Free Play and Mental Health: What We Know, What We Don’t

Introducing the Play Collaborative Review Doc

Source: Shutterstock

There has been much debate about the effects of social media and the “phone-based childhood” on mental health. Jon Haidt and Zach Rausch have published numerous Google Docs, After Babel posts, and academic papers addressing that question. Yet there has been far less discussion or debate about the other half of The Anxious Generation: the loss of the “play-based childhood.” Unlike the debates over phones and social media, this part of the book has faced essentially zero pushback. Nobody that we know of has doubted the importance of free play for healthy childhood development. But the absence of critique means that few people are scrutinizing the evidence base. Without debate, without researchers trying to disprove claims or poke holes in studies, we don’t learn where the evidence is strong and where more work is needed.

The Anxious Generation presented four new norms to foster healthier childhoods in the digital age. The fourth norm — greater independence, responsibility, and free play in the real world — is the one we focus on here, particularly its play component. How strong is the evidence that the reduction or loss of free play in childhood can increase depression or anxiety? That more free play would improve children’s mental health? That it might protect them from mental illness in future years? And looking ahead to public policy, how confident can we be that if schools and families give children more free play in the real world, children’s mental health will improve in the aggregate?

Jon created a preliminary Google Doc to aggregate relevant studies back in 2020, and YeJin has been working with him since 2023 to update it and get it ready to present to the public and the research community. Our hope is that many readers will add to it and critique it, and help sharpen the evidence base for one of the most important questions surrounding child development today. In the rest of this post, we introduce the collaborative review doc, Play and Mental Health: A Collaborative Review, and we preview six early takeaways from the literature. We highlight what we still don’t know (and the studies we most need next), and close with a few practical ways schools and families can make more room for real-world free play. We’ll also invite researchers to contribute studies and critiques so the review stays updated, accurate, and useful.

Introducing the Play Collaborative Review Doc

We’ve launched the Play Collaborative Review, a living document designed to bring together the best available research on free play and its impact on children’s mental health and socio-emotional development. This project has three main goals:

1. Aggregate evidence: Collect and organize the correlational and experimental studies that examine links between free play and mental health outcomes, so researchers and policymakers can quickly get up to speed through curated abstracts and excerpts.

2. Identify gaps: Highlight where the research is thin, particularly regarding the long-term effects of free play.

3. Guide future research: Clarify what kinds of studies (especially experiments) are most urgently needed to answer open scientific questions and inform policy.

We hope this review will serve as a practical resource for educators, policymakers, parents, and anyone concerned about the mental health of today’s young people.

The review includes both correlational and experimental studies examining the effects of play for children ages 3 to 18, examining a range of socio-emotional outcomes. We are especially interested in evidence of causal or long-lasting effects, but such studies are nearly nonexistent. Most existing research focuses on immediate changes during or just after play sessions. Given this scarcity, we’ve also included relevant experimental studies with nonhuman subjects when they provide clear insight into the mechanisms underlying play's impact on relevant socio-emotional outcomes.

Below is a preview of the document’s Table of Contents, which gives a sense of the areas of research currently included. (You can click on any section to jump directly to that part of the document.)

Clickable Table of Contents

INTRODUCTION
1. THEORETICAL ARTICLES RELATED TO THE IMPORTANCE OF PLAY FOR CHILDREN’S DEVELOPMENT
2. MAJOR REVIEW ARTICLES ON THE EFFECTS OF PLAY
3. THE DECLINE OF PLAY
4. CORRELATIONAL STUDIES RELATED TO PLAY AND MENTAL HEALTH OUTCOMES
5. EXPERIMENTAL STUDIES RELATED TO PLAY AND MENTAL HEALTH OUTCOMES
6. QUALITATIVE STUDIES RELATED TO PLAY AND MENTAL HEALTH OUTCOMES
7. THE EFFECTS OF PARENTAL OVERPROTECTION
8. DISCUSSION
APPENDICES

What Can We Learn From the Review So Far?

