Child-spanking at age 5 may have a detrimental effect on child externalising behaviour and cognitive development at age 9
- Department of Criminal Justice and Criminology, Sam Houston State University, Huntsville, Texas, USA
- Correspondence to: Dr Brian B Boutwell, Department of Criminal Justice and Criminology, Sam Houston State University, PO Box 2296, Huntsville, TX 77341, USA;
Commentary on: 
Implications for practice and research
Despite the apparent negative consequences, a large number of American parents utilise corporal punishment.
The adverse effects of spanking on behaviour seem clear, yet the effects on verbal intelligence remain less certain.
Questions remain about the causal effect of spanking on child development; further research with proper controls for heritability is required.
The use of corporal punishment, despite repeated caution from health professionals and developmental researchers, persists as a widespread parenting strategy in the USA.1 As Mackenzie and colleagues note, the prevalence of spanking by American parents appears surprising when one considers the litany of studies linking corporal punishment with deleterious outcomes across the life course. Mackenzie and colleagues extended this body of research by examining the association between the use of corporal punishment and the development of receptive vocabulary skills and externalising problems in childhood during the first decade of life.
Using survey data drawn from the Fragile Families and Child Wellbeing study, Mackenzie and colleagues examined the relationship between spanking and externalising problems and cognitive development in children at 9 years of age. Spanking was assessed at ages 3 and 5 by asking parents how frequently in the past month had they spanked their child. Externalising problems were assessed using the Child Behaviour Checklist (CBCL) at ages 3 and 9.2 Receptive vocabulary skills were measured using the Peabody Picture Vocabulary Test (PPVT) at ages 3 and 9.3 The impact of spanking was estimated using a series of regression equations. Additional control variables included a range of items intended to assess parental-level traits (eg, maternal IQ) and family-level variables (eg, family structure).
Two broad findings emerged from the data. First, maternal spanking (both frequent and infrequent) impacted on the child's behaviour at age 9, regardless of the control variables. The effect remained consistent after the researchers included a prior measure of behavioural problems. Second, the influence of spanking on receptive vocabulary development appeared to have much less of an impact compared to behaviour. High-frequency paternal spanking, however (which represented only one out of several spanking variables), did emerge as a predictor of PPVT scores.
Mackenzie and colleagues’ study represents the latest in a long line of research examining the effect of spanking on child development. In this case, the outcomes of interest represent two key milestones: behavioural development and verbal intelligence. The findings dovetail with prior research highlighting the impact of spanking on problematic behaviour; in particular a meta-analysis that explored associations between the use of corporal punishment and child outcomes undertaken by Gershoff.1 The results, however, differ from prior findings in that Mackenzie and colleagues’ explored the influence of spanking on intelligence. Overall, the analyses appear rigorous, well conceived and well executed. The main contribution of the study is that it reveals yet another apparent relationship between spanking and externalising problems. While this is an important contribution to the literature, methodological problems exist which the authors cannot overcome and which prevent any firm conclusions regarding causality from being drawn.
One aspect in particular of Mackenzie and colleagues’ analysis warrants further discussion: the inability to control for heritability (ie, the influence of genes on individual differences). The influence of heritability on virtually all human traits is pervasive.4 Findings from research undertaken by Barnes et al clearly suggest that the covariance between spanking and externalising problems can, at least to some extent, be accounted for by genetic factors. Heritable influences represent a potential source of confounding that Mackenzie and colleagues cannot address because the original Fragile Families study did not sample more than one child per family, a requirement for behaviour genetic modelling. In addition, the Fragile Families study has not, to date, made the genetically sensitive measures (ie, participant genotypes) widely available.
It is difficult to determine whether the current findings would remain if genetic influences on behaviour, intelligence and the use of spanking were directly modelled; some evidence suggests that they might not.5 However, one cautious note, despite questions about the causal influence of spanking; no consistent line of scholarship has emerged suggesting that spanking represents a factor capable of boosting childhood IQ scores. Nor does spanking improve behaviour in a majority of children, but refer to Lansford and colleagues for evidence concerning race based differences.6 At best, the effects of spanking could be neutral; at worst, they could be devastating.