Raising Canada 2026: Does the Evidence Really Establish Online Harms as Canada’s #1 Threat to Children?
Caveat: This week, we were contacted by several news outlets asking for our thoughts on the newly released Raising Canada 2026 report. Given that we had not yet had an opportunity to read and review the report in full, we did not believe it would be responsible for us to comment. We have now carefully reviewed the report, thus the reason for this article.
First, we want to acknowledge the importance of the work, the Raising Canada 2026 report was produced by Children First Canada in collaboration with PolicyWise for Children & Families, The Hospital for Sick Children, and the University of Calgary, drawing upon evidence and perspectives from researchers, clinicians, governments, community organizations, and young people across Canada. The report identifies serious challenges affecting Canadian children and calls for greater attention, investment, prevention, and accountability. We agree with much of that objective (1). We also agree that the ten categories identified in the report represent important threats and challenges facing Canadian children and youth. Many are issues that we at The White Hatter have spoken about extensively over the years.
Where we have questions is the decision to identify “Online Harms” as the #1 threat facing children in Canada. Our concern is not whether online harms belong on the list, because they absolutely do. Our question is more fundamental, “Does the evidence actually allow these ten very different categories of childhood risk to be ranked from #1 through #10 in the first place?”
Children First Canada states that, for the first time since the Raising Canada report was launched in 2018, online harms have become the number one threat facing Canadian children. The organization explains that online harms are increasingly intersecting with or accelerating several other threats, including poor mental health, declining physical activity and play, bullying, racism and discrimination, violence and child abuse, and vaccine misinformation (1)(2). We acknowledge that interconnectedness deserves serious attention, and that digital environments can absolutely amplify existing risks.
However, being interconnected or cross cutting does not necessarily establish that something represents the greatest overall threat. Poverty is also cross cutting as it can influence nutrition, housing stability, educational opportunity, physical health, mental health, family stress, recreation, access to health care, and vulnerability to violence. Child maltreatment is similarly cross cutting as it can affect physical health, mental health, educational attainment, relationships, substance use, socioeconomic outcomes, and functioning well into adulthood. Climate change can affect respiratory health, infectious disease, food security, physical activity, displacement, extreme weather exposure, and mental health. If the ability of one threat to amplify other threats is part of the methodology used to determine the #1 ranking, that criterion should be clearly defined and consistently applied across all ten categories.
This leads to what we believe is the central methodological issue that we have with the ranking, that being the ten threats identified in Raising Canada 2026 are not necessarily comparable units of risk:
Unintentional injury is primarily a mechanism and outcome of physical harm.
Poverty is an upstream social determinant of health.
Mental illness is a health outcome.
Bullying is an interpersonal behaviour and form of victimization.
Physical inactivity is a behavioural health exposure.
Systemic racism and discrimination are social and structural exposures.
Vaccine-preventable illnesses are a category of disease.
Climate change is an environmental and systemic threat.
Online harms are an extraordinarily broad environmental umbrella encompassing behaviours, exposures, victimization, crimes, health concerns, and social experiences, which can create an apples to oranges comparison. It would be somewhat like attempting to rank smoking, cancer, poverty, car crashes, air pollution, and depression on a single list and declaring one as the definitive “#1 health threat.” All are important, but they operate at different points in the causal chain and are measured differently.
This is why we believe the most important question is not whether online harms should be #1, #5, or #8. The question for us is, “What empirical methodology allows these fundamentally different categories to be placed into a meaningful ordinal ranking at all?”
If the ranking is based primarily upon mortality, preventable injuries would likely move toward the top.
If it is based upon population prevalence, poverty, food insecurity, maltreatment, bullying, physical inactivity, or certain mental-health measures could move substantially.
If it is based upon long-term disability and morbidity, the hierarchy could change again.
If it is based upon children’s rights, another ordering might emerge.
If it is based upon future cumulative risk, climate change might move much higher.
If it is based upon preventability, economic burden, inequality, developmental consequences, or the ability of one risk to amplify other risks, still other rankings could emerge.
None of those approaches is inherently unreasonable. The problem arises when a report presents something as the “#1 threat facing children in Canada” without making sufficiently clear to the public what #1 actually represents and how the categories were compared and weighted.
