To appear in Murray A. Straus, The Primordial Violence:  Corporal Punishment By Parents,

Cognitive Development, And Crime, Walnut Creek, CA:  Alta Mira press, In Press.



Chapter 14





Murray A. Straus

Family Research Laboratory, University of New Hampshire

Durham, NH 03824  603-862-2594 





A1.  If My Dad Hadn't Taken Me Over His Knee, I Would Be In Big Trouble Now. 3

A2.  Spanking Is Needed As A Back-Up Or Last Resort 3

A3.  The "Eight Strongest Studies" Demonstrate "Beneficial Outcomes" Of CP. 4

A4.  "Authoritative" Parents Use CP And Their Children Have The Best Outcomes. 5

A5.  Parents Spank In Countries Such As Japan That Have Low Crime Rates. 5



B1.  I Was Spanked and I'm OK. 5

B2.  If CP Makes Things Worse, Why Does Misbehavior Decrease As Children Grow Older?. 6



C1.  Since CP Has Been Abolished In Schools, Discipline Has Declined And School Violence Is Up. 6

C2.  Crime By Youth Is Up Because Parents Are Not Spanking. 6

C3.  No-Spanking Means No Discipline And Children Running Wild. 7



D1.  Low Education Parents Lack the Cognitive And Linguistic Skills Needed To Avoid CP. 7

D2.  Spanking Is Necessary Because Toddlers Don’t Understand Explanations And Reasoning. 8

D3. “Never Spank” Is A Negative Approach And Won’t Work.  It Is Better To Use A Positive Approach And Teach Alternatives. 8

D4  Spanking Is Universal And Can't Be Dealt With By Policies Or Laws. 9



E1.  The Real Problem Is Bad Parenting. 9

E2.  The Important Issue Is The Overall Pattern Of Parent Behavior 9

E 3. The Link Between Spanking And Mental Health Problems Is Genetic. 10

E 4  Overlap Of Corporal Punishment And Physical Abuse. 10



F1.  CP Is Not Harmful If It Is Perceived As A Legitimate Exercise Of Parental Authority. 10

F2.  CP Is Appropriate In The Context Of African American Culture And Life Circumstance. 11



G1.  Spanking Is A Response To Misbehavior Rather Than A Cause Of Misbehavior 11

G2.   CP By Itself Has Only A Small Effect And Is Not Worth Worrying About 11

G3  There Is No Evidence That “Moderate” CP, Used Only Rarely, Has Harmful Effects. 12

G4. We Should Wait For More Conclusive Research Before Advising Parents To Never- Spank. 12

G5.  If All Children Are Spanked, It Is A Constant and Cannot Explain Anything. 12

G6  Inadequate Controls For Other Variables. 12

G8  The Effect Size Is Too Small To Make A Difference. 13

G9 Even Prospective Research Does Not Prove A Causal Relation. 14

G10  The Results May Be Due To Recall Bias. 14



H1  The Bible Tells Parents To Spank. 14

H2  We Should Focus on More Important And Crucial Threats To Children Such As Poverty And Racism   14




            The preceding chapters make clear that public opinion and an increasing number of parents have shifted away from corporal punishment (CP).   As a result, those who continue to favor corporal punishment now have had to defend what was once a taken-for granted and near universal belief.  As a result, CP has become the focus of a sometimes-intense debate (see Chapter 1).  The debate has taken place in scientific journals (Friedman, Schonberg, and Sharkey 1996), in the mass media (Lemonick and Park 1997), and in the courts (Associated Press 1997).  The purpose of this chapter is to show that most of the arguments used in this debate to defend CP do not hold up under scrutiny, and that others are no more valid than equally plausible arguments against CP.


Many of the defenses of CP analyzed in this chapter seem on cursory examination to be scientifically sound.  Consequently, even social scientists who are opposed to CP are likely to wonder if they are really correct in opposing CP.  One example is Robert Larzelere’s review of the literature on the effects of CP (Larzelere 1996).  He winnowed the many studies down to the “eight strongest studies” and concluded that they all show “beneficial effects.”  However, when one reads the actual studies, what they show is that, without exception, non-corporal methods were equally effective (see section A3 of this chapter).


Altogether, the chapter discusses 32 defenses of CP, grouped under eight headings.  I do not anticipate that many readers will read all 32.  It is just too much.  Instead, the chapter is intended to be a reference source;  that is, as a place to look up a particular defense of CP, not a document to be read from start to finish.  I suggest using the Table of Contents to locate defenses of interest.  In addition, because few readers will examine all 32 defenses, some redundancies were left.   For example, the fear that children who are not spanked will be out of control and engage in delinquency and adult crime has been almost an article of faith in American culture.   Some of the ways this idea is expressed are dealt with in the defenses A2 and B1, 2, and 3.




A1.  If My Dad Hadn't Taken Me Over His Knee, I Would Be In Big Trouble Now

(Insert Figure 14-1 about here)

If “big trouble” means being convicted of a serious crime,  Figure 14-1 shows that when parents use CP to try to straighten out a son, the probability of crime later in the child’s life is increased.  The data are for a sample of boys in a “high risk” urban area who were followed for 35 years (McCord 1991).  This is the kind of environment in which parents often feel that they need to use “strict discipline” (i.e., CP) to protect their children.   But rather than protecting them, Figure 14-1 shows a much higher rate of criminal convictions for the boys whose parents used CP.   A similar boomerang effect is also shown in Chapters 5 through 9.


Why does CP increase the probability of delinquency and crime?   Several processes are involved.  For example, Chapter 6 shows that CP undermines the bond between a child and his or her parents.  That bond is extremely important in motivating children to heed what their parents have to say.  Another process is illustrated in Chapter 9 and 10.  Those studies show that CP slows down mental development and then reduces educational attainment.  The importance of this is that not doing well in school is a major risk factor for delinquency.


Of course, there are instances in which CP does seem to have “straightened out” a delinquent child.  But it should not be assumed that the change for the better was the result of the CP.  Studies of juvenile delinquency show that crucial elements in delinquency prevention include a combination of clear and consistent parental rules and parental monitoring and love and support (Sampson and Laub 1993a).  If the parents had shown the same level of love, concern, and involvement; had laid down clear rules, kept tabs on the child’s behavior; and had used non-violent methods of discipline when those rules were broken, the outcome is likely to have been even better.



What about adding spanking to “good parenting”?   The studies summarized in section E1 indicate that when parents spank, if they also provide love and support, are consistent in stating and enforcing rules and monitor children, the probability of later delinquency and psychological problems is reduced, but spanking remains associated with an increased probability subsequent misbehavior and psychological problems (see chapters 5, 6, and 7).

A2.  Spanking Is Needed As A Back-Up Or Last Resort

            Even parents who are not in favor of CP usually feel that there may be rare occasions when it is necessary.  Many parents say, “I don’t believe in spanking, but sometimes there is no alternative.  You can’t let a toddler run out in the street and you can’t let a child get away with things.“  The research evidence shows that this is a myth.


To understand what the research says about this belief, it is necessary to distinguish between the short-run or immediate-situation effectiveness, intermediate-term effectiveness  such as the subsequent 8 hours, and long-term effects such as months or years later.


