Growing up in the social media age

2. What is social media and its impacts?

Social media can be understood as the “forms of electronic communication (such as websites for social networking and microblogging) through which users create online communities to share information, ideas, personal messages and other content (such as videos)” (Merriam-Webster Dictionary, 2025[5]). However, academics have defined social media in various ways that have evolved over time (Aichner et al., 2021[6]). Today, social media is typically characterised by several key features. First, it is understood to enable the creation and exchange of user-generated content, typically linked with the user via a profile. Second, it is understood to be networked, meaning that it enables both one-to-one and one-to-many communication with other users. Third, it is often understood to enable interactivity, whereby users can interact with content and other users, including through liking or messaging (Kapoor et al., 2018[7]; Aichner et al., 2021[6]).

Social media use is driven by motivations that are both positive (e.g. strengthening social connections) and negative (e.g. escaping unpleasant offline experiences) (OECD, 2025[8]). For example, social media can reinforce existing relationships, as well as provide access to new people and communities, including those with shared identities and interests (Odgers and Jensen, 2020[9]; Berger et al., 2022[10]). Across the OECD, 40% of 11-15 year-olds have friends they met online and communicate with at least once a week (Health Behaviour in School-aged Children Study, 2023[11]; OECD, 2025[8]). It also facilitates access to information, with many people sourcing their news from social media platforms (OECD, 2024[12]; Pew Research Center, 2025[13]). In addition, some researchers have found social media platforms and online communication help reduce social isolation and cope with anxiety, including for elderly and young people (Cauberghe et al., 2021[14]; James et al., 2023[15]; Ibsen et al., 2025[16]).

At the same time, social media use brings risks, particularly for young people who may be more susceptible to harms from negative peer feedback loops, unrealistic body images and harmful content online. Girls are especially at risk from the negative effects of social media use (OECD, 2024[17]; Molly Rose Foundation, 2025[18]; Anses, 2026[19]). Some researchers also argue that intensive social media engagement may be used to withdraw from offline social interactions, potentially increasing negative feelings like loneliness (Noland, Necka and Cacioppo, 2018[20]; Roberts, Young and David, 2024[21]; OECD, 2025[8]).

Social media clearly presents a multitude of opportunities and risks for young people (OECD, 2024[17]; OECD, 2025[22]; OECD, 2025[8]). This section presents a brief review of the literature on the impacts of social media and well-being outcomes for young people.

The research on social media use and well-being outcomes is complex and nuanced

Researchers typically examine large, nationally representative datasets, including self-reported data on both social media use and outcomes such as mental health or assessments of academic achievement. Most analyses of such data find that excessive social media use is correlated with negative outcomes.

However, researchers disagree on the size of the correlations (Twenge et al., 2020[23]; Orben and Przybylski, 2020[24]; Vuorre, Orben and Przybylsi, 2021[25]). Correlations do not prove causality. For example, lower-performing students or those already depressed might be more likely to use social media rather than social media use being the cause of lower test scores or unhappiness. Some researchers suggest that social media use is less likely to affect the mental health of young people relative to other economic factors, such as income level or parental unemployment (Odgers, 2024[26]). Furthermore, other circumstances, such as parental engagement, homelessness or hunger, may also play a role. OECD work indicates that both contextual and protective factors play a role in understanding the impacts (OECD, 2025[8]).

Other disagreements about correlations further heighten the need for nuance. The relationship between social media and well-being outcomes may be subject to omitted variable bias, where a third variable is correlated with both variables of interest. Researchers also disagree on the extent to which other factors may explain potential correlations. Some see such factors having a direct causal relationship on young people’s outcomes, while others view them as a pathway to negatively affect such outcomes (e.g. closeness with parents, negative attitudes towards school, cyberbullying, disrupted sleep, limited time with friends offline, etc.) (DoE, 2019[27]; Orben and Przybylski, 2019[28]; Twenge et al., 2022[29]). Systematic reviews using rigorous statistical techniques and that account for covariates find a mixture of no, small or mixed correlative associations (Ivie et al., 2020[30]; Odgers and Jensen, 2020[9]; Orben, 2020[31]; Hancock et al., 2022[32]; Valkenburg, Meier and Beyens, 2022[33]). Moreover, the associations should be further nuanced as moderating factors, such as age, sex and pre-existing vulnerabilities, may influence the strength and direction of the relationship between social media use and well-being outcomes (OECD, 2025[8]).

