US State Policies Regarding Social Media: Do Policies Match the Evidence?

Abstract

Policy Points.

• State policymakers have moved rapidly to regulate adolescent social media use, frequently stating concerns about mental health, harmful content exposure, and developmental impacts.

• The degree to which policymakers’ arguments in favor of state social media regulations correspond to the state of the current scientific evidence remains questionable.

• The evidence to substantiate policymakers’ assertions has substantial limitations in the ability to answer causal questions, but some promising directions are emerging around targeted protections for highly susceptible youth.

Context

The potential adverse effects of social media use for adolescents have received substantial attention. In response, a growing number of state‐level social media regulations are emerging in the United States. These policy interventions are being implemented in the context of mixed scientific evidence, forcing policymakers to weigh the need for proactive regulation against the limitations of extant research. We explore policymakers’ publicly stated rationales for social media regulations and contextualize their claims within extant scientific literature.

Methods

We conducted a media content analysis of elected government officials’ statements about 69 state social media legislative initiatives that were adopted or enacted prior to September 2024 using Google News. Subsequently, we critically reviewed the strength of the evidence underlying common themes.

Findings

We screened 637 documents, included 161, and extracted three main themes connected to claims about social media‐related adolescent harms: (1) harm to adolescents’ mental health (mentioned in 55 articles), including six subthemes (e.g., social media addiction, self‐harm or suicide, anxiety and depression); (2) exposure to dangerous online content (73 mentions), including five subthemes (e.g., access to pornography, risks of sexual exploitation); and (3) harm to adolescent development (38 mentions), including three subthemes (i.e., negative impacts on learning, social relationships, and brain development). We identified some evidence to support associations between social media use and adverse outcomes, particularly for vulnerable youth, but, overall, the current research base has significant limitations and cannot definitively establish causal effects.

Conclusions

State policymakers have moved rapidly to regulate adolescent social media use, often citing concerns about mental health, harmful content, and developmental impacts. The evidence to substantiate these assertions remains preliminary, but some promising directions are emerging around targeted protections for highly susceptible youth. We formulate a research agenda to inform evidence‐based policy.

The role of social media use in youth health and development has generated substantial scholarly and public attention. Use of social media platforms has increased significantly in the past decade, with one in five youth reporting use of social media and other digital technology platforms almost constantly. ¹ Concurrently to the rise in social media use, depressive symptoms, psychological distress, and other indicators of compromised mental health have markedly increased. ² , ³ Given the ecological connection between these two trends, an intuitive hypothesis has emerged that social media use may causally impact deteriorating youths’ mental health and development. While evidence around the potential impacts of social media continues to accumulate, federal and state public health officials and policymakers have emphasized the need for proactive social media regulation to protect youth. ⁴ Given that definitive evidence may take years if not decades to emerge, and because the absence of regulation may result in avoidable harm to the public, these policy interventions have to be designed in the context of imperfect information. ⁵ Despite gaps in evidence, emerging regulation should be informed by the best currently available evidence on the connection between social media use and adolescent mental health. It is, therefore, critical to evaluate the strength of evidence used in ongoing debates on policy interventions and to formulate research agendas that fill key evidence gaps to inform future regulatory efforts.

In evaluating the evidence base on the effects of social media use on adolescent health and development, two distinct areas of research need to be considered. First, evidence regarding the effects that social media use has had, and is having, on adolescent outcomes—the “realized effects” of social media use—should be considered. ⁶ Second, evidence regarding the effects that intervening on social media use would have on adolescent outcomes—the “intervention effects” of regulating social media—should also be considered. ⁷ To put the difference between realized and intervention effects into concrete terms, consider the following two questions. A causal question about a realized effect would ask what the mental health of adolescents would have been today if social media had never become widely adopted (compared with what it actually is given that social media has, in fact, become widely adopted). In contrast, a corresponding causal question about an intervention effect could ask what the mental health of adolescents would be a year from now if, after social media use has become widely adopted, an intervention to block or reduce adolescent social media use was implemented (compared with what adolescent mental health would be without such an intervention). Quite plausibly the effects on adolescent mental health that are implicit in these two questions would be very different even though the comparison between two alternative scenarios, adolescents do and do not use social media, is similar. In other words, if social media use has effects on important adolescent outcomes, it is important to consider the potential effects of policies to restrict social media use as a separate matter. This distinction poses an especially pertinent question when considering the current landscape of social media policy: to what degree is research on the potential realized causal effects of social media use being used to justify interventions to reduce or regulate social media use?

We explore the evidence on realized and intervention effects of social media use and the associated policy implications in five parts. First, we provide an overview of current US state‐level social media regulatory policy and provide a framework to organize regulation around the entities that are affected (i.e., platforms versus users of platforms) and whether legislative initiatives directly or indirectly impose regulations on social media use. Second, we conduct a media content analysis of how elected officials discuss the evidence around social media use and youths’ mental health, development, and learning, extracting common themes. Third, we provide an overview of the strength of the evidence underlying common realized causal effect themes that emerge from the content analysis. Fourth, we present the strength of the evidence for the intervention causal effects of policy interventions on social media and other digital technology use among adolescents. Fifth, we provide recommendations for advancing the understanding of social media impacts on youth and outline a research agenda for evaluating policy interventions.

