What Should Policymakers Do about Social Media and Minors?

Barkley , Taylor

Introduction

Policymakers at the state and federal levels are clamoring for solutions to problems with minors and social media. In March 2022, in his State of the Union address, President Biden talked about “the harms of social media” on children and called on Congress to ban targeted advertising and strengthen privacy protections for children.¹ In September, California passed a bill, the first of its kind, that adds myriad new requirements to online platforms in order to keep users under 18 safe.² As of the end of 2022, three major bills in Congress have advanced through Senate and House committees that focus on or include portions intended to protect teens on social media.³ Finally, all member states in the European Union adopted the Digital Services Act, which requires platforms “to take appropriate and proportionate measures to ensure the privacy, safety, and security of children on their services.”⁴ Even with these significant events, this summary accounts only for the more serious proposals and enacted laws. Many more could be included.

Much of this action can be credited to calls for stricter regulations coming from pundits, organizations, and the media. The documentary film The Social Dilemma debuted in 2020 and has since garnered an estimated 100 million views.⁵ The Wall Street Journal’s publication of “The Facebook Files” in September 2021 prompted congressional hearings and high-profile interviews with former Meta employee Frances Haugen, and it brought to light Meta’s internal research on the effects their services might have on teens.⁶ It is a topic of society-wide importance.

Clearly policymakers want to do something about minors and social media. What are the problems they are seeking to solve, and what are the solutions that will keep children and teens safe online?

Lawmakers are currently most concerned about three issues on this topic: negative mental health effects, privacy (mostly in the realm of targeted advertising), and illegal content such as CSAM (child sexual abuse material) and drugs.

Policymakers at all levels of government should be concerned about the negative impacts new technologies might have on children and teens. Throughout history there are many examples of government officials recommending state action to prevent or correct harms against minors in the context of emerging technologies. As populations in the United States moved from rural to industrial occupations, labor unions and governments were at their best when advocating for workplace protections for and then bans on workers under a certain age. On the other hand, policymakers wasted a lot of time on concerns over comic books and video games.⁷ The level of skepticism and scrutiny is already high for new technologies.⁸ Policy responses often follow a predictable “panic cycle,” which usually ends in a healthy assimilation of new technologies into society.⁹ These panic cycles, as the Information Technology and Innovation Foundation describes them, proceed through four phases for many new technologies: trusting beginnings, rising panic, deflating fears, and moving on. Between rising panic and deflating fears there is the height of hysteria.¹⁰ The nature of the first half of this cycle can lead to overreaction, which will likely mean worse problems and certainly stymie future benefits.¹¹

Studies on Mental Health Effects Do Not Warrant Bans

Teens’ social media use and the potential negative mental health effects lead current policy discussions. Early concerns about health and well-being for minors on social media focused on inappropriate content and child predators.¹² In 2008, to protect teens and children from predators, the social media platform MySpace led efforts to study age verification and safety features; the results of those studies were then reported to state attorneys general. This effort ended with a final report in 2008 that found:

Technology can play a helpful role, but there is no one technological solution or specific combination of technological solutions to the problem of online safety for minors. Instead, a combination of technologies, in concert with parental oversight, education, social services, law enforcement, and sound policies by social network sites and service providers may assist in addressing specific problems that minors face online.¹³

Analysis of additional task force reports on the topic came to a similar conclusion.¹⁴ The earliest research studying the mental health effects of internet use was in 1998, and research on teens’ social media use came later, in 2014.¹⁵ The issue really broke out of the academic realm into the popular realm in 2017 when Jean Twenge published her book iGen and wrote a popular article in The Atlantic, “Have Smartphones Destroyed a Generation?” This work, concurrent with the rise of the “techlash” in 2016, brought the issue of teen mental health effects to prominence and even sparked “techlash 2.0,” according to former USC researcher Nirit Weiss-Blatt.¹⁶ Many additional books, articles, studies, films, and advocacy efforts between 2017 and the present date together built to a crescendo, with comparisons to tobacco and claims of the loss of free will.¹⁷