At the same time, several longitudinal studies show that changes in digital technology or social media use are associated with changes in mental health and academic outcomes (Babic et al., 2017[34]; Boers, Afzali and Newton, 2019[35]; Shakya and Christakis, 2017[36]), while others do not (Heffer, Good and Willoughby, 2019[37]; Orben, Dienlin and Pszybylski, 2019[38]; Mitev et al., 2021[39]). Some experimental research,1 including natural experiments and controlled trials, finds positive correlations between participant outcomes and social media use with varying experiment specifications (Hunt et al., 2018[40]; Allcott et al., 2020[41]; Fioravanti, Prostamo and Casale, 2020[42]; Mosquera et al., 2020[43]; Lambert et al., 2022[44]). Others find negative correlations (Hall et al., 2021[45]; Collins and Eggers, 2022[46]; Mahalingham, Howell and Clarke, 2023[47]).

Additional methodological issues complicate the analysis of digital technology use or social media use and young people. First, most cross-sectional studies rely on self-reported data of time spent online. However, a body of research shows that self-reported data on digital technology use may be highly inaccurate (Sewall et al., 2020[48]), with subjects routinely overestimating their use of digital technologies. Such data also provide aggregate relationships that may mask effects based on age-specific, cultural, family or other characteristics (so-called group-level effects). Overall, the academic research on social media use and well-being outcomes is complex and nuanced.

Social media use is ubiquitous among young people, although girls are more intensive users

Mobile phones are as ubiquitous in the lives of young people as they are for adults. Young people need a digital device to access social media, and this primarily takes place through mobile phones. At 15 years of age, 97% of students have their own smartphone on average across the OECD (OECD, 2025[49]). However, most receive their first mobile phone between the ages of 10 and 12, with some getting one before turning 5 (McClain et al., 2025[50]; Hungarian National Media and Infocommunications Authority, 2026[51]). The first acquisition of a digital device is mainly motivated by educational and entertainment uses (e.g. video games). The creation of social media accounts comes later, driven by the desire to interact with family members and peers (Arfaoui and Elbaz, 2025[52]).

The use of devices for educational purposes persists as children grow older. According to data from PISA for 2022 for 15-year-old students across the OECD (Box 1), devices are increasingly used for educational purposes, particularly since the COVID-19 pandemic. Most young people use them for learning activities on average: regularly in the classroom (two hours), before and after school (1.5 hours) and on the weekend (1.6 hours) (OECD, 2025[8]). Research suggests that some digital technologies can enhance learning outcomes, but their impact depends highly on context. If poorly designed or implemented, digital technologies risk becoming more distracting than beneficial (UNESCO, 2023[53]). Moderate use of digital technologies for learning is also shown to be positively associated with academic performance (OECD, 2024[54]): 15-year-olds who use digital tools for learning for up to three hours a day tend to outperform peers who do not, particularly in mathematics.

The use of devices for entertainment purposes likewise continues for children as they age. Limited use of digital technologies for leisure activities (on average under two hours) is associated with improved feelings of belonging and academic outcomes (OECD, 2024[54]). However, as daily hours of leisure spent in digital environments increase, academic scores decline more rapidly than they do for those who spend an equivalent number of daily hours using digital technologies for learning.

Furthermore, not all leisure activities carried out on digital technologies have the same relationship with students’ academic and non-academic outcomes. Students who spend over seven hours per day communicating on social media, for example, score 15 points lower in mathematics than those spending a similar amount of time playing video games. This is the case even after controlling for socio-economic profiles of students and schools (OECD, 2024[54]).

Box 1. The OECD Programme for International Student Assessment (PISA)

The PISA survey is a cyclical, cross-national and cross-sectional study analysing 15-year-old students (OECD, 2023[55]). The Programme assesses educational competencies in reading, mathematics, science and other domains like creative thinking and financial literacy. The survey collects detailed information about students, including socio‑economic background and reported life satisfaction and the feeling of belonging.