What Are the Current and Proposed Policies?

State Social Media Regulation: an Evolving Landscape

Adolescent social media regulation has become a salient legislative priority in many US states. Beginning in 2023, the number of related state legislative initiatives has grown markedly. At the time of this writing, the National Conference of State Legislatures (NCSL) has tracked more than 50 adopted or enacted bills related to adolescent social media regulation in more than half of all US states, with well over 100 additional pieces of regulation under deliberation. ^{8 , 9} Additionally, an excess of 200 social media‐related bills were introduced but failed to pass state legislatures. ⁸ , ⁹

Overwhelmingly, the introduced, pending, or enacted pieces of legislation included elements designed to reduce adolescents’ social media use or to shape adolescents’ experiences on social media by imposing either restrictions or mandates on social media platforms, adolescent users, parents, school administrators, or other stakeholders. However, there is substantial heterogeneity in the type and scope of proposed regulations and, consequently, in the magnitude of their potential positive and unintended adverse effects. Generally, the regulations fall into one of four categories. These categories are distinguished by the degree to which the regulation directly or indirectly targets adolescent social media use and by whether the regulation imposes restrictions or mandates primarily on social media platforms or other technology corporations versus on adolescent users and other noncorporate stakeholders. Figure 1 depicts a heuristic schema for categorizing regulatory proposals along these axes. In the following sections, we discuss examples of regulation in each category.

Figure 1.

Schema for Categorizing State Social Media Regulation for Adolescent Protection

Direct Social Media Regulations Targeting Social Media Platforms

A wide range of direct restrictions or mandates for social media platforms have been proposed. Many of these proposals aim to alter core aspects of adolescent users’ experiences on social media in ways that conflict with the attention economy‐oriented business models of major social media corporations. ¹⁰ For example, legislative proposals have included provisions to prohibit social media platforms from algorithmically suggesting content to minors (e.g., California S287, failed), ¹¹ ban targeted advertisement to minors (e.g., Georgia S351, enacted), ¹² mandate enhanced privacy protection for minors on social media (e.g., Arizona H2858, pending), ¹³ or mandate content moderation strategies to protect minors from harmful content (e.g., Florida S454, failed). ¹⁴ Other proposals aim to reduce the access of minors to social media. Enhanced requirements for social media platforms’ age verification process, including specific stipulations of allowable age verification mechanisms and providers, represent a common example (e.g., Arkansas S396, enacted). ¹⁵ Notably, policies regulating social media platforms’ age verification process are frequently combined with restrictions imposed on users, such as social media minimum age requirements or parental consent requirements, which are discussed in the next section.

Direct Social Media Regulations Targeting Users and Noncorporate Stakeholders

Restrictions or mandates directly affecting adolescent social media users and other noncorporate stakeholders, including parents and schools, have also been proposed. These regulations primarily seek to reduce adolescent access to social media. Examples include state‐level social media age restrictions (e.g., ban of accounts for users under age 14; Florida H3, enacted), ¹⁶ parental consent requirements to create social media accounts for minors (e.g., Arkansas S396, enacted), ¹⁵ or nighttime social media curfews for minors (e.g., between 10:30 PM and 6:30 AM; Alaska H271, failed). ¹⁷ School‐focused regulatory proposals include prohibiting access to social media platforms on school‐owned devices or internet (e.g., Mississippi H925, failed) ¹⁸ or requirements to address social media safety in kindergarten through 12th grade (K‐12) curricula (e.g., Georgia S351, enacted). ¹⁹ Although these regulations directly impose restrictions or mandates on adolescent users and other noncorporate stakeholders, parallel regulations mandating enforcement by social media platforms (see previous section) are often required.

Indirect Social Media Regulations Targeting Social Media Platforms and Other Technology Corporations

Some state regulatory proposals also seek to shape adolescent social media access and experiences through broader restrictions or mandates imposed on both social media and nonsocial media technology platforms. Examples of these proposals include broad data privacy protections for minors on the internet, inclusive of social media (e.g., California A2529, pending), ²⁰ requirements to implement age verification for all websites hosting content designated as harmful to minors (e.g., websites hosting sexually explicit content, including some social media platforms; Alabama H164, enacted), ²¹ or the proposal to mandate smart device manufacturers to build adult content filters into their software (e.g., Indiana S201, failed). ²²

Indirect Social Media Regulations Targeting Users and Noncorporate Stakeholders

Regulations falling into the last category are neither specific to social media nor do they impose restrictions or mandates on social media platforms. Instead, the regulations pertain to adolescent users and noncorporate stakeholders. Cell phone bans in schools, such as the one passed in Indiana (S185), ²³ represent a prominent example. Other examples include requirements for K‐12 curricula to broadly address digital citizenship and media literacy (e.g., California A787, pending) ²⁴ or broadened definitions of negligence that can hold adults (including parents) liable for exposing minors to smart devices or other technology without adequate protections against cyberbullying and other abuse (Puerto Rico S1240, pending). ²⁵

These current and proposed state social media regulation initiatives have largely emerged in response to the thesis that social media use, if unregulated, causes harm to adolescents. ⁴ To evaluate the scientific evidence underlying this thesis, it is helpful to disaggregate the broad claim of “harm to adolescents” into narrower claims regarding specific domains of social media‐related harm that are cited by elected officials in support of current social media legislation. These specific claims regarding social media‐related harm can be more effectively examined relative to extant empirical literature. To this end, we sought to understand common themes in elected officials’ statements supporting current social media legislation.

