their risk of developing PTSD in the event of trauma.

As for the third point, there has been no study done to our knowledge that tests whether trigger warnings inhibit professors from covering certain topics or students from discussing taboo topics in class, but there is surely strong anecdotal evidence that that is happening.

In her 2014 article “The Trouble With Teaching Rape Law,” published in The New Yorker, Harvard professor Jeannie Suk Gersen wrote that “[a]bout a dozen new teachers of criminal law at multiple institutions have told me that they are not including rape law in their courses, arguing that it’s not worth the risk of complaints of discomfort by students.” Also, seven humanities professors (teaching at institutions including Princeton, Georgetown, and UC Davis) co-authored an opinion piece for Inside Higher Ed, stating that even in 2014, trigger warnings were “already having a chilling effect on our teaching and pedagogy.” And in 2016, James Madison University professor Alan Levinovitz noted in The Atlantic that trigger warnings and safe spaces were inhibiting student speech in the classroom. The “spirit of tolerance and respect that inspires these policies,” he wrote, “can also stifle dialogue about controversial topics, particularly race, gender, and, in my experience, religious beliefs.”  

So trigger warnings are at best neither helpful nor harmful. Increasingly, they have been shown to either provide zero benefit, or have a negative effect. Nonetheless, according to one survey, about half of professors use them. Just like parents making accommodations for their anxious children may make anxiety worse, the same pattern may be repeating in higher education. This should not be too surprising, as a lot of these accommodations communicate the implicit message “I don’t believe you can really handle this.” That’s bound to make anyone feel less empowered and more anxious. 

We will revisit Kate Julian’s excellent article in greater detail when we cover “paranoid parenting,” the third of the six causal threads we discuss in COTAM. In short, the behaviors we address are generally those of parents and campus administrators. There is nothing intuitive about exposing the people you love most in the world — your children — to things that could hurt them, emotionally or otherwise. It’s not easy, and none of us is exempt or immune. But in COTAM, we examined what happens when we don’t.  

The rise in anxiety and depression since the publication of COTAM

When Haidt and I first started working on the original article back in 2014, we at FIRE were hearing reports of skyrocketing rates of anxiety and depression among college students, in the news as well as from friends and contacts we had on campus. By Haidt’s high standards the data was not yet conclusive. After we published the article in 2015, however, the data came in loud and clear, and the modest bump in anxiety and depression that we expected was far more dramatic.

What we discovered in COTAM was that rates of anxiety, depression, and self-harm began rising around 2012, and that they kept rising fairly steadily right up through the latest data available to us in early 2018, as we were finishing the book. There were a record number of students suffering from anxiety and depression, with rates of teen suicide rising rapidly as well. When the book came out, the most recent then-available data from 2016 showed 12.8% of youths aged 12-17 had a major depressive episode in the prior year (that is, exhibited five of nine symptoms of depression every day for two weeks or longer); now, data from 2018 shows a new high of 14.4%. But to really understand the trends, we need to look at the data by gender.

Let’s take a look at the first graph we have updated from the book relating to depression rates for 12 to 17-year-olds. In COTAM, we found that about 19% of females in the age group had experienced a major depression in 2016 — an increase of just over 60% from the 2010 rate. As you can see, the most recent data from 2018 now places the number even higher at about 21.5%. The rate for men was about 6.6% in the book (which was about 55% higher than the 2010 rate), and stands at nearly 8% in the most recent data. Overall, that means that in the two years of data made available since COTAM, the depression rate for girls increased by almost 10%, and the rate for boys increased by over 20%. (You can find a comprehensive list of published studies, with graphs of trends in the USA, UK, and Canada, in this open source literature review curated by Haidt and San Diego State University psychology professor Jean Twenge.)

Now let’s look at the grimmest statistic, the rise in suicide. When we published COTAM, the suicide rate for young women was already at a high of 5.09 per 100,000. It peaked at 5.35 in 2017, and decreased to 5.21 from the most recent data.

One disturbing trend we did not see coming was the sudden spike in the suicide rate for young men in 2017. In the book, the most recent available data came from 2016, showing a rate of 14.81 per 100,000 male teens. The suicide rate for young men spiked to 17.94 in 2017, the highest rate in nearly 30 years and approaching the highest ever recorded (18.1, in 1991). In 2018, it dipped slightly to 17.32, fortunately not rising any further, but also not returning to the 2016 level. 