Teens use social media sites regularly, though habits and popular platforms change over time. In 2022, according to Pew Research, YouTube was the overwhelming favorite, with 95 percent of teens “ever visiting” the app and 19 percent using or visiting it “almost constantly.” TikTok, Instagram (owned by Meta), and Snapchat were next in line. Among teens, nearly all of whom have access to smartphones, Facebook (owned by Meta) is only half as popular as Instagram, and Twitter is even less popular.¹⁸

Such near-ubiquitous use, combined with increases in teen suicide and in negative mental health indicators, has prompted inquiries into whether these technologies are having an impact on the mental and physical health of teen users. When a chart showing rising suicide rates is overlayed with rising social media use, the narrative can be quite convincing. However, this could be an oversimplification, given the dozens of factors that put teens at risk of mental health issues. Researchers have yet to adequately isolate social media as a potential cause among the variables.¹⁹ In fact, the American Psychological Association has changed their official position on the link between violent video games and violent behavior. In 2015 they reported a link between the two, but in 2020 their position was updated to state that there is “insufficient scientific evidence to support a causal link between violent video games and violent behavior.”²⁰

Despite the headlines and dominant narratives, a conclusive, causal relationship between social media and negative mental health effects for the typical teen user has not been made.²¹ The linkage is correlational at best. A summary of evidence linking social media use to negative mental health was provided by the social psychologist Jonathan Haidt in congressional testimony in 2022. Haidt is a leading proponent of this connection and in his testimony stated, “Correlational studies consistently show a link between heavy social media use and mood disorders, but the size of the relationship is disputed.”²² The most significant, although still disputed, connection is among females aged 12–15 and 19–21.²³ Even suicide rates—among the most tragic of statistics—do not offer conclusive narratives. Suicide rates among boys in the early 1990s were much higher than they are now. For girls, suicide rates are slightly higher now, but not by a large margin.²⁴ There are less data on self-harm and depression among teens over time, and there are still questions as to why rates went up or down.²⁵

To further complicate the matter, “teens themselves paint a more nuanced picture of adolescent life on social media,” as the November 2022 Pew survey on “Connection, Creativity, and Drama: Teen Life on Social Media in 2022” summarized it.²⁶ Even in the famous “Facebook files” published by the Wall Street Journal and discussed by Frances Haugen in media appearances and congressional testimony, the number of teens who said Instagram made them feel better about themselves was almost double the number of those who said it made them feel worse.²⁷ Experience varies from individual to individual, and teen testimonials reflect this. In a mix of interviews, surveys, and focus groups with educators and teens conducted between 2017 and 2021, MIT researchers Emily Weinstein and Carrie James came to this conclusion about the dominant narrative of negativity: “You’ve been misled. Alarmist headlines that shape the public narrative about teens and tech often lead us astray.”²⁸

Nonetheless, these matters are of serious concern and more research is being conducted. However, the tenuous evidence does not warrant strong government action, and if states or the federal government were to enact strong rules, the costs will likely outweigh any potential benefits.

The Harms of Data Collection for Teens Are Not Clear

Privacy for teens is a policy area that Congress and federal regulatory bodies like the Federal Trade Commission (FTC) have already addressed. In 1998, Congress passed the Children’s Online Privacy Protection Act (COPPA) and delegated enforcement to the FTC. Essentially, Congress decided that any products or services designed for users under the age of 13 need to be compliant with much stricter rules and require parental consent. In effect, most platforms and services are designed for users age 13 and up. There are proposals from policymakers and experts that recommend raising the age threshold of COPPA so its requirements would apply to teens. However, the costs of COPPA would apply to more users, and the age verifications discussed below would apply. Of concern are the growing number of children ages 8–12 who report using social media platforms: 38 percent in 2021, up from 31 percent in 2019, according to Commonsense Media.²⁹