Elected Officials’ Statements Supporting Current Social Media Legislation

A Content Analysis of State Elected Officials’ Statements

We examined the publicly stated rationales by elected officials for the adoption of social media restrictions and protections for minors as reported in the media. Specifically, we conducted a content analysis of elected government officials’ statements in reference to 56 pieces of state social media legislation, which were adopted or enacted prior to September 2024 and indexed by the NCSL. ⁸ , ⁹ In addition, we included 13 pieces of state regulation designed to ban or restrict the use of cell phones in schools, ²⁶ which were not indexed by the NCSL. For each legislative bill or regulation (total N  =  69), we identified news articles citing state elected officials’ statements on Google News, using the following search syntax:

“[state]” AND (“[bill number]” OR “[key term from bill name]”) AND (“governor” OR “senator” OR “representative” OR “speaker” OR “mayor” OR “assemblywoman” OR “assemblyman” OR “councilwoman” OR “councilman” OR “judge” OR “attorney general”) AND (“said” OR “told” OR “stated” OR “claimed” OR “explained” OR “assured”)

We screened the first ten news articles returned by Google News (as ranked by relevance) for inclusion in the content analysis. We selected the first ten to ensure our review included articles that are representative of the algorithmically prioritized search results that members of the public would be presented with for an analogous search and to maintain the total number of articles to be screened within a feasible range. Articles were included if they provided a direct quote regarding the law of interest from an elected government official. When news articles provided links to official government statements regarding the law of interest, these statements were also included in the analysis. Articles were excluded if they covered an unrelated topic, were inaccessible because of a paywall, did not provide a direct quote from an elected official, were not in reference to the law of interest, or were a duplicate of a previously included article. We used inductive coding techniques to categorize elected officials’ statements into themes and subthemes. The online supplement (Supplement 1–3) provides additional details about our methodology.

Results and Emerging Themes

A flow chart for the content analysis is depicted in Figure 2. We screened 637 documents, of which 161 were included in the analysis. Three main themes connected to claims about social media‐related adolescent harms emerged, including harm to adolescents’ mental health (mentioned in 55 articles), exposure to dangerous online content (73 mentions), and harm related to adolescent development (38 mentions). Table 1 provides exemplary quotes for each major theme and their subthemes.

Figure 2.

Flow Chart for Content Analysis of Government Officials’ Statements [Colour figure can be viewed at wileyonlinelibrary.com]

Table 1.