The immediate-situation effectiveness of spanking is not in dispute.  However, non-violent control strategies, such as explaining to the child, depriving a privilege, or just walking up to a child and saying “No” or “Stop,” or putting a child back in a time out chair, work just as well in the immediate situation (Day and Roberts 1983; Larzelere, Sather, Schneider, Larson, and Pike 1998; Larzelere, Schneider, Larson, and Pike 1996; LaVoie 1974; Roberts and Powers 1990).


The short-run effectiveness of spanking is also no better than the effectiveness of alternatives.  Figure 14-2 shows that a typical 2-year old is likely to repeat whatever misbehavior is corrected within the same day, regardless of the method of discipline (none of the differences between modes of correction are statistically significant).  Or putting it another way, with toddlers, all methods of correction, including spanking, have a very high short-run failure rate. The “recidivism rate” for toddler misbehavior is about 80% within the same day and 50% within two hours regardless of whether spanking or some other corrective step has been used (Larzelere, Schneider, Larson and Pike 1996).  For some children and on some occasions for all children, it is within two minutes.


As for long-term effects, the research clearly shows that non-violent disciplinary strategies work better.  This was shown in chapters 5, 6, and 7 for juvenile violence and delinquency (see also Figure 14-1 and (Gunnoe and Mariner 1997); and in Chapters 8 and 9 for violence and other crime by adults and much other research (see also Chart 14-1 Brezina 1999; Gunnoe and Mariner 1997; Simons, Lin, and Gordon 1998; Straus, Sugarman, and Giles-Sims 1997).  On average CP boomerangs and results in an increase in misbehavior.  In addition, CP has many harmful long-term side effects including an increased probability of depression (DuRant, Getts, Cadenhead, and Emans 1995; Straus 1994), substance abuse, and suicide (Straus and Kaufman Kantor 1994), and a slower rate of cognitive development and a lower probability of graduating from college (see chapters 10 and 11).   Finally, non-corporal modes of control also have side effects, but they are likely to be positive, such as a better-developed conscience (Sears, Maccoby, and Levin 1957), higher self-esteem (Coopersmith 1967), and a closer bond between the child and the parent (see Chapter 7).


A3.  The "Eight Strongest Studies" Demonstrate "Beneficial Outcomes" Of CP

The review of research on CP by Larzelere (Larzelere 1996) purports to show that “the eight strongest studies” found that CP has a beneficial effect.  Because the review was published in a recognized journal and seems to reflect a careful evaluation of a large number of studies

, it is important to examine the degree to which the conclusion of “beneficial outcomes” for CP is justified.  To do this, it is again necessary to distinguish between short-run and long-term outcomes of CP (see the discussion of immediate-situation, short-run and long-term effectiveness of CP in section A3).  Moreover, the crucial question is whether the outcomes from using CP are superior to the outcomes associated with non-violent discipline strategies.


Short-Term Effectiveness.  Seven of the “the eight strongest studies” refer to short-term compliance with a parental request.  Of these seven, five compared CP with alternatives.  All five found the alternatives to be just as “beneficial” as CP.  For example, Day & Roberts (1983) compared spanking as a back-up for leaving a time-out chair with placing the child in a room with a waist-high barrier held across the door for one minute.  They found that "both spank and barrier procedures were equally effective at increasing compliance" (p. 141), and that "There was no support for the necessity of the physical punishment….” (p. 150).  A replication (Roberts and Powers 1990) obtained the same results.   Of course, taken literally, Larzelere is correct in saying that these studies found “beneficial outcomes” in the sense of the spanked children complied with the parental requests.  But I think it is misleading to not have more clearly stated that non-corporal discipline worked just as well.


Long-Term Effects. Only one of the eight “strongest” studies was about long-term effects (Bernal, Duryee, Pruett, and Burns 1968).  It does show a beneficial effect of spanking.  However, it takes a leap of faith in CP to regard it as a “strong” study.  It is about a single case.  Moreover, that case was a child with severe conduct disorder, and possibly schizophrenia.  So even that one case does not provide information that applies to children in general.  Most important of all, only a minor part of the intervention was the use of CP.  The major part was training the mother in how to respond appropriately, such as when to not take the bait provided by this child’s misbehavior.  The mother was also trained to reinforce positive behaviour and to issue commands confidently and consistently.  Thus spanking was confounded with other interventions and the study provides no evidence that the spanking part of the intervention was what improved the child’s behaviour.  In fact, the intervention might have been even more effective if the spanking part of the intervention had been omitted.


Larzelere also reviewed ten prospective studies and 17 retrospective studies.  He summarizes the prospective studies as follows:  “Three (30%) found predominantly detrimental effects, whereas the other seven (70%) found predominantly neutral outcomes.  In short, the results were either no benefit or determent.   Of the retrospective studies, nine (53% of the 17) r found predominantly detrimental outcomes, 7 (41%) found predominantly neutral outcomes, and 1 (6%) found predominantly beneficial outcomes.”   So, contrary to Larzeler’s implication of beneficial effects, examination of the actual studies, reveals that none of the 10 prospective studies and only one of the 17 retrospective studies found a beneficial outcome.  Moreover, there were substantial percentages (30% and 53% with detrimental outcomes).


            The New Prospective Studies.  Since the publication of Larzelere’s review, almost a revolution has occurred in research on the effects of CP.  There are now five prospective studies.  All show that the long-term effect of CP is counterproductive in the sense of higher rates of misbehavior two and 4 years later for children who were spanked versus lower rates for children whose parents avoided CP (Gunnoe and Mariner 1997; Straus, Sugarman and Giles-Sims 1997), or that CP had harmful side effects, including slowing the rate of children’s cognitive development ((Straus and Paschall 1999) and violence by adolescent boys towards their parents (Brezina 1999) or towards a dating partner (Simons, Lin and Gordon 1998).  These studies are summarized in the concluding chapter to the 2nd edition of Beating The Devil Out Of Them (Straus 2000b).


A4.  "Authoritative" Parents Use CP And Their Children Have The Best Outcomes

Diana Baumrind (1996) classified the parents in her prospective study of 32 Children into three groups:  Permissive, Authoritarian, and Authoritative.  The children of the Authoritative parents were the best behaved and best adjusted.  Baumrind defends CP by pointing out that the Authoritative parents used CP.  She implies that this was part of what made them effective parents.  However, she has never published any results on the use of CP by these parents.  I do not doubt that they used CP, but my guess is that if she were to publish that data, it would show that all three groups used CP, and that the Authoritative parents did it less than the two other groups.


My interpretation of the use of CP by the authoritative parents is that the better outcome of the children of Authoritative parents  occurred despite their using CP.  If and when Baumrind publishes the correlation it will be an example of a “spurious correlation.”  These children did well because an Authoritative style of parenting includes being supportive and responsive to children while also exercising firm, loving, control.  These are characteristics that have repeatedly been shown to be beneficial for children.  The children of the Authoritative parents might have done even better had the parents followed an Authoritative approach while also avoiding CP.  This is an easily tested hypothesis.  I described it in a letter to Diana Baumrind in the early 1990's.  She has had ample time, and also a grant to cover the costs of testing the spurious correlations hypothesis, but has not.