Were mortality and hospitalization considered?
How was prevalence weighted against severity?
How were long-term developmental consequences measured?
How were children’s rights incorporated?
Was preventability considered?
Were population-level and individual-level risks treated differently?
How were overlapping categories handled?
Without a transparent and reproducible methodology, the numerical ordering can communicate a level of epidemiological precision that may be greater than the underlying evidence allows.
As an example, consider preventable injury. The Canadian Paediatric Society states that injury remains a leading cause of death and morbidity among Canadian children and youth aged 1 to 19. Its 2026 injury prevention framework reports that in 2018, 643 children and youth died from injuries, 23,305 were hospitalized, and more than 1.2 million were treated in emergency departments. The CPS also reports that serious injuries can have lasting consequences for development, quality of life, education, and families (3). More recent data reviewed by the CPS continue to identify major causes of unintentional injury death and hospitalization among Canadian children and adolescents, including drowning, motor-vehicle incidents, poisoning, falls, burns, and off-road vehicle incidents (3).
Deaths, hospital admissions, emergency-department visits, and disability are measurable outcomes. That does not automatically mean injury must be ranked #1. However, it demonstrates something more important, that being the answer depends upon what is being measured. If mortality and serious physical morbidity receive substantial weighting, injury becomes extremely important. If another framework gives greater weight to children’s rights, emerging risks, cross cutting effects, or rapid changes in exposure, a different category might rise. That is why we believe that methodology needs to be visible before a numerical hierarchy can be properly interpreted.
Child poverty and food insecurity are similarly difficult to compare with online harms because poverty is not a discrete event. It is an upstream condition capable of influencing numerous dimensions of childhood development. Children First Canada’s own Raising Canada 2026 material reports that approximately 1.4 million Canadian children, or 18.3%, live below the poverty line, while approximately 2.4 million children, or 30.8%, live in food insecure households.
Poverty can influence nutrition, housing stability, educational opportunity, mental health, family stress, physical health, recreation, and access to resources. Its significance therefore extends well beyond the number of children living below a particular income threshold.
This raises the same methodological question, “How do we compare an upstream social determinant affecting potentially millions of children with a broad digital category containing everything from relatively minor negative online interactions to serious criminal sexual exploitation?” There is no obvious mathematical answer unless a methodology establishes one.
Recent Canadian evidence also demonstrates the substantial prevalence and consequences of child maltreatment. A 2026 peer reviewed Public Health Agency of Canada study using the nationally representative Canadian Health Survey on Children and Youth examined self-reported maltreatment among youth aged 15 to 17. It found that 44.9% reported emotional abuse, 22.3% physical abuse, and 5.3% sexual abuse, while 39.4% reported exposure to emotional intimate partner violence between caregivers (4). The researchers noted that child maltreatment can produce immediate physical, emotional, behavioural, and academic consequences while increasing longer-term risks involving mental health, substance use, chronic illness, and socioeconomic outcomes.
Again, this does not prove that child maltreatment is “really” #1, however, it demonstrates why declaring any one of these very different categories #1 requires a clearly articulated comparison framework.
Mental health disorders represent another substantial source of morbidity among Canadian youth. The Canadian Institute for Health Information reports that approximately one in five Canadians is diagnosed with a mental illness by age 25, with approximately 70% experiencing symptoms before age 18. CIHI also reports changing patterns in physician visits, emergency department use, hospitalizations, and medication use associated with child and youth mental health (5).
Mental health is particularly difficult to isolate because it overlaps with several other categories on the Raising Canada list. Poverty, maltreatment, bullying, discrimination, family circumstances, physical activity, sleep, and certain online experiences can all interact with psychological well being.
The relationship between digital technology and adolescent mental health also requires careful language. A 2026 umbrella review synthesizing 72 reviews and meta-analyses found that general social-media use showed weak and inconsistent associations with well-being and ill-being, while problematic social-media use was much more consistently associated with poorer outcomes. The authors emphasized the importance of conceptual precision when studying these relationships (6) and that distinction matters. “Using social media,” “using social media heavily,” “problematic social-media use,” and “experiencing an online harm” are not synonymous exposures.