Themes and Exemplar Quotes From Government Officials’ Statements

Theme	Subtopic	Select Exemplar Quotes
Mental health	General mental health	“If you look at the increased rates of depression, anxiety, self‐harm since about 2012, across the board but especially with young women, we have just seen exponential increases in those mental health concerns. Again, the research is telling us over and over and over again that it is not just correlated, but it's being caused, at least in part, by the social media platforms.” (Governor, Utah)
		“Kids spend so much time on social media and using their smartphones that it's taking a toll on them mentally, emotionally and academically. Smartphone restrictions have proved successful in reversing these trends.” (State Senator, Pennsylvania)
	Anxiety/ depression	“Three or more hours of social media use per day doubles adolescents’ risk of poor mental health outcomes, such as depression and anxiety.” (State Senator, Colorado)
		“Studies show that students’ anxiety and stress related to social media are reduced when cell phone access during school hours is prohibited.” (Governor, South Carolina)
	Addiction	“The program will address the symptoms of youth mental illness as well as one of its largest causes: screen addiction and social media.” (Governor, Arkansas)
		“We're talking about products that are not only addictive, these products are deliberately designed to be addictive.” (State Representative, Florida)
	Bullying	“I am proud to have prioritized this legislation to help protect our children online and combat the very serious epidemic of cyberbullying which plagues this country.” (Lieutenant Governor, Georgia)
		“The feedback from school administrators, students, and parents that I have spoken with as I have toured the state hosting roundtables on this topic has been unanimous. You …reduce bullying…when you get phones out of the classroom.” (Lieutenant Governor, Ohio)
	Loneliness	“[Social media is a] product that is shown to increase depression, loneliness, and anxiety.” (Governor, Arkansas)
		“Youth depression, anxiety, and isolation have risen across the country.” (Governor, Arkansas)
	Self‐harm/ suicide	“I think we were just talking earlier, about the number of suicides…and the number of suicides and social media, and the correlation, is quite scary.” (State Representative, Tennessee)
		“These [social media] companies are killing our kids.” (Governor, Utah)
Harmful content	General harm	“[If social media] is the new town square, then God help us.…For our children, social media is no town square,…it is a dark alley.” (State Representative, Florida)
		“And it is a battle worth waging, as the harm social media is causing our children continues. Let's be clear: social media companies could voluntarily, at this very moment, do everything that the law put in place to protect our children. But they refuse to do so. Instead, they continue to prioritize their profits over our children's wellbeing.” (Governor, Utah)
	Pornography	“Exposing children to pornography is a form of child abuse in my opinion,…and is harmful to their mental health and developing healthy relationships. We need to apply the same standards in the cyber world that are in place in the physical world. No one under 18 can walk into an adult establishment or buy adult‐oriented materials. We need to broaden this to include the same standards for cyber opportunity.” (State Senator, Tennessee)
		“The availability of pornography has changed…We know the average kid sees porn at age 11, and that kids as young as five access it.” (State Representative, Louisiana)
	Exploitation	“As technology changes and online platforms provide opportunities for predators to disguise their identity, our approach has to evolve as well.” (State Representative, Kentucky)
		“[Social media is a] product that…puts our kids in human traffickers’ crosshairs.” (Governor, Arkansas)
	Drug/substance use	“The reality is that dealers don't need to lurk in dark street corners or remote parking lots. Now they are connecting with our kids on platforms like Snapchat, Instagram, Facebook and WhatsApp.” (State Assembly member, California)
		“This is not about an overdose…This is poisoning by people who are dealing these drugs on social media and murdering our children.” (Mayor, California)
	Misinformation	“Teaching media literacy is a key strategy to support our children, their families, and our society that are inundated with misinformation and disinformation on social media networks and digital platforms.” (State Assembly member, California)
		“I've seen the impact that misinformation has had in the real world—how it affects the way people vote, whether they accept the outcomes of elections, try to overthrow our democracy…This is about making sure our young people have the skills they need to navigate this landscape.” (State Assembly member, California)
Development	Learning/education	“Research also shows an increase in performance in the classroom when these devices are removed from the students. And the strong indication in this—and what is awesome—is that the performance of low achievers increases twice as much as high achievers when you remove a mobile device from a student's hand (while they're in class), so we're going to bring low achievers…up twice as much.” (State Representative, Florida)
		“It's important that students are focused on their lessons rather than ‘likes.’” (Lieutenant Governor, Ohio)
	Social relationships	“Being normal kids, like kids were prior to social media, is important. The social media (causes) more problems than it solves, and I think it causes more harm than good. So, let's have our education system be as much about traditional education as we can.” (Governor, Florida)
		“The other states that have done it the conversation is wow, the kids are talking to each other in the hall, they're having conversations in the cafeteria instead of each of them sitting there staring at the phone.” (State Senator, Louisiana)
	Brain	“We have a longstanding tradition in our country of drawing lines around ages for brain development when it comes to certain activities. We don't let kids smoke or drink or drive a car before certain ages, because we know the danger and the damage that is being done there, and the science will back that up.” (Governor, Utah)
		“Unlike an adult who can make an adult decision say I drank too much last night, I need to drink less or stop drinking altogether, a child in their brain development doesn't have the ability to know that they're being sucked in to these addictive technologies, and to see the harm, and step away from it. And because of that, we have to step in for them.” (State House Speaker, Florida)

The six mental health‐related subthemes included unspecified general mental health harm (n  =  31 articles), the dangers of social media or smartphone addiction (n  =  23), self‐harm or suicide in connection to social media use (n  =  10), increases in adolescent anxiety or depression (n  =  9), bullying connected to social media (n  =  5), and the impact of social media on adolescent loneliness (n  =  4). The five subthemes related to exposure to dangerous online content included unspecified general harm for adolescents as a consequence of social media use (n  =  45), adolescent access to pornography (n  =  18), the risk for adolescents to become subject to sexual exploitation on social media (n  =  12), social media as a venue for the sale of illicit substances (n  =  2), and exposure to misinformation (n  =  2). The three subthemes related to adolescent development included social media and smartphone use as obstacles to education and learning (n  =  31), negative impacts on adolescent social relationships (n  =  16), and adolescent brain development (n  =  2).

Beyond the three primary themes, additional themes related to motivations for supporting social media legislation emerged from the elected officials’ statements, but not all contained claims for which verification against extant scientific evidence is necessary. For example, 13 articles referenced the precautionary principle that views emerging threats to adolescent health and well‐being as sufficiently dire to act, even in the absence of definitive evidence about the role of social media. A Florida state representative explained, “I don't know if this is going to be the fix‐all…Maybe this may not be the bill and we've got to come back next year and the year after, but at least we are acting. At least we are moving the needle.”

Another prominent theme was the desire to improve data safety and privacy protections, especially for minors (n  =  23 articles). For example, a California state assembly member hoped the new legislation would “help students to be more responsible digital citizens, more intentional about what they put online, and better understand online safety and privacy.”

In 12 news articles, elected officials referenced empowerment of individual actors to take independent (rather than state‐level) action in protecting minors from the harms of social media, for example, in the form of locally flexible enforcement of rules (“Every school district's different, so this will be extremely flexible”; state representative, Ohio) or in the form of parents’ rights to sue (“It's going to be used against a corporation that has disseminated this material, so a single solitary single citizen going toe‐to‐toe with a corporation…This is the parents that are put in the driver's seat to protect their children, it's not meant to be the state”; state senator, Idaho).