First launched in 1997, with the first data collected in 2000, PISA assesses the academic competences of 15-year-old students across the world. Students also complete a background questionnaire to collect information about their attitudes, homes and schools, and activities. Optional questionnaires are sometimes distributed to students (e.g. additional modules focusing on topics like well-being or financial literacy) or to other people in students’ lives, including their parents, teachers and school administrators.

The 2022 assessment surveyed approximately 700 000 students in 81 countries and economies, representing about 29 million 15-year-olds. Some caution is required when interpreting estimates for Australia, Canada, Denmark, Ireland, Latvia, the Netherlands, New Zealand, the United Kingdom and the United States. In these countries, one or more PISA sampling standards were not met (OECD, 2023[56]).

The optional “ICT questionnaire”, in which 53 countries and economies participated, surveyed students on their use of digital devices. This questionnaire included questions on the availability and use of digital resources, as well as time spent in specific leisure activities on the weekend and weekdays. These included activities such as playing video games, browsing social media networks, and communicating and sharing digital content on social media networks.

In addition, 15 countries distributed the “Well-being questionnaire” to students. This included questions about health, body image, number of friends, physical and mental health, and social connectedness (OECD, 2025[57]). Compared with the other questionnaires, the Well-being questionnaire has a smaller sample of participating students and covers only nine participating OECD Member countries. These data can be used to provide insights on the relationship between social media use and body image or social connectedness.

Source: OECD (2023[56]), PISA 2022 Results (Volume I): The State of Learning and Equity in Education, https://doi.org/10.1787/53f23881-en; OECD (2024[58]), PISA 2022 Technical Report, https://doi.org/10.1787/01820d6d-en; OECD (2025[49]), PISA 2022 Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html; OECD (2025[59]), PISA 2022 ICT Familiarity Questionnaire.

Young people use digital devices to access social media platforms often. On average across the OECD,2 about 95% of surveyed 15-year-old students report browsing social media on a typical day. Meanwhile, 88% report communicating and sharing digital content on social media, with relatively little difference on weekdays and weekends (OECD, 2025[49]). This may suggest a revealed preference for time spent on social media. They also report spending more time browsing social media, and for longer periods of time, than communicating and sharing digital content on social media or other online communication platforms (Figure 1).

The intensity of social media use varies across countries. A low share of respondents across the OECD (6%) report spending no time at all browsing social media networks on either the weekday or the weekend. However, more than 10% of surveyed 15-year-olds in Israel, Japan, Korea and Türkiye report spending no time at all on social media (OECD, 2025[49]).

Figure 1. Almost all young people use social media, but they spend more time browsing than communicating and sharing digital content

Percentage of 15-year-old students by time spent browsing social media networks, or communicating and sharing digital content on social media or any communication platform, on weekdays and weekends, OECD average, 2022

Note: Endnote.3

Source: Calculations based on OECD (2025[49]), Programme for International Student Assessment Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 24 June 2025).

Although communicating and sharing digital content on social media or other online communication platforms is also widespread, 16% of 15-year-olds do not spend any time carrying out this activity either on the weekday or on the weekend (Figure 1). More so, a third of 15-year-olds in Israel and around a quarter in Costa Rica, Czechia and Türkiye report neither communicating nor sharing content on social media on either weekdays or weekends. In contrast, approximately 12% of those surveyed report spending more than five hours communicating and sharing digital content on weekdays and 13% report spending more than five hours doing so on the weekend. In Chile and Slovakia, around one in five 15-year-olds spend more than five hours on this activity.

Taken together, 15-year-old students in OECD Member countries spent on average almost 35 hours per week browsing, communicating and sharing digital content via social media in 2022 (Figure 2). This cohort also reports different average rates of usage of social media across countries. For example, on average, 15-year-old students spend less than 26 hours per week on social media in Japan and more than 45 hours per week in Chile. By way of comparison, the average 15-year-old OECD student spends almost 24 hours per week learning in regular school lessons (OECD, 2023[55]) and 10.5 hours per week completing homework (OECD, 2025[49]).

Figure 2. The average 15-year-old in the OECD spends 35 hours a week using social media

Percentage of 15-year-old students by average time spent on social media per day (left axis), and overall average weekly hours of social media use, total and by sex (right axis), 2022

Note: See Endnote.4

Source: Calculations based on OECD (2025[49]), Programme for International Student Assessment Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 24 June 2025).