Julian and others have speculated that the increase may be related to the opioid crisis, with the suicides of young men reflecting a larger rise in “deaths of despair.” Julian raises a second possibility: that the 2004 addition of warning labels on once-overprescribed antidepressants, noting that they could cause suicidal ideation in teenagers, may have led to an undertreatment of adolescent depression. 

These trends are reaching younger age groups, too. In 2016, the suicide rates for males and females in the 10-14 age range was 2.52 and 1.69 (per 100,000), respectively. Now, we have 2018 data showing those rates are 3.66 and 2.02, respectively. The 2018 suicide rate for boys is more than double, and the rate for girls nearly triple, the average from 2001 to 2010.

One factor not reflected in this data is the effects of the coronavirus on young adults. At present, it’s too early to tell how that will impact mental health in the long term. In the short term, however, an increase in anxiety seems all but inevitable. A survey of 500 students from 129 colleges, conducted around the third week of March, found 75% reported even higher anxiety than before the virus. We will have to see whether that has increased or decreased as people have settled into lockdown and social distancing measures.

The role of social media and ‘screen time’ in the rise of anxiety and depression

The role of smartphones, screen time, and social media in the rise of depression and anxiety among Gen Z has been the most hotly contested and researched issues we brought up in the book. In chapter 7, where we presented graphs such as the ones above, we focused on two main causes: overprotection and social media. In a section titled “Anti-social media,” we discussed the work of Jean Twenge, particularly her analysis of large datasets that showed that teens who were heavy users of smart phones and other digital technologies had much worse mental health than those who were light users or non-users. In the following section, titled “Why is it mostly girls who suffer?,” we tried to explain why mental health has deteriorated so much more rapidly for girls than for boys. We focused on the fact that girls use social media platforms (such as Instagram) far more than boys (who spend far more time gaming). We offered four reasons why social media is particularly harmful for girls:

1. Girls may be more adversely affected by the constant display of perfect “curated” lives;
2. Girls may be more sensitive to FOMO (the Fear of Missing Out);
3. Girls may be more harmed by the constant focus on beauty and appearance; and
4. Boys' aggression is more physical, while girls' aggression is more relational (and is therefore amplified by social media). 

We were appropriately careful about offering advice to parents, based on the state of research in 2017. In a section titled “Screen time: A caution about caution,” we wrote

One conclusion that future research is almost certain to reach is that the effects of smartphones and social media are complicated, involving mixtures of benefits and harms depending on which kinds of kids are doing which kinds of online activities instead of doing which kinds of offline activities.

And we said 

So we don’t want to create a moral panic and frighten parents into banning all devices until their kids turn twenty-one. These are complicated issues, and much more research is needed. In the meantime, as we’ll say in chapter 12, there is enough evidence to support placing time limits on device use (perhaps two hours a day for adolescents, less for younger kids) while limiting or prohibiting the use of platforms that amplify social comparison rather than social connection.

One day after our book was published (by coincidence) a psychiatrist wrote an essay in the New York Times titled “The big myth about teenage anxiety.” Three months later, two empirical studies were published asserting that there is in fact no more than a trivial relationship between screen time and mental health problems. In that same month, other studies were published reaffirming a relationship between social media (specifically) and mental health problems, particularly for girls. 

We think the case is now much stronger that girls who become heavy users of social media have worse mental health, and the relationship is not just correlational, it is causal.

In response, Haidt created two open-source literature reviews where he pasted in the abstracts of all relevant studies he could find, on both sides, and he invited other researchers to add to the document. The first literature review assembled all published data on the prevalence and changes over time in rates of depression, anxiety, self-harm, and suicide. The second literature review assembled all published research bearing on the question of whether social media was a contributing factor. Both documents grew rapidly, and we can now say that a consensus is emerging:

1. 1. There really is a mental health crisis engulfing adolescents, particularly girls, in all of the major English-speaking countries. There was hardly any pushback on the mental health literature review.
      