The last emerging theme pertained to parental rights and responsibilities (n  =  28 articles). Some elected officials expressed expanded parental control over their adolescent children's social media use as a goal of the legislation in question. For example, the state senate president from Utah stated, “We in Utah are leading the fight against social media's harmful impacts on the unsupervised industry while providing parents with more resources and tools.” Other elected officials expressed a preference for parents, rather than the government, to take on the responsibility of protecting adolescents from social media‐related harms. For example, the Colorado Governor said, “The government can't parent kids. It's really up to the responsibility of parents to step up…Many parents don't understand the full threat of different social media for their kids. And I think more parents need to step up and take on that responsibility.”

Central themes emerging from the content analysis include claims that social media damages youths’ mental health and is addictive, that youth are exposed to harmful content on social media, and that social media negatively affects education and learning. We sought to contextualize these claims within extant scientific literature. In the next sections, we review the evidence related to potentially realized causal effects of adolescent social media use in these domains. Subsequently, we review the evidence for potential intervention causal effects of policy interventions on social media and other digital technology use among adolescents.

Realized Causal Effects: Does the Evidence Support That Social Media Causes Youth Harm?

Does Social Media Harm Youth Mental Health?

There is relative consensus across reviews and meta‐analyses that the correlation between time spent on social media and common youths’ mental health symptoms, such as depressive symptoms and psychological distress, ranges from approximately 0.1 to 0.2, ²⁷ , ²⁸ , ²⁹ , ³⁰ consistent with a statistically significant but relatively small effect size. The magnitudes of association may be higher for some groups, in particular girls, ²⁸ , ³¹ and potentially for some platforms. ³² However, the utility of this effect size for conclusions about the role of social media in youths’ mental health remains controversial. ³³ In terms of design, most studies on social media and youths’ mental health are cross‐sectional, which limits the ability to assess temporality and bidirectional associations. Furthermore, time spent on social media is nonlinearly related to adolescent mental health, ³⁴ and the meta‐analyses of time spent uniformly include broad time periods over which platforms and activities have shifted. ²⁷ , ²⁹ , ³⁰ , ³⁴ In terms of measurement, consensus is now building that “time spent” on social media is limited in explanatory variance. Variation in user experience and effects occur across type of device, specific applications, and specific features within application, and the quality and content of interactions within social media are heterogeneous axes for potential benefits and harm. ³⁵ Substantial research confirms that experiences on social media are context and person specific, ²⁸ , ³⁶ , ³⁷ yet the research areas within each of these subdomains are more limited. Compounding issues of exposure definition, meta‐analyses of “time spent” often include a range of “screen time” experiences (e.g., video game use, television watching, all mobile phone use). ³⁸ , ³⁹ However, even among studies specific to social media, little evidence of strong effects emerges. ⁴⁰ An important line of research that would theoretically aid conclusions is experimental evidence wherein users are randomized to social media patterns, and the available summaries of experimental evidence indicate null effects of experimentally reducing social media use on mental health. ⁴¹ However, these studies too are limited, often the timespan of the experiments are short, and mental health outcomes are heterogeneous. Mechanisms that link social media use to negative adolescent mental health outcomes have been hypothesized, ⁴² but conclusive evidence remains elusive. Studies of brain development suggest areas of the brain are activated by specific kinds of social media use (e.g., “checking” or frequency of logging into social media), ⁴³ , ⁴⁴ , ⁴⁵ yet associations typically are only found for certain groups, sample sizes remain small, and the role of identified areas of the brain in the development of mental health problems are not firmly established. ⁴⁶

Greater attention to the direction of effects is critical. Youth with mental health challenges are those who are at the greatest risk of adverse online and social media experiences. ⁴⁷ , ⁴⁸ When examined longitudinally, youth with depressive symptoms and depression tend to increase their use and engage in less social activity on social media (e.g., more scrolling content without connecting with others). ⁴⁹ , ⁵⁰ Youth with attention‐deficit/hyperactivity disorder and other externalizing symptoms have substantially different patterns of social media and phone use. ⁵¹ Indeed, although some longitudinal studies report positive associations between social media use and subsequent depressive symptoms, ⁵² there are also plenty of null findings, ³⁸ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ and research is consistent in identifying negative online experiences and patterns of use among adolescents with mental health challenges. ⁴⁸ , ⁵⁶

Are Youths Developing Social Media Addiction?