A5.  Parents Spank In Countries Such As Japan That Have Low Crime Rates

            The idea that the low crime rate in countries such as Japan is related to their use of CP is presumes that parents in the US do not spank; whereas, as shown in Chapter 2, 94% of parents of toddlers spank and they do so frequently.  Moreover, CP does not cease on average for American children until age 13, and one out of four still experience CP at age 16.   Thus, it would be more accurate to change this to “Parents in societies such as the USA spank frequently and they have high crime rates.


Beyond the erroneous factual basis and erroneous logic of this defense, not much more can be said because there has been no study comparing the amount of CP in Japan and the USA.   Probably a number of things combine to produce the low crime rate in Japan.  One may be what can be called “attachment parenting, “including extended breastfeeding, co-sleeping, and carrying the baby close to mother [Estrada, 1987 #5134; Conroy, 1980 #5132; Hess, 1987 #5133].   Another possibility is  the close supervision and monitoring of children and the degree of social control characteristic of Japanese society as a whole.  Adults as well as children are closely monitored and supervised by informal social networks and the police.  I lived in Japan for a month and it was a wonderful experience.  However, I would not want to live permanently in such a closely supervised society.  If the price is more crime, that is a price I and most other Americans are willing to pay – at least up to a certain point.   A society, in effect, chooses the balance between freedom and order.  In the US, we opt for more freedom and we pay the price with less order.  Japan is moving to more freedom, and that will contribute to more crime.  In the US, there has recently been a movement towards more survailance and more order and that may have contributed to the decline in crime since the mid 1990s.





B1.  I Was Spanked and I'm OK

            One of the most frequent and intuitively attractive defenses of CP is  “I was spanked and I’m OK.”   Let us start by assuming that the “I’m OK” part is correct.  However, the intended implication – that therefore spanking is OK – is not correct.  The correct implication is that the person saying this is one of the lucky ones who were not adversely affected by CP.  To understand this, one has to first understand that CP is a “risk factor,” not a one-to-one cause.  This means that the more CP experienced, the greater the risk or probability of a harmful side effect.  Fortunately, as with other risk factors, CP does not guarantee a harmful effect.  The research evidence indicates that most people who were spanked are OK,  but as shown repeatedly in this book, when compared to others, the percent who are OK is lower than among those who were rarely or never spanked.


A well-known example of a risk factor is smoking.  People who smoke more than a pack a day have about a one in three chance of dying of lung cancer or some other smoking-related disease (Matteson, Pollack, and Cullen 1987).  That same statistic also means that two thirds of heavy smokers will not die of it.  They will be able to say, as millions said when the research on smoking first appeared,  “I’ve smoked all my life and I’m OK.” The  “I’m OK” part will be correct two-thirds of the time, but the implication that therefore smoking is ok is not.  The real implication, as I said, is that they are one of the lucky ones.


            Even the “I’m OK” part of this argument is often wrong.  In this context, “I’m OK” refers to being able to perceive that a problem results from spanking.  The idea that spanking could cause depression just doesn’t cross anyone’s mind.  However, there is now excellent research linking spanking with an increased probability of depression (DuRant, Getts, Cadenhead, Emans, and Woods 1995; Holmes and Robins 1987; Kessler and Magee 1994; Straus 2000a).  Of course, many other biological and social factors increase the risk of depression, so it would be equally wrong to say that someone suffering from depression must have been spanked a lot.

B2.  If CP Makes Things Worse, Why Does Misbehavior Decrease As Children Grow Older?

            The explanation for the seeming paradox that children who are spanked do, on average, improve their behavior over time lies in the fact that, as children grow older, so does their ability to understand, to reason, and to control their behavior.  Spanking and other trauma slows down development of these abilities, but except in very rare cases, not enough bring development to a halt.  The abilities of spanked children improve with age despite the spanking.  This is illustrated by the study in Chapter 10 showing that CP is associated with a slowing of mental development.  The fact that the children in the high CP group fell behind the cognitive development of the no-CP group by six points does not indicate that, after two years, the children in the high CP had less cognitive ability than at the start of the study or that they got dumber.   On the contrary, the children in the high CP group, like all normal children, increased their cognitive skills tremendously in those two years, but less than the children who were not spanked.  The same principle applies to rates of misbehavior.  On average, children whose parents spank, like all normal children, are much better behaved at age 4 or 6 than at age 2, but children whose parents spank, on average, fall behind the behavioral improvement of children who experienced little no CP.


            Actually, the process is more subtle than the previous two paragraphs suggest.  Children do learn from CP, especially when it is consistently applied, just as they learn from explanations, reasoning, and non-CP discipline when they are consistently used.  The difference is that the learning process is slower when parents use CP than when they rely on other teaching strategies.




C1.  Since CP Has Been Abolished In Schools, Discipline Has Declined And School Violence Is Up

            The idea that abolishing CP in schools has caused the behavior problems experienced in American schools is another fantasy in the minds of those favoring CP.   The research evidence shows the opposite.


In fact, CP continues to be used in schools in about half the states.   Although disciplinary problems may have increased in schools, it is because of many crime-enhancing changes in American society (see Blumstein, 200), not because CP is no longer used in half the states.  I have done two studies comparing the rates of school violence in states that do not use CP with states that authorize CP in schools under very limited conditions and those that authorize wide use of CP.  The first of these studies (Straus 2000a, p. 112) used data from the 1980s and found that the more CP was authorized, the higher the rate of violence and vandalism by students.  The second study (see Chapter 1) was done when the 1999 Columbine, Colorado school shootings, led some editorial writers and some members of state legislatures to call for a return to CP in the schools.  It too found that the more CP was authorized, the higher the rate of homicides by juveniles.  Both these studies can be interpreted as illustrating the effect of the modeling of violent behavior by persons in positions of moral authority.

C2.  Crime By Youth Is Up Because Parents Are Not Spanking

            The belief that crime has increased because parents no longer spank is based on the false premise that crime by youth is increasing.   In fact, crime by both adults and youth decreased rapidly since the mid 1990s (Blumstein, 2000).   In response, defenders of CP could point out that even the major decreases in the 1990s brought the rates back only to about where they were in the 1960s.  This, in turn can be countered by the fact that, a generation ago, violent crime rates were already so high, and Americans were so concerned about the high level of violence, that President Johnson appointed a national commission in 1968 to investigate the cause and to recommend steps to reduce the level of violence in the United States.  The importance of this bit of history is that in 1960s almost everyone believed in CP and most parents used CP, even with teen-agers (see Chapter 13).  So, in 1968, when CP was nearly universal, crime rates were just as high as they are now.