Physical inactivity is another major population-level concern. ParticipACTION’s 2024 Report Card found that only approximately 39% of Canadian children and youth aged 5 to 17 met the recommendation of at least 60 minutes of moderate to vigorous physical activity per day (7). Physical activity is associated with cardiovascular and metabolic health, physical fitness, sleep, psychological well-being, and healthy development. Yet physical inactivity is difficult to compare numerically with sexual exploitation, poverty, injury, or mental illness because once again we are comparing fundamentally different types of measures.
Interestingly, Raising Canada also identifies declining physical activity as one of the risks that online environments may amplify. That may be true for some children, but it further illustrates the difficulty of treating “online harms” as a discrete independent category while simultaneously incorporating its potential effects into other categories.
The overlap problem may be one of the most important methodological issues in the entire ranking. As mentioned in the report, and something that we agree with, online harms are not mutually exclusive from several of the other nine categories. Cyberbullying can simultaneously be classified as bullying and an online harm. Online child sexual exploitation can simultaneously constitute an online harm and child abuse or sexual victimization. Online hate can overlap with racism and discrimination. Problematic digital behaviour may overlap with mental-health outcomes. Sedentary recreational screen use may overlap with physical inactivity.
If online harms are simultaneously being treated as their own umbrella category and as contributors to several other categories, there is a potential problem of conceptual overlap and, depending upon how data were aggregated, possible double counting.
Statistics Canada, for example, reported that in 2019 approximately 25% of Canadian youth aged 12 to 17 experienced at least one of three measured forms of cyberbullying during the previous year. Those experiences included being threatened or insulted online or by text, exclusion from an online community, and having hurtful information posted online (8). That is important evidence of online victimization.
But if those experiences help establish the burden of “online harms” while also helping establish the burden of “bullying,” the categories are no longer independent. The same conceptual problem appears elsewhere.
Statistics Canada reported that among victims identified in police reported cyber related hate crimes between 2018 and 2022, 23% were youth aged 12 to 17. Importantly, this does not mean that 23% of Canadian youth experienced online hate. It means that youth aged 12 to 17 represented 23% of victims within that specific police reported cyber-hate dataset (9). This is why we believe that clear denominators matter when comparing risks.
None of this should be interpreted as minimizing serious online victimization. Statistics Canada’s 2026 analysis of police reported online child sexual exploitation demonstrates why this issue deserves significant national attention. Online child sexual exploitation encompasses offences including luring, non-consensual distribution of intimate images involving minors, invitation to sexual touching, and offences involving child sexual abuse and exploitation material (10).
Children First Canada reports that police recorded 16,905 incidents of online child sexual exploitation in 2024, while police reported incidents increased substantially over the preceding decade (2). These are serious crimes involving real children.
We at The White Hatter have spent almost 3 decades educating families, schools, law enforcement, and young people about sextortion, grooming, online sexual exploitation, image-based sexual abuse, cyberbullying, fraud, harmful content, privacy violations, and other technology facilitated risks. We have also assisted families directly affected by several of these harms. Online harms absolutely belong in any serious discussion about childhood safety in Canada. However, recognizing the seriousness of online harms is different from establishing that this enormous collection of experiences represents the single greatest threat facing Canadian children.
This distinction becomes particularly important when large numbers are used to communicate the scale of online sexual exploitation. A report received by an organization is not necessarily equivalent to one unique Canadian child victim. A police reported incident is not necessarily equivalent to population prevalence. A victim count is not the same statistical measure as the percentage of Canadian children experiencing poverty, physical inactivity, bullying, or food insecurity. Different datasets are measuring different things.
This does not make those statistics unimportant, quite the opposite. They provide critical information about the scale and nature of exploitation. However, they cannot necessarily be placed beside percentages for poverty, maltreatment, bullying, or physical inactivity and treated as directly comparable units of population risk. This is another reason why a transparent ranking methodology matters.
There is one statistic within the Raising Canada 2026 discussion of “online harms” that we believe deserves additional context. The report states, “Canadian adolescents now average seven or more hours of screen use per day—nearly four times national guidelines” (1). The number itself is not fabricated. However, it is presented without sufficient qualification, it can sound more nationally representative and current than the underlying research allows.