Although still preliminary, we offer six insights that emerge from the studies listed in the review document (and we welcome corrections or suggestions if we have overlooked important work):

1. Free play Is declining

Children in the U.S. today have far fewer opportunities for self-directed, unsupervised outdoor play — especially with peers. One study found that, compared to their own childhoods, 85% of U.S. mothers say their kids now spend significantly less time playing outside (Clements, 2004). Time-use data also support this shift: children’s discretionary time in free play has steadily shrunk over the past few decades (Hofferth & Sandberg, 2001). Several researchers also note a sharp historical decline in child-led social play (Gray, 2011). Similar declines are happening in the UK and Canada, though long-run time-series evidence is sparse (Baines & Blatchford, 2023; Beaulieu & Beno, 2024; Dodd et al., 2021; Firth & Powell, 2025; Kuzik et al., 2023). While there are promising cross-national initiatives to assess the state of free play (e.g., Global Matrix highlights substantial gaps in active-play measurement, or Policy Studies Institute’s cross national study of children’s freedom to get about and play in their local neighborhood unaccompanied by adults), more rigorous research is needed about the rest of the world (Lee et al., 2024; Shaw et al., 2015).

2. Correlational studies show that play is associated with beneficial mental health and social-emotional development outcomes.

The bulk of available research is correlational, and these studies consistently find that outdoor play and nature connection are associated with lower psychosomatic symptoms, especially in girls (Piccininni et al., 2018), and adventurous play (vs. cautious or overly supervised play) is associated with fewer internalizing problems and higher positive affect, particularly for children from lower-income families (Dodd et al., 2022). Gray (2020) also found that during the early pandemic lockdown, children who had more free time and autonomy for play and family contribution showed lower anxiety and higher happiness.

3. Longitudinal evidence suggests that play’s effects on mental health persist in the short to medium term.

Longitudinal data on play’s impact across different developmental stages remains limited. But the few studies that have been conducted are promising, suggesting that more time in free play during early childhood predicts better self-regulation two years later, even after controlling for earlier self-regulation (Colliver et al., 2021). Another study links time spent in physical activity and outdoor play at age 9 to improved peer relationships through adolescence, though the effects on other socio-emotional outcomes are more modest (Hilliard, 2021).

4. Experiments point to causal evidence of play’s effects on children’s emotional resilience, mental health, and cognitive function.

There is some experimental evidence that play, especially when it’s physically active, imaginative, and self-directed — may reduce depression risk, improve emotion regulation, and lay the groundwork for mental wellness. In neuroscience labs, one study found that rough-and-tumble play among rats boosted resilience to depression by enhancing synaptic plasticity in the medial prefrontal cortex, a part of the brain linked to mood and self-control (Burdorf et al., 2017). In real-world settings with children ages 2 to 5, similar benefits have been observed. When researchers redesigned preschool play yards to include more nature and opportunities for “risky” play, children showed lower levels of depressed affect and antisocial behavior, and teachers reported improvements in focus, creativity, and self-regulation (Brussoni et al., 2017). Even in pediatric hospitals, adding just a few sessions of unstructured play for children ages 4 to 7 led to clinically significant reductions in anxiety within days (Al-Yateem & Rossiter, 2017).

Other experiments highlight the unique power of outdoor play. A randomized trial in New Zealand found that children at schools with more physical freedom — fewer playground rules, more loose parts, and more opportunity for rough-and-tumble play — were happier, more socially engaged, and possibly more resilient to bullying (Farmer et al., 2017). And a 2025 study comparing indoor and outdoor physical activity found that outdoor play consistently improved attention, memory, and cognitive control, with measurable differences from as little as 45 minutes after play (Walters et al., 2025).

5. Overprotective (or “helicopter”) parenting may be part of the problem

Across dozens of studies, overprotective parenting — often driven by anxiety or excessive concern — has been linked to higher rates of subsequent anxiety, depression, social withdrawal, perfectionism, and even disordered eating in children and adolescents. Longitudinal data show that low autonomy-granting parenting in adolescence predicts worse psychosocial outcomes in adulthood, including depression, suicidality, and substance abuse (Raudino et al., 2013). Overprotective or "accommodating" parenting — especially for children with anxiety or OCD — tends to reinforce distress and impair emotional development, both in the child and the parent (Thompson-Hollands et al., 2014; Pontillo et al., 2020).