These fluctuations in crime rates result from a multiplicity of causes.  CP is one of those causes.  The research evidence in this book shows that the more CP, the greater the probability a child will be delinquent.   If so, the decreasing use of CP during this period contributed to decreasing rather increasing crime.  If that is the case, why did crime rates go up during part of this period?   It is because CP is only one of a large number of things that influence the crime rate and the crime reduction effect of less CP is not sufficient to overcome the many other crime-enhancing factors.  Here are some examples:


* The growth in unemployment, especially among poor inner-city youth and young men. A stable job may be the most important crime prevention known (Anderson 1999)


* The growth of illegal businesses, especially the crack epidemic which involved many youth, and provided them with guns (Blumstein, 2000)


* Decline in neighborhood cohesion and monitoring of children and lack of social ties and trust, crime increases (Sampson, Raudenbush, and Earls 1997)


* The increased exposure of children to the trauma of parental divorce and the personal and financial stress of bringing up children without the aid of a partner, as well as the personal loss experienced by children (Wells and Rankin 1991).


* The growth of teen births and births to poor unmarried mothers who, more often than other parents, lack the material and social resources to provide an adequate home environment (???)


* Poverty -- one out of five American children grow up in poverty.  Messner et al (2000) for example, found that the rate of homicides by children goes up when the percent of children in poverty increases and goes down when child poverty decreases


* Endless hours watching TV violence and practicing killing in the form of video games (Anderson and Dill, 2000)


* Public distrust of the police and failure of the police to keep public order (???)


            * The personal trauma and example of growing up in a family in which there is violence between the parents Geffner, Holden etc book, 19???> (Straus 1992)


* Other examples of incivility and self‑centeredness set by parents and adults (???)


* Adult violence.  The rate of violence by youth in a neighborhood, region, or nation mirrors the rate of adult violence (see Figure 14-3).   Consequently, ending juvenile violence means adults must set an example of non-violence.


(Insert Figure 14-3 about here ---Not yet ready)


The list can go on and on.  If CP was to be completely eliminated, and everything else remained the same, crime would be less frequent but the rate would still be high because all the other causes of crime remained.  Thus, a major decrease in the juvenile crime rates requires doing something about a large number of these risk factors for crime.  No one of them, including CP, is sufficient by itself to bring about a major change or to overcome increases in the others.  Consequently, in the period from the mid 1960s to the mid 1990s when so many of the risk factors were increasing, the reduction in CP was not nearly strong enough to outweigh the crime-promoting changes in American society during that period.

C3.  No-Spanking Means No-Discipline, Permissivenss,  And Children Running Wild

            If no-spanking meant no-discipline, the research evidence would favor spanking.  Children do need clear rules and expectations, consistency in enforcing the rules, and parents who do not tolerate disrespect for others.  But no-spanking does not mean no-discipline or “permissiveness” in the sense of tolerating misbehavior (Straus 2000a, pp 157-159).  Moreover, the research evidence in this book, together with the results from many other studies, clearly indicates that children whose parents use only non-violent modes of discipline are, on average, better behaved than the children of parents who spank (see Chapters ?? and ??).


Diana Baumrind is a leading child psychologist who tries to equate no-CP with “permissiveness,” i.e.,  no-discipline (Baumrind 1996), but she presents no research evidence.  She is right to be concerned about no-discipline.  But contrary to what Baumrind implies, researchers and organizations opposed to spanking also believe that rules and discipline are necessary.  One of these organizations even incorporates that idea in its name – the Discipline (see their web site:




D1.  Low Education Parents Lack the Cognitive And Linguistic Skills Needed To Avoid CP

            The idea that abolishing spanking would be harmful to children of low education and low income parents Larzelere (19??) is based on assumption that these parents lack the verbal skills needed to adequately control children by cognitive methods alone, or even enough space in the household to use strategies such as sending a child to his or her room.  The dubious nature of this argument is suggested by the results of research showing that parents in non-violent tribal societies do not use CP (Montague 1978).   Because these societies are also non-literate, it demonstrates that not even literacy is needed to bring up a child without hitting.


There is also a cruel irony in the idea of allowing use of CP by low education parents because, as shown in Chapters 10 and 11, CP slows down children’s mental development.   CP is also associated with a reduced chance of graduating from college.  And, among those who do graduate, CP is associated with a reduced chance of securing a well paying job.  Thus, as Holly Gimple and I concluded “The demand for workers in low level occupations who can adapt to the monotony of unskilled factory work is disappearing.  Jobs which require a strong back and obedience to authority are becoming so rare that men and women who, in a previous historic era, could have a stable place in society may find no place in the post-industrial labor market.  CP, which helped socialize previous generations of factory workers may now be helping to create the next generation of the chronically unemployed” (Straus and Gimpel 1994).  It is possible that children of low socioeconomic status parents will be the ones to gain the most from being brought up without spanking.

D2.  Spanking Is Necessary Because Toddlers Don’t Understand Explanations And Reasoning

            Contrary to the belief that “you can’t reason with a toddler,”  there are many studies showing that toddlers do reason, engage in moral reasoning, and have empathy for others, for example (Aronfreed 1976; Edwards 1980; Hetherington and Parke 1999; Hoffman 1984; Kochanska 1995; Kochanska and Thompson 1997; Parke 1977; Rheingold 1982).


In addition to these formal studies, almost every parent who has had a toddler tell them something is “not fair,” or had a toddler tell them "because he hit me" to explain why he is crying, knows that children do understand explanations and reasoning, even though imperfectly.  As this chapter was being written, for example, a heavy snow forced cancellation of 1st and 2nd grade classes for two of my grandchildren.  One went to spend the day with a friend and other went with his dad.  However, the 3-year-old’s preschool was not canceled.  He too wanted to go to a friend’s house or go with his dad.  He insisted on knowing why he could not.  His parents explained why he had to go to school -- in fact, several times.  The conversation left no doubt that Eric understood the reasons.  He also understood the complex notion of fairness and found that this was unfair.  And he was right!  However, understanding something and doing something are not the same.  Adults often fail to do what they know they should do, or do things they know they should not do.  This is even more true of children.  So, parents have to be in charge on things that matter.  In this case Eric’s parents, while repeating the explanation, never wavered from the fact that Eric was going to school that day.  This example also illustrates the combination of consistency and cognitively based strategies to influence and control the child, which may be part of the explanation for the fact that children who are not spanked, on average are better behaved (see Chapter ??) and experience faster cognitive development than other children (see Chapter 10).


            Finally, aside from the question of the rationality of toddlers, there are many non-violent aversive sanctions, as well as non-aversive methods of shaping the behavior of child that are equally effective in the short run and more effective in the long run (?? cite needed).


D3. “Never Spank” Is A Negative Approach And Won’t Work.  It Is Better To Use A Positive Approach And Teach Alternatives

            The idea that you can’t tell parents not to spank and instead must focus on teaching alternatives is a false choice.  Both are needed, but to my surprise, an exclusive focus on teaching alternatives and rejection of the idea that parents should be advised to never spank is something I frequently hear from child maltreatment scholars and parent educators, including some of the most distinguished.  They reject the idea of telling parents to never spank, and also no-spanking messages on milk cartons, on posters in pediatrician's offices, and a warning notice on birth certificates.  When I ask if they favor posters and warning notices about cigarettes, the answer is almost always yes.  They typically go on to explain that a “negative approach” will not succeed for spanking because parents must first be taught alternatives.  There are at least three reasons for thinking this is a specious argument.  


