The idea that ending spanking and other legal corporal punishment by parents can make a major contribution to preventing physical abuse of children seems preposterous to all but a few of the parent educators, child psychologists, pediatricians and other professionals I talk to. Paradoxically, it is rarely because they are in favor of spanking. Even more paradoxical is that most also think that spanking is an undesirable mode of parenting and many advise parents to use alternatives to correct misbehavior.

Two of the reasons for this paradox are presented in the next section. The section after that explains another paradox: that parents who "don't believe in spanking" do spank toddlers. A third paradox to be explained is that ending a risk factor with a very low effect size, such as spanking, can make a greater contribution to ending physical abuse than ending risk factors that have a much stronger relation to physical abuse. The paper concludes with a discussion of some of the many other benefits of not spanking.

Paradox 1: Even Professionals Who Are Against Spanking Don't Believe That No-Spanking Can Result In a Major Reduction in Physical Abuse

Part of the explanation for this paradox is that, spanking has only a relatively small effect on abuse. That is, only a tiny percent of parents who spank "lose it" and become physically abusive as defined by current law and practice. Despite that, the research shows that most cases of physical abuse started out as corporal punishment.

David Gil's pioneering study of 1,380 children (1970) found that 63% of the abuse incidents were an "immediate or delayed response to specific (misbehavior) of the child." Kadushin and Martin's in-depth study of 66 cases of physical abuse (1981) also found that two thirds were instances of corporal punishment (CP) that had escalated out of control. Perhaps even more important their detailed cases studies revealed the causal processes in the escalation of CP into physical abuse. Most recently, a national Canadian study (Trocmé et al., 2001) again found that, in about two thirds of physical abuse cases, the abusive incident started out as CP to correct misbehavior. Less direct evidence is the study by Straus and Yodanis (Straus & Yodanis, 2001) of a nationally representative sample of American parents. They found that the more CP these parents had experienced as children, the greater the probability that, in bringing up their own children, they would go beyond legally permissible CP and engage in severe physical attacks on their children.

Studies like those just cited, plus clinical observation, led some leading scholars to conclude that that reducing CP is essential to reducing physical abuse and minimizing child mental health and social relationships problems (Feshbach, 1980; Gelles & Straus, 1988; Haeuser, 1991; Kaufman & Ziegler, 1987; Maurer, 1976; Williams, 1983; Zigler & Hall, 1989). This negative view of CP has become part of American culture and permeates the media and the work of professionals who advise parents. There has been a cultural transformation of spanking from something that instills "moral fiber" and is inherently good for children (Straus, 2001), to something that should be avoided but is sometimes necessary. Moreover, even the "sometimes necessary" view is declining. That opinion, and most other aspects of CP, have declined to about half of what they were in the 1960s and 70s. A national Survey of American adults in 1968 found that 94% of the US population agreed that "A good hard spanking is sometimes necessary" (Straus & Mathur, 1996). That has declined to 52% (Straus & Mathur, 1996). In 1975, two thirds of a national of sample of early teen-agers (age 14) had experienced CP that year. Since then CP of early teen-age children has dropped to "only" one third (Straus & Stewart, 1999). Hitting children with traditionally approved objects such as a hairbrush or belt has also declined drastically. It is arguable that these decreases in CP are part of what explains the decreases in physical abuse in recent decades (Finkelhor, 2008; Straus & Kaufman Kantor, 1995).

Paradox 2: Most Aspects of Corporal Punishment Have Decreased By 50%, But Almost All Parents of Toddlers Use Corporal Punishment

These changes raise the most intriguing paradox of the many paradoxical aspects of CP. It is that despite the decrease in use of CP and in cultural approval of CP, almost all American parents (94%) hit preschool age children to correct misbehavior (Straus & Stewart, 1999), which is about the same percentage as in 1975. Moreover these physical assaults on children by the people who love them the most and who, on average, the children love the most, is rarely just an isolated experience. It is chronic in two senses. First, among those who spank toddlers, while some do it very rarely and others do it every day or more often, on average it occurs about three times a week (Giles-Sims, Straus, & Sugarman, 1995; Holden, Coleman, & Schmidt, 1995). Second it is chronic in the sense that it continues on average for 12 years (Straus & Stewart, 1999, Figure 1). Moreover, violent child rearing under the euphemism of "spanking" and "discipline" (including use of hair brushes and belts) continues to be legal in every state of the U.S. provided the child is not physically injured (Bitensky, 1998). In addition, the child abuse statutes in most states specifically declare that CP, when properly administered, is not child abuse. These contradictory trends raise two important questions. First, why hasn't the law changed, despite the drastic declines in most aspects of CP and in the belief that spanking is sometimes necessary? Second, why do almost all parents continue to hit toddlers? In this brief commentary I will focus on the latter question.

