The American Academy of Pediatrics (AAP) is a relatively conservative organization, in the sense that they require a significant body of evidence before making a statement or changing a recommendation. For instance, they equivocate on infant circumcision, ignoring both the violent protestations of those who feel the practice is immoral as well as the body of evidence that suggests it’s of modest health benefit (see this post and this one for more discussion on this topic). One topic on which the AAP does take a firm stance, however, is that of corporal punishment and spanking. From the Healthy Children website (AAP):
The American Academy of Pediatrics does not recommend spanking. Although most Americans were spanked as children, we now know that it has several important side effects.
- Even though spanking may seem to “work” at first, it loses its impact after a while.
- Because most parents do not want to spank, they are less likely to be consistent.
- Spanking increases aggression and anger instead of teaching responsibility.
- Parents may intend to stay calm but often do not, and then regret their actions later.
- Spanking can lead to physical struggles and even grow to the point of harming the child.
It is true that many adults who were spanked as children may be well-adjusted and caring people today. However, research has shown that, when compared with children who are not spanked, children who are spanked are more likely to become adults who are depressed, use alcohol, have more anger, hit their own children, hit their spouses, and engage in crime and violence. These adult outcomes make sense because spanking teaches a child that causing others pain is OK if you’re frustrated or want to maintain control—even with those you love. A child is not likely to see the difference between getting spanked from his parents and hitting a sibling or another child when he doesn’t get what he wants.
Each of the statements made above is backed by solid scientific evidence (see, for instance, Bender et al, Berlin et al, Bradley et al, Gershoff, E., Knox, M., Slade et al, Strassburg et al, Strauss et al, Taylor et al).
It’s known that physical and emotional abuse and neglect are associated with certain mental disorders, including major depression and other mood disorders, schizophrenia, obsessive compulsive disorder, and the like (see, for instance, Fergusson et al, Kessler et al, MacMillan et al, Scott et al). One question that has remained unanswered, however, is whether children who are spanked are more likely to experience these same types of mental disorders during adulthood than those who are not spanked.
A new study examined the relationship between spanking (and similar punishments, all in the absence of more serious abuse or neglect) and mental disorders in an attempt to answer this question (Afifi et al). The study involved an examination of a large quantity of data that included information on more than 34,000 individuals. Participants who reported having been spanked, slapped, or similar were compared to those who reported not having been physically punished. Individuals who were spanked but who also reported having been physically, sexually, or emotionally abused or neglected were discarded from the data set, leaving a total of about 20,500 individuals whose data were analyzed.
Of these, just over 19,000 reported not experiencing physical punishment, while 1258 did experience such punishments. Once the data were corrected to remove possible confounding variables (such as parents with mental disorders, which would predispose children to mental disorders), the prevalence of mental disorders was approximately 2-7% higher among adults who’d been spanked as children as compared to those who had not.
Unfortunately, while I’m intellectually inclined to believe that there may be an association between physical punishment during childhood and adult mental disorder, I have a number of concerns about this study and its methodology. First, it’s notable that of over 34,000 original records, only 20,500 remained once the authors removed individuals who had been physically, sexually, or emotionally abused or neglected. This is a tremendous portion of the original population, and simply doesn’t make intuitive sense. On further examination, I find two potential explanations. First, the data were pulled from the National Epidemiologic Survey on Alcohol and Related Conditions, which despite the implications of its name, surveyed a large group of individuals who were more-or-less representative of the general population for information about substance use/abuse habits and mental disorders (Grant et al). Grant does note that the survey over-sampled certain minority demographics, and further, the survey was voluntary and depended upon self-report of information (with an approximate return rate of 81%). As such, it’s possible that individuals who were resentful of the way they’d been parented (and who were perhaps more likely to report a mental disorder) were also more likely to return the survey. Unfortunately, there’s no way to know whether this is a significant factor.
Far more likely to produce the startling proportion of survey respondents who were not included in this study because of abuse and/or neglect is the definition of abuse used by Afifi et al. Per the authors, individuals who were removed from the study population were those who (among more traditional definitions) were hit hard enough to leave marks, had objects thrown at them, were sworn at, and/or were made afraid during or by punishments. It’s not difficult to imagine, given this definition, that there would be very few individuals left in the physical punishment category once those in the abuse category had been removed.* Whether the authors’ definition of abuse is appropriate or not, the exclusion from the study of many individuals who were physically punished in ways that would not be considered abuse in the traditional sense weakens the study by reducing the number and the type of participants. It would be interesting to see what the data would reveal if the definition of abuse were more narrow (and more traditional), allowing inclusion of a larger group of those who’d been physically punished in “non-abusive” ways.
*Note that here I do not express either agreement or disagreement with the authors’ definition of abuse; I merely suggest that per the given definition, it’s not surprising that there were relatively few participants left in the punishment condition.