   2. The evidence linking “screen time” to mental health problems, when you combine both genders together, is indeed weak and inconsistent, as our critics charged. But when you limit the analysis to social media effects on girls, the link is much stronger, and is surprisingly consistent, even found in many of the papers that say there is little problem with “screen time.” Furthermore, the data are not just correlational. We found ten true experiments, using random assignment to a condition where people greatly reduce their use of social media for one or more weeks. Seven of the ten show a significant positive effect on variables related to well-being. 

You can read the details about the state of debate in this article at Medium, by Markham Heid, or in this Twitter thread from Haidt.

The kids themselves often recognize the addictive and socially harmful aspects of some of their online activities.

So were we right? In our focus on social media and girls, yes. That link has held up well. We think the case is now much stronger that girls who become heavy users of social media have worse mental health, and the relationship is not just correlational, it is causal. Parents should try to delay the day when their kids (especially daughters) get accounts on Instagram and other platforms where people rate them and comment on their looks and their posts. They should talk with their children about the dangers of these platforms, and they should consider tight time limits on their use (such as one or at most two hours a day). School districts should help parents by strongly suggesting the norm that no kids get social media accounts until high school, to avoid the fear of exclusion that leads many parents to allow their kids to lie about their age and open an Instagram account in sixth grade, or even fifth grade. Middle school is already so hard on kids, and social media makes the worst parts even worse. 

But our suggestion that parents should consider placing “time limits on device use” needs to be modified. The evidence we have found does NOT show that five hours a day of “screen time” or “digital device use” causes depression or anxiety. Some “screen time” is clearly good for mental health, such as Facetime, Zoom, or simple phone calls, because they allow teens to communicate directly, privately, and synchronously. (This is particularly important during the COVID pandemic and our time of physical distancing.) Other kinds of “screen time” are clearly educational or entertaining, such as watching videos on Netflix or taking a course at Khan Academy. The very concept of “screen time” is nebulous and not very helpful. 

But with that said, parents may still want to place limits on total daily use of particular devices or apps, or even on total “screen time,” for the simple reason that these things all have an “opportunity cost.” If your son is spending eight hours a day playing video games and watching youtube videos, these two activities will then push out nearly all other activities, every day, year after year. Add in the fact that many of these things are designed by psychologists in Silicon Valley for the specific goal of keeping users “hooked” for as long as they possibly can. Given these facts, you should talk with your kids and ask them how they want to spend their days. The kids themselves often recognize the addictive and socially harmful aspects of some of their online activities. Working out reasonable limits, together, will help your child cultivate good habits and obtain a wider range of experiences—perhaps even including unsupervised outdoor play. 

We’ll look at social media and its unintended consequences again in our next causal thread: political polarization. 

Conclusion: COVID-19, accommodation and antifragility

Now, let’s apply our analysis of safetyism, antifragility, and accommodation to the potential mental toll that the COVID-19 crisis could take on young people. Under a framework of safetyism and the Great Untruth of Fragility (what doesn’t kill you makes you weaker), this crisis could be framed as something devastating from which young people will never really recover. It could be seen as permanently damaging. 

This unprecedented crisis could leave many young people feeling as though they have made it through a crucible and grown from it.

The trouble with this kind of thinking is that, by ignoring the importance of antifragility, it can become a self-fulfilling prophecy. Telling young people that adversity will lead to inevitable, permanent harm is like teaching them the opposite of the first Noble Truth of Buddhism: “Life consists of suffering, pain, and unhappiness.” Accepting that life will include pain and unhappiness can actually make life feel more bearable when we’re not happy. Conversely, giving young people the message that experiencing pain or unhappiness means they are doing something fundamentally wrong is a formula for making anybody miserable. 

It’s like the difference between repeatedly saying to a child, “I know you’re hurting. But I know that you will get through it. It’s a part of life, and you really can handle it,” versus repeatedly saying, “If you’re hurting, it means that something is seriously wrong with you. Or maybe there’s something wrong with your life. Either way, you really can’t handle any of this.” 

This unprecedented crisis could leave many young people feeling as though they have made it through a crucible and grown from it. Hopefully any short term increase in anxiety will give way to a rising tide of empowered, competent, antifragile young people with a sense of self-efficacy and an internal locus of control. 

But that won’t happen if, in an effort to help and protect our children, we give them the message that we don’t think they are up to the challenge.