Toddlers Get Over It

Part of the explanation is the culturally shared belief that spanking a child at that age, while it may cause upset at the moment, is soon forgotten and does not result in long term harmful consequences. The belief that being physically attacked is less traumatic for a child than an adult (or not traumatic at all) is an aspect of American folk culture for which there is no scientific evidence (Finkelhor, 2008). In fact, what evidence is available, suggests that it may be more traumatic for children than adults. A recent example is the results of a study of 17,252 university students who reported on whether they were spanked before age 12 (Straus, 2008). The results showed that the more CP experienced, the higher the score on a Post-Traumatic Stress Symptoms scale. Moreover, that study also tested the popular belief that when spanking is done by loving and attentive parents, it does not have harmful side effects. It did find that being a child of such parents was associated with lower scores on the Post- Traumatic Stress Symptoms scale. However, it also found that, even among children in the top fifth of positive parenting, the more CP, the more Post-Traumatic Stress Symptoms as a young adult. These findings contradict the culturally ingrained belief that children soon get over the distress of being spanked.

Despite the research, the idea that spanking to correct misbehavior is not harmful to toddlers is so firmly embedded in American culture that it is enshrined in the "consensus statement" recommendations by a conference sponsored by the American Academy of Pediatrics (Friedman & Schonberg, 1996). The third consensus recommendation is that "Spanking a child should not be the primary or only response to a misbehavior." Thus, spanking is permitted if it is not the primary or only method. The fifth recommendation is that "Spanking is not recommended in infants and children under two years because escalation, should it occur, carries a greatly increased risk of causing physical injury." Thus, it is permissible with children age two and older. Moreover, the reason for not hitting children under two is not that it is bad for them, but because of the risk of escalation into physical abuse.

The policy statement by the American Academy Pediatrics (1998) does recognize "the negative consequences of spanking" and advises against it. But the policy statement also implicitly accepts spanking to correct misbehavior. For example, it states that certain types of CP, such as hair-pulling should never be used. It nowhere says that spanking should never be used. Given the years of sometimes bitter debate that preceded the resolution, this was probably as far as the committee could go. Still, it is a sign of progress because it indicates that defenders of CP have accepted the view that CP should not be used with children under two, and the general tone is against any use of CP.

The Deadly Combination

Perhaps the most important explanation for why almost all parents continue to hit toddlers lies in the combination of the behavior of toddlers and the advice given to parents. Almost no one advises parents to never, under any circumstance, spank. Instead, the advice, beginning with Spock (1992), is to "avoid it if you can." But if you are the parent of a two-year old, it does not take long to conclude that you can't avoid it. The recidivism rate for whatever crime a two year old is told to stop doing is about 50% in two hours, and 80% within the same day (Larzelere, Schneider, Larson, & Pike, 1996). That is the average. For some children it is within two minutes. Consequently, on any given day, a parent is almost certain to conclude that so-called alternative disciplinary strategies such as explaining, deprivation of privileges, and time out, do not work. When that happens, and given another false belief in our culture -that spanking works when other things have failed - parents turn to spanking even though they may prefer not to.

The underlying problem is that with a two-year old, in a certain sense, nothing works! More accurately, it is that with toddlers, no matter what the method of correction, on average, the cycle of misbehavior and correction takes a huge number of repetitions. This also applies to spanking. But spankers have an advantage. After the third or fourth repetition of the crime, they do not give up and conclude that they need to try something else, like explaining and diverting. Spanking parents are prepared to spank over and over again until the child learns what to do or not do. But when parents use so-called alternatives, after the typical repetitions of the misbehavior, they turn to spanking. The result is the infamous statistic "94% of parents spank toddlers." We have had 40 years of the "avoid it if you can" message. It is a demonstrated failure for parents of toddlers. We need to move to a policy of advising parents to never spank.