The 7.7 hours cited figure originates from a 2024 study by Poirier and colleagues published in the Journal of Adolescent Health. Researchers examined 28,307 adolescents attending 63 high schools in Quebec, with an average participant age of approximately 14.9 years. The researchers specifically investigated changes in sleep and screen time between 2018 and 2022, including changes accompanying the COVID-19 pandemic (11). The study found that screen time increased substantially over the study period, reaching approximately 460 minutes per day in 2022, or about 7.7 hours.
This was a large and valuable Canadian study. However, it was not a nationally representative survey of all Canadian adolescents. It involved adolescents attending participating high schools in Quebec. The timing of this study also matters. The researchers themselves designed the study partly to examine changes associated with the COVID-19 pandemic. Canadian youth experienced periods of school disruption, remote learning, restrictions on activities, physical distancing, and fewer opportunities for in-person social interaction during this period. Technology consequently became even more important for entertainment, communication, education, and maintaining relationships (11).
This does not mean that the 2022 findings should be dismissed as merely a COVID artifact. However, It does mean that caution is warranted when taking a Quebec estimate measured during an historically unusual period and presenting it in 2026 in language that could easily be interpreted as the current national average for Canadian adolescents.
Another important distinction is what did the 7.7 hours actually measure?
They were not 7.7 hours of social media use.
They were not 7.7 hours of gaming.
They certainly were not 7.7 hours of exposure to online harm.
The estimate was constructed from self reported time across several recreational screen activities. The COMPASS research program measures behaviours including television and video viewing, video or computer gaming, internet use, and communication activities such as texting and messaging. School related screen use was treated separately (11)(12).
There is also a methodological consideration involving simultaneous use. A teenager can watch a television program while messaging friends or browsing another device. Because reported time across activities can potentially overlap, summed screen activity measures should not automatically be interpreted as an exact stopwatch measurement of mutually exclusive chronological hours. We should be cautious, however, about assuming how much simultaneous use occurred because the available data do not establish the size of that effect.
There is another important piece of context for this study. The approximately 7.7 hour figure is a mean, or mathematical average, and the study reported substantial variation between participants . A mean does not tell us what every participant, or necessarily the most typical participant, was doing.
Consider a hypothetical example in which five teenagers reported 1, 2, 2, 3, and 12 hours of screen activity. The mean would be four hours, while the median would be two hours. This simply illustrates why an average can sometimes differ considerably from the experience of a typical participant when a distribution contains high values.
We should not assume that this hypothetical pattern describes the Quebec study. A standard deviation or mean alone cannot tell us whether the distribution was strongly skewed. What the reported variation does tell us is that adolescents differed considerably in their screen use (it should be noted and the study reported a standard deviation of approximately 4.2 hours, which is significant). Some were using screens substantially less than the mean and others substantially more. This matters because saying “Canadian adolescents average 7.7 hours” can easily be interpreted by the public as meaning that most Canadian teenagers spend approximately that amount of time every day. The data from this study DOES NOT establish that, and to say so in this report is extremely misleading. If we where to put it more appropriately, Canadian teens in 2022 average 3.5 - 11.9 hours of screen time daily. The unique date and broad range brings increased scepticism to the statistics.
More recent Statistics Canada research provides useful additional context. In March 2026, Statistics Canada published a longitudinal analysis using the nationally representative Canadian Health Survey on Children and Youth. It reported that median screen time among Canadian youth aged 12 to 17 had increased to approximately five hours per day in 2023. Statistics Canada also specifically noted that some of the increase between 2019 and 2023 could have resulted from COVID-19 precautions, including virtual classes, physical distancing, and lockdowns (13).
This does not mean five hours should now become another simplistic national headline. The measurement methods differ, and screen-time measures cannot tell us precisely what young people were doing throughout those hours. What the national data demonstrate is that screens occupy a significant amount of young people’s lives, while also reinforcing the need for precision when describing exactly what is being measured.