In all OECD Member countries surveyed, girls spend more of their leisure time in digital environments browsing social media than on other activities (Figure A A.1). Conversely, in most countries, playing video games is the top leisure activity in digital environments for boys. On average across the OECD, girls spend 22 hours per week browsing social media and 16 hours communicating and sharing on social media (Figure 3). In contrast, boys spend 18 hours browsing social media and 14 hours on social media communication.

Differences between boys and girls in playing video games are more striking. Boys dedicate 20 hours per week to video games while girls spend only 11 hours doing so (Figure A A.1). Boys and girls alike report spending another 17 hours browsing the Internet. They also spend around ten hours on each of the following: looking for practical information online, learning how to do a specific task or activity (e.g. reading and listening to tutorials and podcasts), and creating and editing online content.

Figure 3. Digital leisure activities differ by sex

Note: See Endnote.5

Source: Calculations based on OECD (2025[49]), Programme for International Student Assessment Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 24 June 2025).

Considering both browsing and communicating on social media, across the OECD, girls spend an average of 37.2 hours per week on social media – over 5 hours more than boys (31.8 hours) (Figure 2). In Ireland, girls spend ten more hours per week on social media than boys, and other studies have highlighted a higher rate of problematic social media use among girls (OECD, 2024[17]). Türkiye is the only surveyed country in the OECD in which boys spend more time on social media than girls. Several authors argue that the negative effects of social media may be sex specific, with girls at potentially higher risk of harms like poor body image and community-based online isolation and exclusion (Napp and Breda, 2022[60]; Svensson, Johnson and Olsson, 2022[61]; Haidt, 2024[62]; OECD, 2024[17]; Molly Rose Foundation, 2025[18]).

The average differences in social media use tend to be relatively small across other socio-demographic characteristics. For example, across OECD Member countries, young people with tertiary-educated parents spend 33.2 hours on social media per week. This is only slightly less time than their counterparts with parents who have no more than non-tertiary post-secondary education (36.5 hours) or secondary or lower education (36.3 hours). Similarly, on average across OECD Members, 15-year-old students with an immigrant background spend slightly more time (up to 2.1 hours) on social media than native-born students, although the overall differences are small (OECD, 2025[49]).

Social media interactions can elicit comparisons with others (e.g. peers and celebrities) and increase peer feedback (both positive and negative). Some researchers argue that social media may be used to withdraw from offline social interactions, potentially increasing negative feelings like loneliness. At the same time, 40% of 11-15 year-olds across the OECD have friends they met online and communicate with those friends at least once a week (Health Behaviour in School-aged Children Study, 2023[11]; OECD, 2025[8]).

Social connection is integral to social media. The PISA Well-being and ICT questionnaires are optional, and five countries (Costa Rica, Ireland, Hungary, Slovenia and Spain) implemented both questionnaires in 2022. The data from these countries shed light on the complex interplay between social media usage and social connectedness. In general, students who are well connected with their peers spend, on average, more time on social media. Those who report having ten close friends or more dedicate one extra hour per day to scrolling and communicating compared to those who say they have two or fewer close friends.

Not surprisingly, students who rarely or never talk to their friends on their phone, send text messages or have contact through social media also report spending less time on social media (around 25 hours per week) than those who communicate several times a day, and who report spending over 38 hours (on average) on social media per week. Interestingly, lower frequency of communication goes together with reduced time spent passively browsing social media.

The more days per week young people spend with their friends after school, the more hours they also spend on social media (Figure 4). Indeed, those who see their friends right after school one day per week or never spend on average around 30 hours per week on social media. Those who see them every day after school spend close to 40 hours per week on social media. This finding suggests that social media does not necessarily reduce in-person interactions but constitutes an intrinsic part of them. Indeed, young people often play video games with their friends simultaneously online, as well as physically located in the same room; they also browse the Internet together, among other online activities.

However, based on PISA 2022 data, social media use among 15-year-olds is higher for those who feel lonely at school compared to those who feel well integrated. Social media can be an escape from loneliness, but it can also reinforce this feeling and lead to even greater isolation at school. Moreover, these data suggest that higher social media usage is associated with lower life satisfaction and more negative body image, in line with existing research (Webster, Dunne and Hunter, 2020[63]).