Addiction‐like symptoms in relation to social media include inability to quit or cut down on use, psychological distress when prohibited from using, craving or urges to use social media, and negative impacts of time spent on social media. ⁵⁷ These symptoms are sometimes referred to as “problematic social media use,” “addiction‐like social media behaviors,” or other labels. Available reviews indicate relatively high prevalence in many countries, including the United States. ⁵⁸ , ⁵⁹ The concept of social media addiction, however, remains controversial. Some classical symptoms of dependence or addiction relevant for psychoactive substances have not been well characterized for social media, and a biological basis for addiction remains elusive. ⁶⁰ Other behaviors that do not involve psychoactive substance use have been proposed for addiction‐like models, including shopping or compulsive buying, food, sex, exercise, internet use, and gambling, and are also met with controversy. ⁶¹ , ⁶² , ⁶³ , ⁶⁴ To date, gambling disorder remains among the few nonsubstance addictions for which there are defined diagnostic criteria in major nosologies. ⁶⁵ , ⁶⁶ Although clinically characterizing a potential phenomenon of social media addiction remains an active area of research, the available evidence indicates that high levels of self‐reported addiction‐like symptoms are correlated with psychological distress, depression, anxiety, peer‐relationship problems, sleep disruption, and other negative consequences. ⁶⁷ , ⁶⁸ , ⁶⁹ These associations are demonstrated in both cross‐sectional and longitudinal data. ⁷⁰ Thus, the literature is consistent that adolescents who self‐report addiction‐like symptoms have significant reports of other mental health challenges. ⁷⁰ , ⁷¹ , ⁷² , ⁷³ , ⁷⁴ However, it remains unclear whether social media “addiction” is a unique pathology that constitutes a disorder, and, furthermore, the evidence base around addiction‐like symptoms and mental health has many of the same limitations as the literature on time spent on social media: the vast majority of studies included in reviews and meta‐regressions are cross‐sectional, thus the bidirectionality of the relationships remain infrequently characterized.

Are Youth Exposed to Harmful Content on Social Media?

Social media allows users to generate their own content, provides a platform for known peer‐to‐peer interaction as well as anonymous interaction, and offers often minimally moderated spaces for sharing of videos, images, and text. Although such spaces can be used to build community and offer support, these spaces also can be harmful to young people in myriad ways. Harms on social media range from self‐comparisons with others, which meta‐analyses indicate can negatively impact self‐esteem and body image, especially for girls. ⁷⁵ , ⁷⁶ , ⁷⁷ Experiences of online bullying and threats are associated with increased risk of psychological distress and psychiatric symptoms, ⁷⁸ , ⁷⁹ but national data on the prevalence of online bullying experiences indicate little change in prevalence of such experiences over time. ⁸⁰ Importantly, available evidence indicates that online experiences often mirror the wider environment of an adolescent: those who have psychiatric disorders, experience bullying, report body‐image distress, and have negative peer interactions offline are the most vulnerable to those experiences online. ⁴⁸ , ⁸¹ , ⁸² , ⁸³ As such, greater attention to identifying adolescents with risk factors for negative online experiences is critical for intervention. Available research indicates that social media is frequently and effectively used by for‐profit companies to advertise addictive products such as alcohol and nicotine, with increases in youths’ uptake of these substances. ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ Material related to self‐harm and suicide is of particular concern, ⁸⁹ , ⁹⁰ , ⁹¹ , ⁹² and young people who survive a serious suicide attempt frequently report that they used internet sources of suicide content prior to their attempts. ⁹³ , ⁹⁴ Further, violent and sexually explicit content is also widely available and often distressing to young people who encounter or even seek out such material. ⁹⁵ Child exploitation and predatory behavior is also of particular concern given that internet connections can often be anonymous and identities falsified. ⁹⁶ Although the prevalence of child exploitation on social media is not well established, case reports are concerning.

Does Social Media Negatively Affect Adolescent Development (e.g., Education and Learning)?

A growing corpus of studies has examined educational outcomes and cognitive development associated with social media use. ⁹⁷ Generally, available reviews find that social media use, when multitasked with studying or learning, is associated with lower academic performance but positive associations with academic achievement when social media is used for academic purposes. ²⁷ , ⁹⁸ However, selection factors preclude causal assessment. It remains a threat to validity that students who use social media for distraction may be overselected for lower academic achievement. There are currently no long‐term studies that examine the impact of different patterns of social media use on academic performance and achievement over time. Outside of school settings, time spent on social media, especially in the evening, is associated with less sleep, ⁶⁹ , ⁹⁹ , ¹⁰⁰ which could in turn impact academic performance. However, longitudinal assessments of social media use, sleep, and academic performance remain limited. A recent systematic review identified five longitudinal studies that assess the connections between social media use and sleep duration, three of which found null associations, and four longitudinal studies concerning social media use and self‐reported sleep problems, three of which found null associations. ⁶⁹ Other reviews and meta‐analyses have found correlations between smartphone “addiction‐like symptoms” with academic performance ¹⁰¹ and that students who report using smartphones while studying report lower levels of learning and academic achievement. Available evidence is mostly cross‐sectional and self‐reported, predominately comprised college‐attending students rather than children and adolescents, and the relevant exposure construct is important to consider given that other meta‐analytic evidence finds that educational programming delivered through smartphones and other digital increases learning.

In sum, the body of literature on associations between social media use and adverse outcomes indicates that evidence of harm is emerging for specific patterns of use, particularly for vulnerable youths, but the current research base has significant limitations that require sustained investment in improved data collection and analysis. However, within the context of the relatively equivocal evidence base, regulations to reduce youth access to smart phones have been enacted, providing opportunities to evaluate the potential impacts of policy interventions.

