There’s pretty much no benefit to spanking children to discipline them, and mountains of evidence show that spanking risks harming children, both short-term and long-term. Kids become more aggressive and anti-social, have more mental health problems from childhood into adulthood and misbehave more—and are more likely to end up abused. That’s the conclusion of the most recent meta-analysis on spanking, published in the Journal of Family Psychology and involving more than 160,000 children—though it’s unlikely to settle the debate that has continued for years over whether spanking is acceptable or whether it hurts children.
A meta-analysis is sort of the mother of all scientific studies: researchers bring together many past studies that studied the same research questions, use specific criteria to narrow down the most similar and/or highest-quality ones, and then analyze all the numbers together to get a sense of what the general consensus in the overall evidence base is. Just four previous major meta-analyses have been conducted on the short-term and long-term outcomes of children who were spanked. This new meta-analysis, authored by Elizabeth Gershoff, PhD, at the University of Texas at Austin, and Andrew Grogan-Kaylor, PhD, of the University of Michigan, set out to address two criticisms of the past research—that spanking has too often been combined or conflated with more abusive punishment or behaviors, and that only poor quality research has found negative effects from spanking.
Despite hundreds and hundreds of studies, the general public, and sometimes even some social scientists, can’t seem to agree on how harmful or beneficial spanking is. Part of the problem is that the research on spanking is messy. It’s hard to separate out all the different factors that might influence how a child turns out, and spanking has a chicken-or-egg problem: Do more difficult children simply get spanked more by their parents, or does getting spanked cause kids to act out and misbehave more? (The handful of long-term studies looking at this question point to the latter.) But the research isn’t so messy that clear patterns haven’t emerged from it: the vast majority of social scientists agree that spanking can lead to problems in childhood and adulthood and doesn’t have any real upsides. Only a handful of social scientists disagree, and their studies tend to be less specific and lower-quality.
All 10 of the other findings showed that spanking was linked to negative outcomes to varying degrees. Children who were spanked had a poorer relationship with their parent and had lower levels of moral internalization, which means they were less able to determine that something was morally wrong for its own sake rather than knowing it was wrong because they’d get smacked otherwise. Spanking was also linked to poorer mental health, higher levels of aggression and antisocial behavior—both in childhood and later on in adulthood. Children were also more likely to become victims of physical abuse and had a higher risk of physically abusing their own child or their husband or wife if they had been spanked.
Considering how many studies were included and how many factors were studied, it would seem that the previous paper covered all the bases in showing that spanking is definitely related to harm of children. Yet naysayers who support spanking continue to criticize the past research for one reason or another, and spanking remains common and supported in the U.S. in general. An ABC News poll of over 1,000 adults two years ago found that 65% of Americans approve of spanking, and half said they sometimes spank their child (though research suggests that’s an underestimate). Another national study in 2014 found that 76% of men and 65% of women agree that children sometimes need “a good hard spanking.”
So this new study by Gershoff and Grogan-Kaylor set out to answer two of the biggest criticisms of past research. The authors’ first research question was whether the harm to children seen in past spanking research was driven primarily by including harsh or abusive discipline methods in the overall definition of spanking. In other words, were the problems that spanked kids had actually linked only to spanking, or were those kids also abused by their parents aside from spanking? The second question looked at whether past studies were of high enough quality and statistically strong enough to show that the negative outcomes really were linked to spanking.
From an initial review of more than 1,500 studies across 50 years, the authors narrowed down the studies they included to 75 research papers. Just over half of these had never been included in the past four meta-analyses, making this study the most comprehensive and up-to-date one available. The authors relied on four criteria to narrow down the studies they examined: a) the study was peer-reviewed, b) it assessed spanking by itself, separate from harsher methods, c) it accounted for other factors that might affect children’s outcomes and d) the study included enough statistical data for the authors to analyze it. Most studies included in this meta-analysis used the term “physical punishment,” defined as “non injurious, open-handed hitting with the intention of modifying child behavior”—or, more simply, spanking.
Out of 17 negative outcomes, the researchers found that spanking was linked to 13 of them. The strongest associations with spanking were a negative parent-child relationship, child mental health problems and an increased risk of a child becoming an abuse victim, followed by low moral internalization, more aggressive child behavior, antisocial behavior in childhood and adulthood, more behavior problems in childhood and mental health problems in adulthood. To a lesser extent, spanking was also linked to lower self-esteem in kids and diminished thinking abilities.
The authors’ statistical analysis also very clearly showed that these outcomes were specifically linked to spanking—not to other kinds of harsher punishment or abusive behavior—and that the studies showing it were strong. All these outcomes are the same ones seen, to a greater severity, with kids who actually are abused. Basically, spanking and child abuse occurred along a continuum that showed the same kinds of short-term and long-term problems, but the problems are worse and bigger in children who are abused.
The biggest weakness of this study, as with nearly all studies on spanking: It’s not possible or ethical to randomly separate families into two groups where one group spanks and one doesn’t and then see how the kids turn out. But the associations and statistics in this study and others is solid enough to strongly suggest spanking does, in fact, cause these problems. The authors conclude, “Parents who use spanking, practitioners who recommend it and policymakers who allow it might reconsider doing so given that there is no evidence that spanking does any good for children, and all evidence points to the risk of it doing harm.”