Figure 4. The more young people see their friends right after school, the more time they spend on social media

Change in weekly hours spent on social media relative to 15-year-olds who usually never see their friends right after school, 2022

Note: See Endnote.6

Source: Calculations based on OECD (2025[49]), Programme for International Student Assessment Database, https://www.oecd.org/en/data/datasets/pisa-2022-database.html (accessed on 24 June 2025).

Notes

← 1.  Experimental designs typically compare statistically similar cohorts after varying exposure to the variable of interest (in this case, social media). Much of this research studies adults (usually of college age) because children are a difficult cohort to experiment with for both technical and ethical reasons. Studies ask a treatment group of experimental subjects to reduce their social media use and examine their outcomes relative to a control group. Another body of work asks participants to use a type of social media in a controlled setting, usually for a limited period, and measures outcomes relative to the control group.

← 2.  These data refer to respondents in the 29 OECD Member countries participating in the PISA ICT questionnaire.

← 3.  15-year-old students were asked, “During a typical weekend day [/weekday], how much time do you spend doing the following leisure activities?”, where two responses were: “Browse social networks (e.g. <Instagram®>, <Facebook®>)”; and “Communicate and share digital content on social networks or any communication platform (e.g. <Facebook®>, <Instagram®>, <Twitter®>, emails, chat)”. OECD average does not include Canada, Colombia, France, Luxembourg, Mexico, the Netherlands, New Zealand, Norway or Portugal.

← 4.  15-year-old students were asked, “During a typical weekend day [/weekday], how much time do you spend doing the following leisure activities?”, where two responses were: “Browse social networks (e.g. <Instagram®>, <Facebook®>)”; and “Communicate and share digital content on social networks or any communication platform (e.g. <Facebook®>, <Instagram®>, <Twitter®>, emails, chat)”. The daily and weekly averages were calculated assuming five weekdays, two weekend days and seven days in a week. The time spent on social media is calculated as a sum of the two variables using range midpoints and assuming eight hours for the upper limit (“More than 7 hours”). Although the items were measured through separate sub-questions, the activities could be performed simultaneously, and their aggregates should be interpreted with caution. OECD average does not include Canada, Colombia, France, Luxembourg, Mexico, the Netherlands, New Zealand, Norway or Portugal.

← 5.  15-year-old students were asked, “During a typical weekend day [/weekday], how much time do you spend doing the following leisure activities?”, where the seven responses were: “Play video-games (using my smartphone, a gaming console or an online platform or Apps)”, “Browse social networks (e.g. <Instagram®>, <Facebook®>)”; “Browse the Internet (excluding social networks) for fun (e.g. reading news, listening to podcasts and music or watching videos)”, “Look for practical information online (e.g. find a place, book a train ticket, buy a product)”, “Communicate and share digital content on social networks or any communication platform (e.g. <Facebook®>, <Instagram®>, <Twitter®>, emails, chat)”, “Read, listen to or view informational materials to learn how to do something (e.g. tutorial, podcast)”, and “Create or edit my own digital content (pictures, videos, music, computer programs)”. The weekly average was calculated assuming five weekdays and two weekend days. OECD average does not include Canada, Colombia, France, Luxembourg, Mexico, the Netherlands, New Zealand, Norway or Portugal. Although the items were measured through separate sub-questions, the activities could be performed simultaneously, and their aggregates should be interpreted with caution.

← 6.  15-year-old students were asked, “During a typical weekend day [/weekday], how much time do you spend doing the following leisure activities?”, where two responses were: “Browse social networks (e.g. <Instagram®>, <Facebook®>)”; and “Communicate and share digital content on social networks or any communication platform (e.g. <Facebook®>, <Instagram®>, <Twitter®>, emails, chat)”. The weekly average was calculated assuming five weekdays, two weekend days and seven days in a week. The time spent on social media is calculated as a sum of the two variables using range midpoints and assuming eight hours for the upper limit (“More than 7 hours”). Although the items were measured through separate sub-questions, the activities could be performed simultaneously, and their aggregates should be interpreted with caution. OECD average does not include Canada, Colombia, France, Luxembourg, Mexico, the Netherlands, New Zealand, Norway or Portugal.