* Parent education and no-spanking education are not mutually exclusive.  Parents can be informed about alternatives at the same time as they are informed about the long-term counter-productive effect of CP and that CP is a risk factor for many psychological and social problem behaviors.


* The belief that training in alternatives is a precondition of advising no-CP implies that parents do not already know and use many other disciplinary strategies.  Almost all parents know and use a wide variety of strategies.   They may not use them consistently enough or skillfully, and learning additional strategies would be a benefit.  Nevertheless, a typical parent already uses a large number of alternatives.  The research evidence indicates that those parents would be even more effective if, as a result of a never-spank message, they left out the spanking part of their approach to discipline.


* The belief that parents must first be taught alternatives presumes that CP has greater efficacy than other discipline techniques and that we should not deprive parents of this tool until we can give them something that is equally effective.  This presumption has been shown to be incorrect, even in research by defenders of CP.  Those studies show that non-corporal methods are just as effective (Larzelere et al. 1998; Roberts 1988).


            Although almost every parent can use additional skills in child management, there is no research evidence that it takes such training to stop spanking.  Consistent with the cultural norms supporting CP, it is just presumed to be true.  The same presumption is not applied to whether prior training is needed before advising parents to never engage in psychological attacks on children.  It has not stood in the way of a “negative approach” to ending psychological aggression as a means of discipline.  We do not have misgivings about telling parents to never call a child  “a filthy pig” if the child spills food because we truly believe that psychological attacks are a harmful mode of discipline.  We apparently lack the same certainty that physical attacks, such as slapping the hand of a child who spills food, is bad for children.  The National Committee for Prevention of Child Abuse conducted an excellent media campaign on psychological attacks by parents that used a “negative” approach.”   It featured a poster with the face of sad child and the message “Stop using words that hurt.”   But the NCPCA has never feature a poster of a sad faced child, which says “Stop spanking.  Other methods work better.


D4  Spanking Is Universal And Can't Be Dealt With By Policies Or Laws

            There are a few societies that do not use CP and they are also non-violent in other ways (Frey, 199??; (Montague 1978).  However, even if spanking were universal, if it is harmful it should be the object of public policies to reduce or eliminate it.  That is the approach we take with other harmful universals such as jealously or murder.  We do not use the fact that these are universal to say that nothing should be done about them.  The difference between spanking and murder is not that one is a universal and the other is not, but that the presumed benefit of spanking “when necessary” is a deeply embedded aspect of American culture.  Consequently, almost all Americans, including social scientists, doubt the wisdom of never hitting a child.  Indeed, this belief is so deeply embedded that they ignore the experimental and other evidence showing that spanking on average is no more effective than non-violent modes of discipline; or when the evidence cannot be ignored, a double standard is used to evaluate that evidence (see Chapter 1).




E1.  The Real Problem Is Bad Parenting

            The idea that the real problem is a general pattern of “bad parenting” by parents who use CP is based on the incorrect assumption that only bad or incompetent parents spank.  If that were the case, 94% of American parents would fall into the bad-parenting category because this is the percent who spank toddlers (see Chapter 2).


            In addition to the factual and logical error, the “bad-parenting” defense has not held up under empirical testing.  A number of studies have found that, even after taking into account different aspects of the quality of parental care,  CP is linked to child behavior problems.  For example, the first study listed in the following table controlled for the degree to which parents were attached or close to the boys in the study.  Even after taking that into account, Brezina found that the more CP used by parents, the greater the probability of the child hitting the parent during the one-year period covered by the study.


Study                                      Problem Linked To CP          Controlled For

Brezina (1999)                         Assault Of Parents                   Parental attachment

Durant (1995)                                      Depression                              Family conflict

Gunnoe & Mariner (1997)                    Antisocial Behavior                  Praise, Rules

Larzelere (1986)                                  Aggression                              Explanation

McCord                                                Conviction for Serious Parental warmth


Simons et al (1998)                             Assaults On Dating Partners   Parental involvement

Straus et al, Chapter 5                        Antisocial Behavior                  Emotional support

                                                                                                            Cognitive stimulation

Straus & Paschall Chapter 10  Slow Cognitive Development   Emotional support,

Cognitive stimulation

Straus & Donnelly (1994)                    Masochistic Sex                       Warmth, Monitoring, Consistency, Reasoning

Turner & Finkelhor  (1996)                              Psych Distress             Parental support             


All of the studies listed above found that, although controlling for quality of parental care lessens the risk of harm from CP,  CP remains related to child behavior problems.  The same finding applies to many other plausible alternative explanations.  See section

F 4 on "Inadequate Controls For Other Variables."

E2.  The Important Issue Is The Overall Pattern Of Parent Behavior

            Defending CP by arguing that only the larger pattern of parental behavior makes a difference has both a conceptual and a statistical basis.  The conceptual basis is the idea that CP is just one of many components in an overall variable identified by names such as harsh or inadequate parenting.  One of the reasons "just a symptom of bad parenting" is not likely to be true, is that over 90% use corporal punishment with toddlers.  No one knows what percent of parents are harsh or incompetent, but 90% is not plausible.  So, almost by definition, there must be sizable number of good, loving, parents who spank.


            The statistical bases for this defense of CP are factor analyses that reveal that CP is part of a more general harsh discipline factor.  The factor score or an additive scale using items in the factor is used to create a scale to measure harsh discipline.  Paradoxically, this prevents finding out about corporal punishment per se.  In a ten-item scale with an alpha of .75, the average correlation of any one item with an overall scale score is about .25 (???).  Thus, only 5% of the variance in the corporal punishment item is explained by the scale score.  Consequently, CP per se could have antecedents and consequences that are different from the overall scale.  Using a harsh parenting factor scale score prevents investigating whether spanking by parents who are not otherwise harsh, i.e. CP by warm and loving parents, has harmful side effects, and the interaction of parental warmth and CP (Straus and Donnelly 1994).  Moreover, “harsh discipline” scales often confound ordinary CP with discipline tactics that are more appropriately conceptualized as "abuse"  (see for example the Maternal Punishment scale in McVoy (???),


            This is not to say that a summary scale of harsh parenting is wrong.  It permits investigating the assumption that what matters for the child is the overall pattern.  This is clearly a plausible and important approach.  However, it does not negate the value of also investigating specific components of the factor such as corporal punishment.   Factor analysis, for example, is widely used to create scales to measure socioeconomic status.  The difference is that specialists in social stratification insist on the importance of also analyzing each of the specific components of a socioeconomic status scale.  Being in a high or low status occupation may have different effects from high or low income even though both have a significant “factor loading” on an overall socioeconomic status factor, just as corporal punishment has a significant loading on an overall harsh parenting factor.  Similarly, in research on cognitive stimulation by parents, it is important to investigate both the effect of specific behaviors such as reading to a child, and also the combined effect of many modes of cognitive stimulation.  In short, the overall pattern of parent behavior is extremely important but it is not an alternative to finding out about the main effects and interactive effects of each part of the overall pattern, including CP.  Both approaches are needed. 

