Legal Banning of Disciplinary Spanking in Canada: Evidence and Human Rights – a Response to Gunnoe and Colleagues

Using scientific evidence to inform the legal banning of disciplinary spanking in Canada is critical. Decades of high-quality research indicate the existence of robust evidence that spanking increases the likelihood of impairment, including developmental, behavioural and mental health problems, physical health conditions, substance use, and suicidal thoughts (1–5). Evidence strongly supports the repeal of Section 43 of Criminal Code of Canada. However, Gunnoe et al. in a recent commentary conclude otherwise due to fallible reasoning and persistent reliance on outdated, flawed studies.

First, the assertion of relying on causal inference from randomized controlled trials (RCTs) is misleading. While it is true that RCTs are often considered the gold standard in health research, there are many issues in health and social sciences where use of such trials is not appropriate due to ethical considerations. The authors highlight results from four outdated studies published over three decades ago. Although they acknowledge the outdated aspect of these studies, they still refer to them as providing the best evidence. These studies are unethical as they randomized children to be spanked. The authors’ recommendation for similar future studies is disturbing and untenable. Furthermore, they discount the importance of high-quality meta-analytic studies, which show evidence against spanking. Paradoxically, while the authors point to RCTs as the gold standard, they also use anecdotal information and speculation to support their arguments.

Second, the authors’ steadfast focus on child compliance following corporal punishment is irrelevant to law reform. The key consideration is whether spanking increases the likelihood of harm to the child as outlined above.

While no study is without limitations, the literature on impairment has expanded exponentially to include hundreds of high-quality studies, with robust conclusions that collectively and consistently indicate that spanking children increases the likelihood of detrimental outcomes across their lifespan (1–5). While we agree with the authors that causal inferences cannot be made because of ethical considerations that preclude use of study designs that meet criteria for causation, this is true for many health and social science issues (smoking is a good example) and does not discount the validity and importance of statistically significant findings for cohort studies. Furthermore, the authors are inaccurate in their assertion that recommendations not to spank children rely on evidence from unadjusted correlations. We would argue that most studies go well beyond simple unadjusted correlations and use more sophisticated modeling, which adjusts for important covariates. The authors also criticize studies using analysis of covariance (ANCOVA) as a statistical method when their argument appears to be about selection bias in recruitment of families who spank, compared with those who don’t. More specifically, baseline difference between groups may not be accounted for in these studies and, therefore, “be biased against corrective actions”. While such a study design may be used more commonly for studies that test compliance following spanking; however, compliance is not the outcome that is relevant to law reform and should not be emphasized over serious harms related to spanking. It should be further noted that ANCOVA is not a common statistical method in studies assessing harms related to spanking.

Third, the purposes of spanking bans are to establish clear social norms, reduce support for and use of corporal punishment, and provide families with interventions that prioritize non-punitive discipline (6). The authors misrepresent the data following spanking bans and inaccurately interpret these findings using speculative arguments. Following the legal prohibition of corporal punishment, data indicate an increase in child maltreatment reporting (7–8). This may reflect greater awareness and readiness to report (7), and changes in child welfare reporting practices. In Sweden, and several other countries, following law reform, there was a steady decrease in public support for corporal punishment and use of spanking, and lower rates of child criminal death and child abuse mortality (6, 9).

Fourth, a critical point not addressed in the commentary is that children have the fundamental human right to be protected from all forms of violence, including corporal punishment. To date, 67 countries/states worldwide have banned corporal punishment in all settings to protect children and uphold their rights (10). Allowing disciplinary spanking by law or otherwise violates the United Nations Convention on the Rights of the Child, which Canada ratified in 1991. Human rights should be paramount when considering such legal reform. In fact, a child’s right to live free from violence should be enough reason to change the laws in Canada.

In summary, this commentary misinterprets the scientific literature on spanking and ignores essential human rights issues. The scientific evidence supporting the legal banning of spanking is robust. We must protect children and promote the best opportunities for them to live happy and healthy lives free from violence.