Meta-analyses confirm that the strongest parenting predictor of childhood anxiety is low autonomy-granting, not lack of warmth. When children are denied opportunities to take age-appropriate risks and develop independence, they may fail to build a sense of self-efficacy — a key buffer against anxiety (McLeod et al., 2007; Van der Bruggen et al., 2008).

The consequences extend beyond childhood. Studies on "helicopter parenting" show that college students who say that they were raised in overbearing family environments report lower self-efficacy, worse decision-making, and more alienation from peers (van Ingen et al., 2015; Luebbe et al., 2018; Odenweller et al., 2014). Some never launch at all: a study on “adult entitled dependence” describes grown children still relying on their parents for daily functioning—and parents struggling to set boundaries (Lebowitz et al., 2012).

In sum, across correlational, longitudinal, and experimental studies, there is growing evidence that free play supports emotional health and resilience. There is a small but growing set of experimental studies increasing our confidence in causal claims: it seems likely that giving kids more and better playtime throughout childhood may improve mental health in adolescence. The evidence is not yet strong enough to quantify how much the decline in play contributes to today’s anxiety levels, nor prescribe the optimal “dose” for adolescents, but the pattern points to free play as a promising, low-cost solution that is easy to implement — and is really fun for the kids.

What research is most needed

Despite promising findings, the evidence base has major gaps:

• Overreliance on correlational studies: Most studies are correlational, and the few experiments that exist mostly just measure very short term changes. Experiments that connect play in childhood with mental health or behavior in adolescence would be extremely valuable, but of course such experiments are difficult to conduct on human children (Pellegrini, 1995). Experiments with other mammals show clear benefits (Pellis & Pellis, 2010; Siviy, 2010).

• School-based experiments: We are hopeful that some forward-thinking states or school districts will implement the kinds of experiments described in Chapter 11 of The Anxious Generation, in which some schools provide all elementary and middle school students with longer and higher-quality recess, as well as expanded play opportunities before and after school, while other schools maintain current policies. We predict that students exposed to enhanced play opportunities in K–8 will show better mental health outcomes in high school.

• Limited longitudinal data: Longitudinal data on play’s impact across different developmental stages remains limited.

• Types of play: Few studies compare the effects of different forms of play (e.g., structured vs. unstructured, physical vs. imaginative, social vs. solitary).

• Digital play: The role of digital play, such as video games or online social interaction, remains contested. Some scholars argue that certain types of digital games foster creativity and connection. Others suggest that screen-based play crowds out physically active and socially rich forms of play. The effects may depend heavily on content, context, and usage patterns, but high-quality causal studies to date are scarce. We also note that since the mid 2010s, “video games” no longer refers solely to playing games with friends; platforms such as Roblox often involve interacting with strangers via text or voice.

• Mechanisms: Few studies explore mechanisms — why or how play improves psychological outcomes (e.g., through enhanced autonomy, social bonding, physical engagement).

• Variation across groups: There is little research on how play interventions affect children in high-risk, low-income, or culturally diverse contexts.

Conclusion

The Play Collaborative Review Doc is a living document designed to aggregate what we know about free play and mental health, surface what we don’t, and outline the studies that could streghten the evidence base on free play’s effects on mental health. If you’re running research or programs, please share missing studies, critiques, or results. (To contribute, request commenter access via the “Request Edit Access” button at the top of the document.) We will continue updating the review so readers can quickly see where evidence is strong, where it’s thin, and what should be tested next.

In the meantime: if you are a parent, teacher, school administrator, or state legislator, we encourage you to visit LetGrow.org for simple, low-cost, low-risk ways to give kids more free play and independence. The research we have reviewed suggests that doing so is likely to be beneficial. And if you’re interested in partnering on experiments that test the benefits of play for adults (e.g., in workplaces, universities, or community settings), please contact YeJin at yejin.park@stern.nyu.edu.