E3. The Link Between Spanking And Mental Health Problems Is Genetic

Mental health problems, like almost all human behavior, are the result of the interaction of genetic characteristics and life experiences.  However, for genetics to explain the findings on the mental health problems associated with CP, one has to assume that parents who spank have genetic predispositions to the psychological and social problems associated with CP, and that these predispositions also cause them to spank.  For example, parents who spank might be genetically disposed to aggressiveness, depression and antisocial behavior.  Given the fact that 94% of parents spank, the idea that this reflects innate psychological problems is unlikely.  Rather than reflecting psychological problems, the high rate of CP reflects the combination of cultural norms permitting and sometimes requiring CP and the fact that controlling and training children is a difficult and often frustrating task.   Moreover, even if spanking is associated with a genetic predisposition for some types of problem behavior, such as aggressiveness, that does not mean that learned behavior and reactions are unimportant.  Both genetic and experiential factors can make their own contributions.


            There is also the possibility that the better behavior and faster cognitive development of children whose parents never spank has a genetic component.   The 94% spanking means that, at best, that only 6% of parents never spank.   That 6% could be genetically superior.   It may take special inborn qualities to buck the cultural norms and the advice from friends and relatives to spank (Carson, 19??,  Walsh, 19??).  However, those special qualities could just as plausibly be socially transmitted. 

E4. Overlap Of Corporal Punishment And Physical Abuse

            The confounding or overlap of CP with physical abuse is an important problem because almost every parent who kicks or punches a child also engages in legal forms of hitting children such as spanking and slapping.  Consequently, what shows up as an effect of corporal punishment might really be due to unknowingly including children in the sample who were physically abused.  Fortunately, there are studies that avoided this problem by removing from the sample children whose parents exceeded ordinary corporal punishment.  These studies, which were restricted to children of parents who did engage in “physical abuse” found harmful side effects for corporal punishment (see chapters 6, 7, and 8 and (Chapters 5, 6, and 7 MacMillan, Boyle, Wong, Duku, Fleming, and Walsh 1999).


Other studies have isolated the effect of CP per se by categorizing parents into three types:  those who used entirely non-violent discipline strategies, those who used CP but did not attack the child more severely, and those whose attacks were severe enough to be considered physical abuse.   These studies find that the effect of CP is parallel to the effect of physical abuse.   The main difference is that the “risk ratio” (the probability of the harmful effect occurring) is lower for CP than for physical abuse (Strassberg, Dodge, Petit, and Bates 1994); Ulman & Straus 2000 ??).




F1.  CP Is Not Harmful If It Is Perceived As A Legitimate Exercise Of Parental Authority

            Gunnoe and Mariner (1997) defend CP by arguing that "…it is not spanking per se but rather the context in which spanking occurs and the meaning that children ascribe to spanking that predicts child outcomes."   They suggest that, f the family and the cultural context lead children to accept the legitimacy of CP, spanking will have no harmful effect.  This “cultural relativity” argument is used by African American defenders of CP such as Polite (1996).  I have been able to locate only two studies that provide direct empirical tests, as compared to offering it as a plausible explanation.  One of those studies was of Afro-Caribbean children in St. Kitts (Rohner, Kean, and Cournoyer 1991).  They found a significant direct relation between CP and psychological adjustment problems.  They did not find a significant mediating effect for children's perception of the legitimacy of CP.  However, the child's perception of CP as indicating rejection by the parents mediated the effect of CP on the child's adjustment.  A second study (Rohner, Bourque, and Elordi 1996) controlled for the child's perception of the justness of the CP they experienced and still found that CP was related to feelings of rejection, which in turn was related to the child's adjustment.  The fact that the relation of CP to maladjustment was mediated by the child's perception of parental rejection provides information on one of the intervening processes that explain the link between CP and child behavior problems.  Since the child's perception of rejection is itself undesirable, rather than disproving the harmful effects of CP, it adds to the evidence.

F2.  CP Is Appropriate In The Context Of African American Culture And Life Circumstance

Some African American social scientists (e.g. Polite 1996), and some who defend CP on religious grounds (e.g.  Larzelere 1993; Larzelere 1994) use the idea of cultural relativity described in the previous paragraph to argue that efforts to end spanking will undermine the effectiveness of African American parents.  Moreover, because so many African Americans believe in CP (see Chapter 12 and (Flynn 1996; Straus and Mathur 1996), they argue that no-spanking advice amounts to an unethical imposition of cultural beliefs and values. 


The empirical evidence is not that clear.  Several studies found that CP is linked to social and psychological problems among African American children (Rohner, Bourque and Elordi 1996; Rohner, Kean and Cournoyer 1991; Straus and Lauer 1992; Straus, Sugarman and Giles-Sims 1997).  One study found a harmful effect for one of the two outcome variables studied, and a positive effect for a second outcome (Gunnoe and Mariner 1997).  Another study (Deater-Deckard, Dodge, Bates, and Pettit 1996) found no relation between CP and behavior problems among African American children.


A plausible interpretation of the two studies that found an interaction of CP with race is that the effect of CP is moderated by the cultural legitimacy of CP in the African American context.  However, there is an alternative plausible interpretation starts from a closely related assumption, which is that in the context of African American culture and life circumstances, CP is virtually synonymous with discipline.  In that context, no-spanking may mean no-discipline.  If that is correct, in view of the research on the importance of parental monitoring and control for prevention of delinquency (Patterson, DeBaryshe, and Ramsey 1989; Sampson and Laub 1993b), it is no wonder that CP has no harmful effect or a positive effect among a group where no-CP tends to mean no-discipline.  This issue urgently needs further research.  Such research could include data to test both the no-spanking means no-discipline explanation and the cultural legitimacy explanation.  In the meantime, it is a sufficiently challenging goal to inform Euro-American parents of the benefits to themselves and their children of avoiding CP.





            As explained in Chapter 1, criticism of the methods used to obtain evidence in support of a theory is a typical and essential part of all fields of science.  Although a critical examination of methods is a standard first reaction to any new evidence, it takes on added importance for proponents of alternative theories who hope to indirectly bolster their position by attacking the evidence for an alternative theory. 

G1.  Spanking Is A Response To Misbehavior Rather Than A Cause Of Misbehavior

            The idea that spanking is usually a response to misbehavior is correct.  However, that does not mean that spanking cannot also be a cause of misbehavior.  Spanking may stop misbehavior in the immediate situation, but studies that have followed up large samples of children show that CP has the long-term effect of increasing the probability of misbehavior, as well as psychological and social problems.  For example, the study in Chapter 5 of over 3,000 children shows that the more CP used during the first year of the study, the greater the tendency for Antisocial Behavior to have increased two years later.  The same tendency for CP to boomerang was also found by three other recent prospective studies summarized in the 2nd edition of Beating The Devil Out Of Them (Brezina 1999; Gunnoe and Mariner 1997; Simons, Lin and Gordon 1998; Straus and Paschall 1999; Straus, Sugarman and Giles-Sims 1997).  All of them controlled for the presence of the behavior problem at the time of the spanking and found that the spanking, on average, was associated with a change for the worse two or more years later. 