As is the case for all users, the harms of data collection among teens must be made clear. For instance, the American Data Privacy and Protection Act includes a portion that prohibits targeted advertising to minors known by the platform to be under the age of 17. Outright bans on television advertising to children were famously considered and then rejected by the FTC during the “KidVid” proceedings in the late 1970s and early 1980s. This rulemaking procedure was a three-year process that produced 6,000 pages from stakeholders at all levels of civil society, academia, and government. Ultimately it was not clear that a ban on television advertising for children would be effective or helpful, and the rulemaking was abandoned.³⁰ Of course, technologies today are more ubiquitous and much more advanced than the television of 50 years ago. However, the base assumptions about advertising to children and its effects are similar. Restrictions then and now would likely diminish the welfare of consumers, including teens.³¹

Finally, targeted advertising for teens is not universally nefarious. Benefits could include educational opportunities and mental health resources. Targeted advertising can even promote safety on these platforms because advertising can be vetted as age-appropriate. A ban on teen-targeted advertising means teens will likely see advertisements meant for adults. Ultimately, for the policymaker concerned about teen data collection, any system that would restrict teen use of social media for mental health reasons would stand in contradiction to those principles.

More Resources Are Necessary to Enforce CSAM

In most policy areas, policymakers should focus on the clearest and most egregious problem first. As outlined previously, there are three major issues of concern: negative mental health effects, privacy (mostly in the realm of advertising), and CSAM (child sexual abuse material). In a ranked order, CSAM mitigation and the marketing of illegal drugs are incontrovertibly in first place. Next are privacy issues for minors, and last would be negative mental health effects. Such a rank order is not to diminish concerns over privacy or mental health. Rather it is a means for policymakers to focus limited time and energy among myriad competing priorities at the local, state, and federal levels.

Therefore, policymakers concerned for the safety of minors online should first focus on combating CSAM. To be clear, social media can be a place where CSAM is developed and trafficked, but the problem is much larger than mainstream social media platforms.

Indeed, tech companies of all sizes are engaged in removing known CSAM images and videos thanks to the work of the National Center for Missing and Exploited Children (NCMEC) and various voluntary standards and organizations. Since 1998, NCMEC has maintained a cyber tip line for individuals and companies to report CSAM. In 2021 there were 29 million reported instances of CSAM, up from nearly 17 million in 2019.³² The sheer volume of images and videos online makes distinguishing CSAM from other images a technical challenge, so NCMEC maintains a library of known CSAM material marked with unique hash values, which are basically digital fingerprints that can be automatically screened. Private companies then use that information, as well as information shared with other companies, to remove the content.³³

As stated, this issue area can itself be triaged, and there are pitfalls to avoid. Make no mistake: the vileness of CSAM prompts solutions such as bans on live streaming, encryption, or peer-to-peer file sharing. All those technologies are used every day by law-abiding citizens but they are also relied on by CSAM consumers and traffickers. Well intentioned attempts to combat CSAM can result in more harm than good. For example, the EARN IT Act did not outright ban encryption, but the requirements would result in a de facto ban.³⁴ Yet there are solutions, as mentioned, that retain the technologies’ benefits while mitigating harm with increasing effectiveness.

Ultimately, CSAM is a growing problem; the effects are devastating to victims, and law enforcement is already overwhelmed. There remains much to be done.

The Trade-offs of Age Verification Requirements

Fundamentally, any sort of public policy-based ban, restriction, or requirements for minors will require platforms to know the ages of those using their platform. Policy organizations and pundits have suggested states or the federal government place restrictions on minors’ access to social media during certain times or fully ban social media use by minors.³⁵ Banning minors from social media assumes the negative effects of social media outweigh the benefits and that any nongovernmental solution is inadequate. This is likely not the case, and outright bans have clear downsides. For example, proposed bills like HB 896 in Texas would prohibit those under 18 from using social media, making any teen who circumvents the age requirement a lawbreaker.³⁶ Either way, COPPA as currently written means we live in a world where users over 13 need to be verified.