We also believe that screen time is not the same thing as online harm. This is perhaps the most important point surrounding the screen time statistic. Watching a movie, FaceTiming a grandparent, playing an online game with friends, creating digital art, scrolling TikTok, researching an interest, learning how to code, messaging classmates, viewing pornography, being cyberbullied, or being sexually extorted can all involve a screen, however, they are not remotely equivalent experiences.
This is why we have consistently argued that discussions about youth and technology need to move beyond simply counting hours. How much time a young person spends online can matter, particularly when use displaces sleep, physical activity, school, relationships, or other important parts of development. But what they are doing, why they are doing it, who they are doing it with, and what is being displaced by that use can matter just as much, and sometimes considerably more.
The research increasingly supports making these distinctions. A 2026 umbrella review found weak and inconsistent associations between general social media use and adolescent mental health outcomes, while problematic social media use was much more consistently associated with poorer outcomes (6).
A large 2024 JAMA Pediatrics systematic review and meta-analysis similarly found associations between social media measures and internalizing symptoms, while emphasizing substantial limitations and gaps in the evidence base (14).
The appropriate conclusion is therefore not that digital technology is harmless, it’s that screen time, social-media use, problematic use, and exposure to online harms are different constructs and should not be treated as interchangeable evidence (18) .
We could take these same ten categories and create a different ranking. For example, if we gave substantial weight to mortality, serious morbidity, population prevalence, disability, developmental consequences, and long-term health effects, a plausible ordering might place preventable injury, poverty and food insecurity, maltreatment, and mental-health disorders above online harms.
But we want to be very clear, “We are not claiming that such an ordering would represent the scientifically “correct” Top 10. In fact, doing so would risk reproducing the very methodological problem we are questioning.
The point of this article is that changing the criteria changes the ranking. For example:
A hierarchy weighted heavily toward immediate mortality and hospitalization might place preventable injury near the top, given its well established burden of childhood death and hospitalization (3).
A hierarchy emphasizing population prevalence and upstream determinants might place poverty and food insecurity near the top (1).
A hierarchy emphasizing prevalence combined with lifelong developmental consequences could place child maltreatment very high (4).
A hierarchy emphasizing morbidity and health-system burden might elevate mental-health disorders (5).
A population-health model emphasizing widespread behavioural exposure could move physical inactivity substantially higher, given that only 39% of Canadian children and youth meet the recommended level of daily moderate-to-vigorous activity (7).
A children’s-rights framework could generate yet another ranking.
A lifetime-risk framework might move climate change much higher. The Canadian Paediatric Society has identified children as particularly vulnerable to climate-related health effects involving extreme heat, air pollution, wildfire, infectious disease, natural disasters, and mental health (15).
Our point, there may not be one scientifically correct ranking. The position of vaccine preventable illnesses mentioned in the report demonstrates another challenge with static rankings. Canada achieved measles elimination status in 1998. Following sustained transmission beginning in October 2024, however, Canada lost that status on November 10, 2025, after more than a year of transmission, largely within unvaccinated and under-vaccinated communities (16). By July 11, 2026, Canada had already reported 1,102 measles cases during 2026 (17). A risk that appears relatively low during a period of strong vaccination coverage can become substantially more important when vaccination rates decline and outbreaks return.
Rankings therefore depend not only upon what is measured but when it is measured. Climate change, also mentioned in the report, demonstrates why the time horizon matters. Climate change creates the opposite methodological problem. If the ranking measures immediate annual mortality or hospitalizations, climate change may appear lower than other threats. If it measures cumulative health risk across the lifetime of a child born today, its position could change dramatically. The Canadian Paediatric Society has warned that children are particularly vulnerable to climate related health effects and identifies risks involving extreme weather, heat, wildfire smoke, air pollution, infectious disease, displacement, and mental-health consequences (15).
Again, the question is not whether climate change should be #10 or #2, the question is what metric and time horizon were used to determine its position.
Children First Canada is an independent national charity focused on children’s rights and well-being, and we congratulate the organization and its research partners for continuing to bring national attention to the challenges facing Canadian children.