Along these same lines, the authors apparently fail to recognize the effect that the elimination of this large group of participants might have had on their data, leading them to make some bizarre observations. For instance, they note that physical punishment is more likely in families in which parents are better educated (39.2% of physically-punished children had a parent with a post-secondary degree) than in less-educated families (8.8% of physically-punished children had parents who hadn’t completed high school). The authors expressed surprise at this finding, as would I were I faced with the raw data. However, a more reasonable explanation (and of course, without the 14,000 records that were not included in the study, this is purely speculative) is that individuals from lower-education families were more likely to be punished physically in a way the authors deemed abusive, and were thus thrown out of the study. From a scientific perspective, the lack of awareness of these sorts of possibilities on the part of the authors distresses me; generally speaking, scientists value the work of other scientists more highly when the researchers are aware of potential confounding factors and methodological limitations.
Another concern I have with the study is that the authors make the mistake of confusing (at least in their language) correlation with causation. In the abstract results, they state:
Approximately 2% to 5% of [one category of mental] disorders and 4% to 7% of [another category of mental] disorders were attributable to harsh physical punishment.
Based upon the study design, it’s impossible to say that the disorders were attributable to harsh punishment; the design allows only the statement that harsh punishments were positively correlated with mental disorders. Regardless of this error, however, the findings nevertheless somewhat strengthen the (already considerable) body of evidence suggesting that spanking and other corporal punishments aren’t effective, and are likely harmful.
As a scientist, I’m not particularly impressed by this new research; I’d want to see what sort of data the very large group of survey respondents (about 14,000 in all) who were thrown out for abuse might have contributed if the definition of abuse had been a bit more in keeping with traditional usage. Again, I want to make it clear that just because a punishment isn’t classically defined as “abuse” doesn’t make it right. In classical terms, however (and here I refer, for instance, to the social mores regarding appropriate discipline techniques in the 50s, 60s, 70s, and 80s) “spanking” was one thing, and “abuse” was another. Clearly, the delineation between the two isn’t absolute, but I nevertheless feel that by defining “abuse” too broadly, we lose the opportunity to extract potentially valuable information about the damage done by “non-abusive” physical punishment.
The weaknesses of this particular study aside, there is nevertheless a tremendous body of compelling science that shows spanking and other physical punishments are simply not effective, and likely contribute to at least some lasting negative effects through childhood and into adulthood. In lieu of spanking and corporal punishment, the AAP recommends techniques such as natural consequences and time-out. For those who don’t mind a slightly more technical read, there’s an interesting paper by the AAP’s Committee on Psychosocial Aspects of Child and Family Health here. Those looking for a less technical guide to effective and positive discipline may find some good information in Elizabeth Pantley’s The No-Cry Discipline Solution (McGraw-Hill), which has suggestions for children of all ages, and Harvey Karp’s The Happiest Toddler On The Block (Bantam), which focuses on children aged 12 months to 4 years.
Science Bottom Line:* The new study of physical punishment and mental disorders in adults doesn’t establish a strong link due to flawed methodology and interpretation, but the existing body of research that discredits spanking as a viable discipline technique is large enough to be compelling without the addition of these results.
How do you feel about spanking children?
Afifi et al. Physical Punishment and Mental Disorders: Results From a Nationally Representative US Sample. Pediatrics. 2012 Jul 2. [Epub ahead of print]
Bender et al. Use of harsh physical discipline and developmental outcomes in adolescence. Dev Psychopathol. 2007;19(1):227–242
Berlin et al. Correlates and consequences of spanking and verbal punishment for low-income white, African American, and Mexican American toddlers. Child Dev. 2009;80(5):1403–1420
Bradley et al. The home environments of chil- dren in the United States part II: relations with behavioral development through age thirteen. Child Dev. 2001;72(6):1868–1886
Fergusson et al. Exposure to childhood sexual and physical abuse and adjustment in early adulthood. Child Abuse Negl. 2008;32(6):607–619
Gershoff, E. Corporal punishment by par- ents and associated child behaviors and experiences: a meta-analytic and theoret- ical review. Psychol Bull. 2002;128(4):539–579
Grant et al. Prevalence, correlates, co-morbidity, and comparative disability of DSM-IV generalized anxiety disorder in the USA: results from the National Epidemiologic Survey on Alcohol and Related Conditions. Psychol Med. 2005;35(12):1747–1759
Kessler et al. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychol Med. 1997;27(5):1101–1119
Knox, M. On hitting children: a review of corporal punishment in the United States. J Pediatr Health Care. 2010;24(2):103–107
MacMillan et al. Childhood abuse and lifetime psychopathology in a community sample. Am J Psy- chiatry. 2001;158(11):1878–1883
Scott et al. Prospectively ascertained child maltreatment and its association with DSM-IV mental disorders in young adults. Arch Gen Psychiatry. 2010; 67(7):712–719
Slade et al. Spanking in early childhood and later behavior problems: a prospective study of infants and young toddlers. Pediatrics. 2004;113(5):1321–1330
Strassburg et al. Spanking in the home and children’s sub- sequent aggression toward kindergarten peers. Dev Psychopathol. 1994;6(3):445–461
Strauss et al. Spanking by parents and subsequent anti- social behavior of children. Arch Pediatr Adolesc Med. 1997;151(8):761–767
Taylor et al. Mothers’ spanking of 3-year-old children and subsequent risk of children’s aggressive behavior. Pediatrics. 2010;125(5)