Current industries like the adult entertainment, gambling, and sports betting industries already have to deal with such legal requirements. Any age verification method necessitates more data will need to be collected by the companies or a third-party firm. Far from alleviating concerns regarding minors and social media companies, these efforts may create more problems, certainly on issues of privacy and cybersecurity. Additionally, in order to be compliant, age requirements for social media platforms would mean that all users of that platform will be subject to more data collection. Just as internet users are prompted by most websites to “accept all cookies” due to European regulations, all users of social media platforms would be subject to more onerous age verification methods. On platforms as widely used as social media platforms, everyone would be subject to verification.

There are efforts underway to catalog and track age verification methods, but the results do not look promising. For the paper “Making Sense of Age Assurance,” the Family Online Safety Institute (FOSI) polled parents and children on seven types of age verification techniques. In order of popularity: parent verification with text or app, self-declaring date of birth, biometric verification, verifying parental identity with ID, verifying child identity with ID, credit/debit/payment information, and audio or video call verification.³⁷ The French technology industry regulatory body, Commission Nationale de l’Informatique et des Libertes (CNIL), conducted an evaluation of known verification techniques to ensure they include “the following properties: sufficiently reliable verification, complete coverage of the population, and respect for the protection of individuals’ data and privacy and their security.” The only addition to the list of seven types outlined by FOSI was an “offline verification system” like a single-use scratch card purchased at a supermarket used to verify the age of a user. They concluded “there is currently no solution that satisfactorily meets these three requirements.”³⁸

To illustrate the trade-offs, it is helpful to examine only some of the potential steps necessary to implement the most popular method among parents and children in the FOSI report: “parent verification with text or app, e.g. form, text verification, app push notification to approve.”³⁹ In this hypothetical scenario, a platform could require all users who denote their age under 18 to supply the contact information or app of a parent. Of course, age verification would need to be reliable for the parent or alleged parent with the contact information or app. In such a case, it would be the parents or alleged parents who need to provide a more vigorous form of identification, which poses privacy and security risks. It would put platforms in the role of verifying parenthood status of millions of individuals. Furthermore, implementation of this method would require children to alert their parents de facto by signing up for the platform’s services. There could be cases like those of LGBTQ+ youth or youth in abusive parental situations who might not have the ability to safely alert their parents when signing up for a service or app. Indeed, this is precisely why over 90 groups wrote a letter of opposition to the Senate Kids Online Safety Act (KOSA).⁴⁰ As Emma Llanso at the Center for Democracy and Technology said about KOSA, the same can be said for parent verification: it presumes parental oversight is “universally a good thing.”⁴¹

In all likelihood, age verification consistent with the principles of reliability, complete coverage, privacy, and the Constitution is impossible to achieve. All current methods have trade-offs. However, firms are working on implementing solutions with those trade-offs in mind. For example, Meta is currently experimenting with age verification tools on Facebook Dating and Instagram.⁴² Lessons can be learned from industries like online gambling on the challenges and successes of user age verification. Policymakers can use their platform to gather information on functional experiments and to convene practitioners for collaboration. KOSA contains a provision directing the multiple federal agencies with relevant jurisdiction to do just that.⁴³ However, forcing platforms to comply with new age requirements before an effective age verification solution is found is likely to result in more data collection of teens or in teens losing access to platforms they use and enjoy.

What Should Policymakers Do?

Although the harms in mental health and data collection are disputed, the following are some of the many actions policymakers can take to keep teens safe on social media.