We also want to acknowledge the significant expertise involved in Raising Canada 2026. The report was produced in collaboration with PolicyWise for Children & Families, SickKids, and the University of Calgary, drawing upon evidence from researchers, clinicians, governments, community organizations, and young people.
Our criticism should therefore not be interpreted as dismissing the report. In fact, given the expertise and credibility of the organizations involved, we believe greater transparency surrounding the ranking methodology would strengthen, rather than diminish, its contribution.
We absolutely agree that online harms belong on a list of important risks facing Canadian children and youth.
Sextortion is real.
Grooming is real.
Online child sexual exploitation is real.
Cyberbullying is real.
Image based sexual abuse is real.
Online hate is real.
Fraud is real.
Privacy violations are real.
Some forms of problematic technology use are real and deserve attention.
AI is creating additional challenges that families, educators, policymakers, technology companies, and researchers will need to understand.
Parents and caregivers deserve good evidence and practical prevention strategies addressing all of these issues. Our concern is simply with making the much more specific claim that this enormous umbrella category has been empirically established as the “#1 threat facing children in Canada.” If #1 through #10 represents a measured hierarchy of risk, readers should be able to see the criteria, weighting, definitions, data sources, time horizons, and methodology used to determine why one threat ranks above another.
How was mortality weighted against prevalence?
How was severe but relatively uncommon victimization compared with widespread but lower severity experiences?
How were long-term consequences measured?
How were upstream social determinants compared with downstream health outcomes?
How were children’s rights incorporated?
How was future risk compared with present burden?
How were overlapping categories prevented from being counted twice?
How was the cross cutting nature of online harms compared with the equally cross cutting effects of poverty, maltreatment, discrimination, or climate change?
These are not trivial methodological questions. They go directly to what the phrase “#1 threat” means. If the Top 10 instead represents ten important advocacy priorities placed into an editorial order informed by evidence, professional judgment, lived experience, children’s rights, and policy priorities, that is entirely legitimate. However, that should be clearly distinguished from an epidemiological ranking demonstrating that one category produces a greater overall burden of harm than the other nine.
There is also an important policy context. Children First Canada’s release of Raising Canada 2026 explicitly calls upon Parliament to strengthen and pass the Safe Social Media Act, Bill C-34, and describes the report as part of its broader Countdown for Kids campaign (2). There is nothing inherently inappropriate about research informing advocacy, that is common in public health. However, when a numerical ranking is being used to help support public policy recommendations, transparency about how that ranking was generated becomes even more important.
Parents and caregivers deserve to know about sextortion, grooming, cyberbullying, image based sexual abuse, harmful content, fraud, online hate, and other digital risks. They also deserve to know that Canadian children face substantial risks involving preventable injury, poverty, food insecurity, maltreatment, mental health challenges, physical inactivity, discrimination, vaccine-preventable illness, and environmental change, which this report clearly describes.
These risks do not need to compete with one another for our attention. The more useful questions are:
How often does a particular harm occur?
How severe can it become?
Which children are most vulnerable?
What are the short- and long-term consequences?
What factors increase or decrease risk?
What can parents, educators, governments, communities, and industry actually do to prevent it?
Those questions help us move from attention toward prevention. Ultimately, our concern is not whether online harms should be ranked #1, #5, or #8. It is whether the available evidence allows these fundamentally different categories to be placed into a scientifically meaningful ordinal ranking at all.
The evidence clearly establishes that online harms can be serious, that some are increasing, that certain children are disproportionately vulnerable, and that governments, technology companies, schools, parents, and communities have important roles to play in prevention. This is one reason why we like to quote Dr. Pete Etchells, a well respected researcher this field who stated, “So instead of asking, does social media use cause mental health issues? perhaps a better question might be: why do some people prosper online while others get into real difficulty?” (18).
What we do not believe, and has yet to be clearly established is the much more specific proposition that this broad collection of experiences constitutes Canada’s single greatest threat to children, and that distinction matters. It is the difference between identifying an important public policy priority and demonstrating a comparative population-level hierarchy of risk. Both can be valuable, but they are not necessarily the same thing.
Digital Food For Thought
The White Hatter
Facts Not Fear, Facts Not Emotions, Enlighten Not Frighten, Know Tech Not No Tech
References