Intervention Effects: What Is the Evidence on the Effects of Policies to Restrict Social Media Use?

Much of the existing evidence related to the effects of social media use restrictions on adolescent mental health, well‐being, and development pertains to policy evaluation studies in two areas: (1) the effects of cell phone bans in schools and (2) the effect of state anticyberbullying legislation.

Evidence on the Effects of School Cell Phone Bans

Although several US states have already enacted regulations to ban or restrict cell phone use in schools, to our knowledge, no peer‐reviewed studies have examined the effects of school cell phone bans on adolescent outcomes in the US context. Internationally, a limited number of studies have evaluated the effects of school cell phone bans. A rapid review and meta‐analysis identified five European studies that evaluate associations with student's academic performance or social well‐being, finding no significant summary effect in the academic performance domain but a moderate and significant protective association of school cell phone bans with social well‐being. ¹⁰² Interestingly, the most pronounced effects were identified for improvements in externalizing behavioral domains (e.g., reduced bullying, student conflicts). ¹⁰² A recent cross‐sectional observational study in England, which was not included in the rapid review, found no association between school cell phone policies and adolescent mental well‐being. ¹⁰³ Additional nonpeer reviewed studies examining the effects of school cell phone bans have been identified but are at high risk of bias. ¹⁰⁴ , ¹⁰⁵ , ¹⁰⁶

In sum, evidence regarding the effects of school cell phone bans is only beginning to emerge, with randomized trials and studies from the United States largely missing from the current literature. Evidence from a limited number of European studies is mixed, with some convergence around small benefits for reducing externalizing behaviors such as bullying. The school cell phone bans that have been evaluated to date are also characterized by substantial heterogeneity in terms of implementation (e.g., complete cell phone bans versus cell phone bans only for noninstructional use) and with respect to the timing of outcome ascertainment. Implementation differences may meaningfully modify the effects of such policies and need to be better understood.

Evidence on the Effects of State Anticyberbullying Legislation

All US states have passed antibullying laws, with 48 states explicitly sanctioning cyberbullying. ¹⁰⁷ Several studies have examined the effects of anticyberbullying legislation on adolescent cyberbullying victimization. The findings from this research are mixed, finding either a protective effect on reduced cyberbullying ¹⁰⁸ or no effect of cyberbullying‐specific laws above and beyond the effect of regular antibullying laws. ¹⁰⁹ Additional research, which found higher reports of cyberbullying victimization following the introduction of anticyberbullying laws, attributed the change to reporting effects rather than increases in cyberbullying incidence. ¹¹⁰ Additional research is needed before definitive conclusions about the effects of anticyberbullying legislation in the United States can be drawn.

Indirect Evidence on Plausible Effects of State Social Media Regulation

Social media regulatory proposals include mandated implementation of specific programming at the state level, such as school‐based media literacy training and targeted restrictions, such as nighttime social media curfews for minors. No direct evidence from policy evaluation studies to support these proposals is yet available, but indirect evidence suggests these interventions may be linked to some improved adolescent outcomes. Meta‐analyses of school‐based media literacy programs have found benefits for improved adolescent body image. ¹¹¹ , ¹¹² A systematic review of media literacy intervention evaluations also found evidence for effects on reduced adolescent alcohol use. ¹¹³ Furthermore, interventions to prevent or reduce nighttime screen use have been tentatively linked to improved adolescent sleep and increased daytime vigilance. ¹¹⁴

Taken altogether, some evidence to support associations between social media use and adverse adolescent outcomes has been identified, particularly for vulnerable youth, but, overall, the current research base has significant limitations and cannot definitively establish realized or intervention causal effects. On this basis, we formulate research priorities for informing evidence‐based policy.

What Is an Appropriate Research Agenda Moving Forward?

There are two primary research agendas that warrant expansion: one that assesses realized causal effects and a second that assesses intervention effects.

Research improvements are urgently needed to estimate realized causal effects (i.e., the causal effects of actual patterns of social media use on adolescent health and development). Literature on social media use and harm to adolescents has similar issues as many other epidemiological studies outside of the mental health domain: reliance on cross‐sectional data, observational study designs that are vulnerable to confounding and selection, and ambiguity in exposure definition. The current evidence base is largely not designed to test causal questions, leaving policymakers without clear scientific guidance for developing effective interventions and regulations that could meaningfully reduce potential harm while preserving the benefits of social media use among youths. We suggest two priority areas for research.

