Fiscal impact of online harms

The preceding analysis estimates the full economic cost of online harms — the combined burden borne by individuals, households, communities, and governments. These are not incrementally new costs relative to the others we have enumerated thus far, but rather a reapportionment across the economy. The point to retain is that governments themselves directly bear some of these costs.

The analysis proceeds through two channels. The first — direct government expenditure — identifies the share of estimated harm costs that flow through publicly funded programs and services. The second — foregone revenue — applies the approach suggested by the structure of Canada's fiscal system: every dollar of national income that is not created because a young person has left the labour force, entered it at permanently reduced capacity, or spent years in treatment rather than employment, is a dollar that the federal government will not collect 16.0 cents from, and the combined federal-provincial-local government sector will not collect approximately 39.8 cents from.1 Both percentages are drawn directly from Canada's Annual Financial Report 2024-25, which recorded federal revenues of $511.0 billion against nominal GDP of approximately $3.08 trillion in 2024, a revenue-to-GDP ratio of 16.6 percent, with Budget 2025 projecting 16.0 percent for 2025-26.

Channel 1: Direct government expenditure

Health care, education, and justice costs that flow through publicly funded programs represent the most direct fiscal exposure. CIHI reports that 71 percent of total health spending in Canada is publicly funded, with about 75 cents of every public dollar being paid by provincial, territorial and sub-national governments.2 Virtually all K-12 education costs are provincially funded. All justice system costs — policing, prosecution, courts, and legal aid — are government expenditures and are generally shared between federal and provincial governments.

By contrast, QALY-based quality-of-life losses represent private welfare losses rather than public expenditures, private financial losses in the gambling category are borne by individuals and households, and diffuse civic costs are not direct government outlays. On this basis, the direct government expenditure attributable to online harms is estimated at $1.55 billion to $2.70 billion annually.

Channel 2: Foregone government revenue

Productivity losses reduce national income. When a young person enters the labour market with permanently reduced cognitive capacity because of platform-driven addictive design, or spends formative years in mental health treatment rather than accumulating human capital, or withdraws partially from the workforce as a result of the psychological sequelae of online sexual violence or cyberbullying, the resulting annual earnings shortfall represents GDP that is never created. Governments collect nothing from income that does not exist.

The productivity base for this estimate is the combined human capital productivity loss from addictive design ($210M–$490M) and cyberbullying ($95M–$142M). Other harm categories (self-harm, hate, OCSE, gambling) generate costs primarily through health system utilization and direct welfare losses rather than labour market participation shortfalls, and are captured under Channel 1.3

Of the combined fiscal burden, the federal government bears an estimated $676 million to $1.2 billion. Provincial and territorial governments bear the larger share — an estimated $1.0 billion to $1.8 billion — primarily through health and education systems. Canadian governments are already spending and forgoing approximately $1.7 billion to $3.0 billion every year as a consequence of platform-facilitated harms that an effective online safety regime could reduce. This is not a prospective cost — it is a recurring annual expenditure that has been accruing since well before the current policy debate began.