Allocate more funding to law enforcement to combat CSAM at the federal, state, and local levels. Law enforcement agencies are overwhelmed by the amount of CSAM material.⁴⁴ Among other provisions, the Invest in Child Safety Act would allocate 5 billion dollars to the Department of Justice and allow the hiring of more personnel at the FBI and NCMEC to deal with the caseload.⁴⁵ The Child RESCUE Act would establish a federal working group to assess the problem and make recommendations to Congress.
Provide more resources and training for officers at the federal, state, and local levels who handle CSAM cases. For officers at all levels working on cases, sufficient mental health support needs to be made available. One study indicated that stigma against mental healthcare was “the most widely acknowledged barrier discussed in relation to police wellness.”⁴⁶ Policymakers can use their platform to push back against this stigma.
Companies should be given more time to report instances of detected CSAM. The END Child Exploitation Act would “increase the amount of time—from 90 days to 180 days—that a provider must preserve the contents of a report” made to the CyberTipline managed by NCMEC.
Policymakers should empower those closest to the teens who need help, be they parents, caregivers, educators, or teens themselves. Policymakers can and should use their public platform to educate parents, caregivers, and minors on safe social media practices. Parents, caregivers, and local communities, regardless of any future policies, will remain at the front lines of helping teens stay safe online. According to the Family Online Safety Institute (FOSI), three out of four American parents believe the primary responsibility for their teens’ online safety rests with the parents, and two out of four believe it is primarily the government’s responsibility; however, at similar rates, parents want the tech companies and governments to support their efforts.⁴⁷ Two authors with very different perspectives on this issue, Nir Eyal and Jonathan Haidt, co-wrote a list of three common sense screen time rules parents can use with their children.⁴⁸ Work such as theirs can be of use in establishing healthy tech habits.
Policymakers, the media, and the public should hold companies to a high standard and join to call attention to the resources the large platforms do and should provide for parents and caregivers. Indeed, recent changes made by Instagram and Apple seem to demonstrate that public pressure is starting to be felt. It is the companies’ responsibility to ensure parents and caregivers are aware of and are comfortable using their tools. However, parents and caregivers are largely unaware of or are overwhelmed by the tools and resources these companies provide.⁴⁹
Given the inconclusive nature of social media’s impact on the mental health of kids and teens, more studies should be encouraged and perhaps funded by the government. At the federal level, the recently passed Children and Media Research Advancement (CAMRA) Act will direct funding to the National Institutes of Health to conduct longitudinal research on this topic. Research over a multi-year time frame is important to avoid the “Sisyphean trap” (incompletable task) of technology regulation.⁵⁰ States should consider similar, supportive efforts.
Policymakers should not vilify companies that actively study these issues and seek to develop safe apps and services for teens. Meta was critiqued for the results of a study they conducted on the impact their services were having on teens and the Instagram for Kids app project was shut down due to public pressure.⁵¹ Universal benefit to all users is not a realistic goal at the scale of most platforms. If that were the case, far more prosaic technologies like bicycles would face stricter rules. As with any technology, some users will experience downsides.

Conclusion

Policymakers are right to pay attention to the impact new technologies are having on teens. Although sweeping regulations like banning teens from using social media are not recommended, policymakers still have multiple tools they can deploy in efforts to keep teens safe on social media. As the horrors of CSAM and other illegal content demonstrate, there is plenty of work to do. Policymakers should start where the evidence of harm is clearest and work outward, all the while collecting data and advocating for bottom-up solutions that teens, parents, and caregivers can adapt to meet their unique needs. Teen life throughout modern US history has been fraught, but the government should avoid preempting opportunities for teens to learn and grow. Rather, as young people become adults, they should be taught the skills necessary to make wise decisions that are right for them, in the context of their communities.

What Should Policymakers Do about Social Media and Minors?

Executive Summary

Resources

Focuses

Federal Agencies

Introduction

Studies on Mental Health Effects Do Not Warrant Bans

The Harms of Data Collection for Teens Are Not Clear

More Resources Are Necessary to Enforce CSAM

The Trade-offs of Age Verification Requirements

What Should Policymakers Do?

Conclusion

Executive Summary

Resources

Cited by

Related Publications

Related News

Related Blogs

Focuses

Federal Agencies

Introduction

The Concerns about Social Media, Children, and Teens

Studies on Mental Health Effects Do Not Warrant Bans

The Harms of Data Collection for Teens Are Not Clear

More Resources Are Necessary to Enforce CSAM

The Trade-offs of Age Verification Requirements

What Should Policymakers Do?

Conclusion