First, research is hampered because social media continues to shift, and these shifts have implications for the potential mechanisms through which engagement may impact adolescent health. Social media engagement can be one‐to‐one, one‐to‐many, many‐to‐one, can involve pictures and videos or text only, active or passive engagement, peer‐to‐peer or stranger‐to‐stranger, and may or may not involve posting personal information. Throughout the last decade, new platforms have emerged with heterogeneous characteristics and engagement modalities. Combining these various platforms into one category of “social media” is ineffective. Reviews and meta‐analyses that combine studies of youths from the 2000s and early 2010s, when Facebook was the modal social media technology, with youths in present day may be fundamentally combining very different exposures and experiences, which limits the ability to draw conclusions. Importantly, a central strength of existing reviews and meta‐analyses is that they are often preregistered with well‐defined targets—an open‐science approach that holds value in its own right as the research continues to accumulate. The strongest signals for impacts of digital technology on youths’ mental health come from studies that query youths about their experiences online (not just on social media but on many different internet platforms). Studies incorporating a broader range of social media behaviors, such as those that capture what adolescents do on social media via self‐report (e.g., posting pictures, liking, commenting), are an advance ⁵² , ¹¹⁵ as are studies that focus on particular platforms, experiences, and behaviors rather than broader measures of “social media.” Although questions remain around the utility of using an addiction frame for social media and digital technology use, available evidence indicates that psychiatric symptoms and distress are concentrated among youths who report addiction‐like symptoms or heavy use, emphasizing that continued development of exposure measures to accurately reflect use patterns is critical for advancing the field. Although data download packages, which allow users to download and donate to research the archived data that social media platforms collect about them (including social connections, search and feed content histories, inferred user characteristics, among other information), hold substantial promise for expanding evidence bases around social media effects, data privacy, consent, and other ethical (as well as analytic) challenges are abound. ¹¹⁶ , ¹¹⁷

Second, the available literature is clear that children and adolescents with psychiatric disorders, attention deficits, and those with psychological distress use digital technology and social media differently than those who are not experiencing such challenges. The direction and magnitude of the prospective association between adolescent psychiatric symptoms and subsequent changes in social media use is meaningful and a critical area for targeted intervention and prevention. Available evidence indicates that adolescents with the greatest psychiatric burden report the most serious distress associated with social media use, including addiction‐like symptoms, and are most likely to be exposed to dangerous and harmful content online. Young people in suicidal crisis use internet sources to both gain information on lethal means as well as to access sources of support to reduce suicidal crisis, thus the role of internet sources and social media as both a significant risk and significant site for intervention are clear. If a central goal of policy is in reducing psychiatric distress, then how children and adolescents already in distress use digital technology is a critical area of study.

In addition to improved research on realized causal effects, the research landscape experimentally manipulating social media and enacting policies to reduce social media use (i.e., intervention effects) is going to rapidly expand as policies are continuously enacted. Currently, the evidence base around causal effects of intervening to reduce social media use on adolescent health and well‐being remains equivocal. ⁴⁰ Existing experimental studies often do not ask the relevant question of what would happen if children and adolescents were not exposed to social media. Such a question is complex ¹¹⁸ : in the absence of social media, are children and adolescents engaging in a different form of digital technology use, engaging with other children and adolescents in nonscreen activities, or something else? This question is critical to estimating intervention effects, yet there is a limited evidence base in answering these questions. Quasi‐experimental studies including instrumental variable (IV) analyses that can be effectively leveraged for answering questions about intervention effects and identifying and developing appropriate IVs for various metrics of social media use are important areas for future work.

Given that policies around youth engagement with social media are rapidly accumulating, it is an important time to design studies that prospectively evaluate impact. At a most basic level, baseline data need to be collected before policies are enacted. Further, legal epidemiology is needed to evaluate the content of policies, variation, and heterogeneity in content, implementation, and enforcement. State and local jurisdictions often implement seemingly similar policies with vastly different regulatory frameworks, enforcement mechanisms, and timing of implementation, creating heterogeneous “treatment” conditions that are difficult to compare across jurisdictions or aggregate across geographic locations. Policy evaluation is methodologically challenging. Positivity violations in the presence of time‐varying confounding often mean that traditional causal effects cannot be estimated for many policy changes. ¹¹⁹ , ¹²⁰ Advancing the evidence base on social media's effects on youths will require interdisciplinary collaboration between legal scholars, who can precisely characterize variations in emerging platform regulations and content moderation policies across jurisdictions, and epidemiologists who can design rigorous studies that account for these legal nuances to generate actionable causal evidence for policymakers.

Conclusion

State policymakers have moved rapidly to regulate adolescent social media use through various policy mechanisms that target both platforms and users, often citing concerns about mental health, harmful content exposure, and developmental impacts. Although some evidence supports associations between social media use and adverse outcomes, particularly for vulnerable youth, the current research base has important limitations in establishing causal effects. The evidence specifically evaluating policy interventions like school phone bans or anticyberbullying laws remains preliminary; however, some promising directions are emerging around targeted approaches for at‐risk youths. To advance the field, research priorities should be focused on better characterizing heterogeneous social media exposures and experiences, understanding bidirectional relationships between psychiatric symptoms and platform use, and rigorously evaluating policy impacts through interdisciplinary collaboration between legal scholars and epidemiologists. Given the rapid pace of both technological change and policy implementation, maintaining scientific rigor while providing timely evidence to inform regulation remains an ongoing challenge that requires careful balance between precautionary action and evidence‐based policymaking.

Funding/Support

This work was funded by R01‐DA048853 (Keyes).

Conflict of Interest Disclosures

The authors have no relevant financial or non‐financial interests to disclose.

Supporting information

Supplement 1: Detailed media content analysis methodology

Supplement 2: List of state legislation and regulation included in the review, with Google News search terms

Supplement 3: Themes and quotes from government officials' statements – expanded exemplars