The belief that CP is harmless if done "in moderation" is deeply embedded in American culture and in the American psyche. It results in "selective inattention" (Dexter, 1958) to the scientific evidence demonstrating that these beliefs are false. It is not just the public and politicians who ignore the evidence. It is also researchers, parent educators, pediatricians, and child psychologists. This is illustrated by a 1999 special topic issue on prevention of child abuse of the journal Child Abuse and Neglect. CP was not even mentioned. In 2004 the Journal of Interpersonal Violence also published a special topic issue on prevention, and again not a word on CP. I reviewed child development textbooks published over three decades - ten each in the 1980's, 1990's and since 2000 (Straus & Stewart, 1999). None advised never spanking. Unless groups such as the readers of this article embrace a never-spank position, the prospects of ending CP in the next decade, or perhaps even in the next generation, are not good.

Paradox 3: Ending a Risk Factor with a Very Low Effect Size Can Make a Major Contribution to Ending Physical Abuse

It is a well-established principle in public health and epidemiology that a widely prevalent risk factor with small effect size, for example spanking, can have a much greater impact on public health than a risk factor with a large effect size, but low prevalence; for example, physical abuse (Cohen, 1996; Rose, 1985; Rosenthal, 1984). To illustrate this, let's examine the effect of spanking on being depressed as an adult, compared to the effect of physical abuse.

Assume that: (1) 50 million U.S. children experienced CP and 1 million experienced physical abuse. (2) Also assume that spanking increases the probability of being depressed as an adult by only 2%, but that physical abuse increases the probability of depression by 25%. (3) Given these assumptions, and the fact the CP increases the probability of depression by 2%, then CP appears to be linked to 1 million cases of depression each year. This is so because .02 times 50 million is one million. The additional cases of depression caused by physical abuse is .25 times 1 million or 250,000. Thus ending CP causes four times more cases of depression than physical abuse. Or rephrasing this in terms of prevention, ending CP provides the basis for a four times greater reduction in depression than from ending physical abuse.

The same principle applies to the relation of CP to physical abuse. There may be other risk factors that are much more strongly related to physical abuse such as heroin addiction, but heroin addiction occurs at a fraction of the 94% rate for CP. Thus, reducing CP could contribute more to ending physical abuse than reducing heroin addiction. Of course, we can and should do both. Moreover, the experience of the 1979 no-spanking law in Sweden shows that it can be relatively inexpensive to end or greatly reduce CP. The Swedish approach was to make "never hit a child" a national policy, to publicize this to everyone, including children, and to provide help to parents when there is a problem which leads them to use CP. The Swedish law has no criminal penalty and does not require expansion of expenditures for police, courts, and jails. It puts the emphasis on prevention rather than on waiting until abuse occurs and treating abusing parents (Durant, 1999).

Preventing Physical Abuse is Only One of Many Reasons to End Spanking

In addition to a major reduction in physical abuse, there are many other reasons for a national policy of "never-spank," and for professionals who work with parents to advise "never spanking." First is the huge amount of evidence showing many harmful side effects for children. These include an increased probability of the child hitting parents, hitting other children, and later in life hitting dating and marital partners, slower cognitive development, crime as an adult, child and adult depression, and a greater probability of physically abusing their own children (Straus, 2001; Straus & Medeiros, In Press). Gershoff's meta-analysis (2002) of 88 studies revealed a 93% agreement among studies in finding harmful side effects of CP. There may be nothing else in the huge body of research on parent practices (and perhaps anything else in social science) with that level of agreement between studies. Moreover, there are now longitudinal studies which show that these are indeed side effects of CP, not just the problems that led the parents to spank (Straus, 2001; Straus & Medeiros, In Press).

Finally, spanking should be ended, regardless of whether doing so can reduce physical abuse because children, like adults, are entitled to a life that is free of violence, and because hitting children to correct misbehavior contradicts our hopes for non-violent families and a non-violent society. This is happening in many parts of the world. Both the European Union and the United Nations have called on all member nations to prohibit CP by parents. As of September 2008, 24 nations had done so. Unfortunately, as pointed out earlier, there is strong opposition to ending CP in the United States. Consequently, the United States may be one of the last industrial nations to end this most prevalent of all aspects of family violence and therefore one of the last to use this opportunity to reduce the tragedy of physical abuse.