G2.   CP By Itself Has Only A Small Effect And Is Not Worth Worrying About

If, as indicated in the previous paragraph, ending CP by itself has only a small effect in reducing the probability of psychological and social problems, a key question is whether advocating no-spanking is worth the effort.  Physical “abuse,” for example, is much more likely to cause serious psychological and social problems than spanking.  Perhaps it would be better to focus on combating “physical abuse?”   There is now a consensus among leading statisticians and researchers that “small effect sizes”  as conventionally measured, are often extremely important (and McCartney, 19??;  19??).  Here are two of the reasons why it is extremely important to end ordinary “mild” CP.


            Ending CP Is Part of Ending Physical Abuse.  About 2/3 of cases of physical abuse known to child protective services are the result of CP that has escalated out of control [Kadushin, 1981 #4062; Straus, 1994 #4891].  Thus, ending CP and preventing physical abuse are part of the same effort.  Thirty years ago, David Gil, one of the pioneers in research on physical abuse wrote: "Since culturally determined permissive attitudes toward the use of physical force in child-rearing seem to constitute the common core of all physical abuse of children in American society, systematic educational efforts aimed at gradually changing this particular aspect of the prevailing child-rearing philosophy, and developing clear-cut cultural prohibitions and legal sanctions against the use of physical force as a means for rearing children, are likely to produce over time the strongest possible reduction of the incidence and prevalence of physical abuse of children."  (Gil 1970, p.141).


Cumulative Effect Is Large.  It is a well established principle in public health that a risk factor that increases the probability of a health problem by only a small amount (such as spanking) can have a much greater impact on public health if it applies to a large part of the population than a risk factor that, if present, increases the probability of a harmful effect by a large amount (such as physical abuse) but which applies to a small number of people (Rose 1985; Rosenthal 1984).   The following hypothetical example shows that ending CP will do more to reducing the prevalence of depression than ending physical abuse.


* Epidemiological studies suggest that  in any one year, 50 million US children experience CP and 1 million experience physical abuse.

* Assume that the probability of being depressed as an adult is increased 25% for children who are physical abused and by only 2% for children who experience CP.


* Based on these assumptions, if physical abuse were to be ended, the reduction in cases of depression would be 0.25 times 1 million or 250,000.   If CP were to be ended, the reduction in cases of depression would be 0.02 times 50 million, or 1 million. Thus ending CP would be associated with a four times greater decrease.

G3. There Is No Evidence That “Moderate” CP, Used Only Rarely, Has Harmful Effects

            At least two studies investigated the effect of mild corporal punishment, used only rarely.  Both these studies were of large and representative samples of children,  Both also controlled for other family and child characteristics.  One is the study in Chapter 5 of antisocial behavior and impulsiveness of almost a thousand children age 2 to 14.  It compared children who were never spanked with children spanked only once in the past six months.  Even children who were spanked only once in the past six months had slightly higher impulsiveness and antisocial behavior scores compared to the never-spanked children.


The study in Chapter 10 of the relation of CP to the cognitive development of 1,506 children aged 2-4 and 5 to 9 measured CP in two sample weeks a year apart.  Only 6.7% had not experienced CP in either of two sample weeks.  These unspanked children experienced the most rapid cognitive development.  The cognitive development of children who experienced even one instance of CP in those two sample weeks was less rapid.  Thus, even one instance of CP in two weeks one year apart inhibited cognitive development, although by only a small amount.


G4. We Should Wait For More Conclusive Research Before Advising Parents To Never-Spank

The idea that absolutely conclusive evidence of harmful side effects is needed before it is ethical to advise parent to never spank (Larzelere, Baumrind, and Polite 1998) is contradicted by standard clinical practice in pediatrics.  Pediatricians advise parents to avoid a drug if there is evidence that the drug they are using is likely to have harmful side effects when an equally effective drug becomes available that does not have those side effects.  CP is like the old drug.  Consequently, the abundance of evidence indicating that CP has many harmful side effects in this book and in (Straus 2000a) in combination with the definitive evidence that other discipline responses are just as effective in the short-run requires advising parents to not spank.


Also relevant is the principle that, if a procedure is shown to have a toxic effect under some circumstances, the burden of proof shifts.  When that is the case, it becomes the obligation of those who favor the procedure under other circumstances to show that it is safe and effective under those circumstances.   As noted earlier, advocates of CP have retreated to recommending CP only for younger children and explicitly warn against using CP with older children.  Thus, having accepted the evidence that CP is harmful for older children, they have the obligation to provide empirical evidence that it is safe for younger children.  The so-called "beneficial effects" of CP with younger children emphasized by (Larzelere 1996) does not provide that evidence because those effects refer to cessation of misbehavior, not to long-term harmful side effects.  Moreover, as noted earlier and in more detail below, non-corporal strategies to stop misbehavior are equally beneficial in securing short-term compliance.

G5.  If All Children Are Spanked, It Is A Constant and Cannot Explain Anything

Although a variable that does not vary, of course, cannot co-vary with some other variable, that problem does not apply to research on CP for two reasons.  First, despite the fact that  almost all children in the US are spanked as toddlers (94%), there is still great variation in how often this occurs, in the severity of the blows, and in the number of years that parents continue to spank.  Thus CP is a parent behavior which varies greatly.


Second, even when examining the question of the effect of never spanking, although the percentage is very small, there are enough parents who truly never use CP to demonstrate a relationship.  Suppose only one or two percent of parents truly never spank.   With a large enough sample, that is sufficient variance to determine if there is covariance with some presumed harmful or beneficial effects of absolutely no-spanking.  For example, the study in Chapter 6 identified 189 children, who, at least according to the mothers, had never been spanked.  These children had the lowest average antisocial behavior scores and were the least impulsive, even compared to children who were very rarely spanked.  Another study that identified a never-spanked group (MacMillan et al. 1999) found that they had the lowest rate of psychiatric symptoms, even compared to those who were spanked only rarely.

G6.  Inadequate Controls For Other Variables

            Statistical controls for characteristics of the child and the parents are needed because a link between corporal punishment and child behavior problems might be the result of other family and parent characteristics that lead to both corporal punishment and child behavior problems.  For example, parents who physically attack each other are more likely to also hit their children (Ross 1996; Straus, Gelles, and Steinmetz 1980) and what seems to be the effect of CP might really be the effect  of the violence of the parents to each other.  To deal with this problem, the empirical studies in this book and other cross-sectional studies controlled for many variables that could be the “real cause” of the problems linked to corporal punishment.  These controls include:


            Social Characteristics of the Parents and Family

            Educational level of parents

            Income, including very low income

            Racial/ethnic group

            Single parent versus two parent families

Number of children in the family

            Sex of the parent

            Age of the parent


            Parental Role Behaviors

            Adequacy of parent’s supervision of children

            Parental warmth and support

            Whether parents established clear rules and expectations

            Use of other disciplinary strategies such as time out

            Parental consistency in discipline

            Parental use of reasoning

            Parental involvement and cognitive stimulation


            Child Characteristics

            Child’s birth-weight

            Sex of the child

            Age of the child

            Child’s delinquency or antisocial behavior at Time 1

            Child’s cognitive ability at Time 1

            Child-to-parent bond


            Psycho-social Problems

Conflict between the parents

            Violence between the parents

            Violence in family in which the parents grew up

            Parental alcohol abuse

            Parent attitudes approving violence

            Whether the parent also engaged in more severe violence (“physical abuse”)

            Depression of parents


            No single study controlled for all of these possible confounds.  However, it is a well recognized scientific principle (sometimes called “triangulation Webb, Campbell, Schwartz, Sechrest, and Belew Grove 1981) that valid conclusions are possible on the basis of cumulative evidence from studies which, taken one by one, are not definitive.  This is because the weak point of one study may be dealt with in another study.  I think we have reached the point of triangulation concerning corporal punishment.  There have been more than 80 studies examining the effects of corporal punishment, and with rare exception, they have found harmful long-term effects (Thompson In press).



