Loading the library…
parenting & child development
The operative claimParental corporal punishment (spanking) is associated with more child aggression, behavior problems, and other adverse outcomes — never with developmental benefit — across meta-analyses, 69 prospective studies, quasi-experimental matching designs, and 92 countries; average effects are modest and the causal magnitude, though not the direction, is debated.
The research community and pediatric/psychological professional bodies agree that spanking shows no developmental benefit and is consistently associated with worse child outcomes — more aggression, more behavior problems, worse parent-child relationships — in a dose-responsive pattern that generalizes across cultures. A minority methodological position holds that once a child's pre-existing behavior is controlled, the remaining longitudinal effects are trivial in size; that dispute concerns magnitude and causality, not direction, and no analysis finds spanking beneficial.
Applies to parental spanking/physical punishment of children and adolescents as studied in community samples worldwide; distinct from physical abuse in legal definition, though meta-analytic effect sizes for spanking and abuse are similar in kind. Associations are strongest for externalizing outcomes and show dose-response; occasional mild spanking in otherwise warm homes shows the smallest associations, which is where the methodological debate concentrates.
Gershoff ET, Grogan-Kaylor A (2016). Spanking and Child Outcomes: Old Controversies and New Meta-Analyses. Journal of Family Psychology.
Meta-analysis · N = 160,927 · children and adolescents
Thirteen of 17 mean effect sizes were significantly different from zero and ALL pointed toward detrimental outcomes: more aggression, more antisocial behavior, more externalizing and internalizing problems, worse parent-child relationships, and lower moral internalization; no outcome showed benefit. Effect sizes for spanking did not differ substantially from those for physical abuse, and were robust to study-design differences.
Bearing on this claim: Spanking-specific meta-analysis (111 effect sizes, 160,927 children): every significant effect detrimental; none beneficial.
doi.org/10.1037/fam0000191Heilmann A, Mehay A, Watt RG, Kelly Y, Durrant JE, van Turnhout J, Gershoff ET (2021). Physical punishment and child outcomes: a narrative review of prospective studies. The Lancet.
Systematic review · children and adolescents (prospective cohorts)
Seven convergent findings: physical punishment consistently predicts INCREASES in child behavior problems over time; it is not associated with any positive outcome; it raises the risk of child-protective-services involvement; the only child behavior that elicits it is externalizing; quasi-experimental studies show worsening behavior over time; associations are robust across child and parent characteristics; and there is evidence of dose-response. The authors conclude physical punishment is harmful and policy remedies are warranted.
Bearing on this claim: Lancet review of 69 prospective studies: physical punishment predicts increasing behavior problems; no positive outcomes; dose-response.
doi.org/10.1016/S0140-6736(21)00582-1Gershoff ET, Sattler KMP, Ansari A (2018). Strengthening Causal Estimates for Links Between Spanking and Children's Externalizing Behavior Problems. Psychological Science.
Observational · N = 12,112 · young school-age children
Whether children were spanked at age 5 predicted increases in externalizing behavior problems by ages 6 and 8, even after matching spanked and non-spanked groups on a wide range of sociodemographic, family, and cultural characteristics and children's initial behavior problems — the closest to a causal estimate available outside an (unethical) experiment. Spanking predicted deterioration, not improvement, of behavior.
Bearing on this claim: Propensity-score matched national sample (N=12,112): spanking at 5 predicts worse externalizing at 6 and 8 after matching on initial behavior.
doi.org/10.1177/0956797617729816Cuartas J, Gershoff ET, Bailey D, Gutiérrez MA, McCoy DC (2025). Physical punishment and lifelong outcomes in low- and middle-income countries: a systematic review and multilevel meta-analysis. Nature Human Behaviour.
Meta-analysis · children and adolescents in LMICs
Physical punishment was associated with detrimental outcomes across domains — mental health problems, worse parent-child relationships, substance use, impaired social-emotional development, worse academic outcomes, heightened externalizing behavior — with ALL subgroup estimates consistent in direction across 92 countries. Sensitivity checks showed the harms were specific to physical punishment and psychological aggression, not to non-violent discipline. The pattern documented in high-income countries generalizes worldwide; cultural normativeness does not neutralize it.
Bearing on this claim: Pre-registered 189-study meta across 92 LMICs: detrimental associations in every domain, consistent in direction; specific to physical punishment.
doi.org/10.1038/s41562-025-02164-yCounter-authority (1). Studies an adversary may cite — surfaced deliberately so the grade is honest about its limits.
Ferguson CJ (2013). Spanking, corporal punishment and negative long-term outcomes: a meta-analytic review of longitudinal studies. Clinical Psychology Review.
Meta-analysis · children and adolescents (longitudinal samples)
CONFLICTING ANALYSIS: with bivariate correlations, small but non-trivial long-term associations appear (spanking–externalizing r=.14); but with better-controlled partial coefficients (controlling time-1 behavior), effects were statistically significant yet trivial in size (pr ≤ .10 for externalizing/internalizing). Ferguson concludes the long-term impact of spanking on these outcomes is minimal and urges a more nuanced public message. Represents the methodological dissent (with Larzelere & Baumrind) contesting causal magnitude, not direction.
Bearing on this claim: Methodological dissent: with time-1 behavior controlled, longitudinal effects are trivial in size (pr ≤ .10) — contests causal magnitude, not direction.
doi.org/10.1016/j.cpr.2012.11.0024 supporting · 1 counter verified sources
Randomized experiments are impossible, so causality is inferred from converging non-experimental designs; child effects on parenting (difficult behavior eliciting spanking) inflate naive estimates, and the best-controlled designs yield small effects whose practical significance is contested (Ferguson 2013). Most data are parent-reported. Moderation by cultural normativeness, warmth, and severity has been studied with mixed results — large multicultural studies find harm regardless of implementation, but the mildest-spanking subgroups show the smallest effects. This evidence concerns population averages, not any individual family.