G7. Biased Perception Explains the Findings

            Biased perception could result in finding a relationship between CP and psychological problems when, for example, the data on both CP a child behavior problem are both obtained from a parent, or both obtained from a child.  This could happen because parents or children who are willing to report spanking may be more likely than other parents to also report behavior problems.  However, the fact that almost all parents of toddlers in the 1995 national survey reported spanking in the past year, and other studies find that about 70% spanked in the last week, and did so an average of about three times that week (see Chapter ??), shows that spanking is not something that which takes a biased perception of report.


More direct evidence of the link between CP and behavior problems comes from studies that measured the child behavior problem independently, such as by observing children interacting with other children (Strassberg, Dodge, Petit and Bates 1994), by using officially recorded delinquency and crime (McCord, 1991’; 19??), by child-report data on aggression (Simons, Lin and Gordon 1998), and by administering objective tests to children (see Chapter 9).

G8.  The Effect Size Is Too Small To Make A Difference  

The importance of ending CP has been challenged by arguing that the correlations between CP and harmful side effects are so low that ending CP would not make an important contribution to child well-being (Larzelere, Baumrind and Polite 1998).  This defense of CP ignores the fact that a small effect size is typical of the relation of parent behavior to child behavior.  In fact, it could not be otherwise because, given the multiple influences on a child, no single risk factor will account for a large percent of the variance in a dependent variable such as delinquency, depression. For example, parents who were abused have a much greater probability of abusing or neglecting their own children, but two thirds of abused children do not continue that pattern with their own children (Kaufman and Zigler 1987).  Similarly, two thirds of pack-a-day and over smokers do not die of smoking related diseases (Matteson, Pollack and Cullen 1987).  The fact that smoking does not explain two thirds of the cases of lung cancer does not invalidate the conclusion that smoking causes lung cancer.  Similarly, if 90% of spanked children do not manifest harmful effects from spanking, that does not invalidate the conclusion that spanking causes behavior problems.  In addition, the fact that the probability of growing up to be an abusing parent is greatest for those who have been chronically abused, the fact that lung cancer is greatest among heavy smokers, and that the probability of behavior problems is greatest with high frequency spanking, does not mean that occasional physical abuse, occasional smoking, or occasional spanking is harmless.  Perhaps most important is the well established epidemiological principle that elimination of a risk factor with a small effect size for a widespread behavior can have a greater impact on health than elimination of a risk factor that affects a small part of the population (Cohen 1996; Rose 1985; Rosenthal 1984).  Corporal punishment and physical abuse are both risk factors for anti-social behavior.   But elimination of CP could make a greater contribution to lowering the prevalence of depression because it affects such a large proportion of the population.  (For a specific example, see section F7.)


G9. Even Prospective Research Does Not Prove A Causal Relation

Defenders of CP used to be able to argue that the research seeming to show harmful effects is not worth attention because it is all cross-sectional.  They must now retreat to pointing out that even prospective studies do not prove a causal relationship (??).  True enough, but when there are experimental studies, defenders of CP will probably shift from insisting on experimental evidence to pointing out that experiments can also be erroneous.


To evaluate this defense of CP one has to understand that methodologists and philosophers of science generally regard causal inference as a judgment with varying degrees of confidence.  These judgments are on a continuum.   At the weak end of the continuum are inferences from cross sectional data.  If there are no statistical controls for possible confounds, one can only say from a cross sectional study that the findings did not provide data contrary to a causal theory.  This was the limitation of most of the early research on CP.  If there are controls for variables representing key alternative explanations, at least some of the plausible rival interpretations will have been ruled out and one can therefore have greater confidence that there is a causal relationship, but hardly proof.   That was the situation up to 1997 when the first of the recent prospective studies was published.   Those prospective studies took “temporal order” into account, i.e., what came first, and that greatly increases the degree of confidence that CP causes the subsequent behavior problems examined by these studies (see for example, Chapters ?? and ??).  An experiment permits the greatest confidence.  There have been some experimental studies of spanking as a back-up for breaking time-out.  They show that non-corporal methods of enforcing the time-out are just as effective as CP.   Unfortunately, there has not been an experiment on the long-term effects of CP.  Such an experiment is possible and ethical.  It can be done by including a no-spanking component for a random half of parents in a parent-education program.  If it found that the no-spanking component led to fewer parents spanking and that their children were, on average, better behaved and had fewer psychological problems, it would provide the most definitive evidence of the benefits of ending CP.   However, such a conclusions would still be an inference, not certainty, because artifacts can produce erroneous findings in experiments as in other types of research.


The attempt to defend CP by insisting on experiments is an example of scientific absolutism that is inconsistent with the principle that causal inference is a judgment with varying degrees of confidence, not something that can only be achieved in one way.  Much of modern science would have to be disregarded if only random assignment experiments were valid.

G10. The Results May Be Due To Recall Bias

            Recall bias could produce erroneous results if people who have behavior problems are more likely to recall CP.  This is unlikely because numerous studies found that 80 to 95% of adults recall CP (e.g., Bryan and Freed 1982; MacMillan et al. 1999; Straus 2000a).  This means that adults who recall CP are not restricted to the much smaller proportion of the population who experience behavior problems such as depression and physical aggression.


            More direct evidence against the recall bias argument comes from the five recent studies reviewed in the concluding Chapter of the 2nd Edition of Beating The Devil Out Of Them.  All five of these studies used data provided by either the parent or the child at Time 1 of the study.  Thus, the most definitive studies of spanking used data that are not confounded with recall bias.



H1, The Bible Tells Parents To Spank

            Some opponents of spanking believe that the biblical reference to “the rod” refers to a shepherd’s rod that is used to guide sheep, not to hit them.   My less-than-expert knowledge of the Old Testament leads me to believe that it does refer to CP.  In fact, the Old Testament goes much further than endorsing hitting children with a stick.  Leviticus 20.19 says "All who curse father or mother shall be put to death."  Deuteronomy 22.12 says "This son of ours is stubborn and rebellious.  He will not obey us.  He is a glutton and a drunkard.  Then all men of the town shall stone him to death."


Parents or religious leaders who justify their use of CP on the basis of believing in a literal interpretation of the Bible obviously engage in selective literalism because they do not advocate executing stubborn, rebellious, or gluttonous children.  Moreover, they seem to ignore the New Testament.  Jesus was committed to love and forgiveness as both the means and the end of human relationships.  It is hard to imagine Jesus recommending hitting a child as a means of correction.

H2.  We Should Focus on More Crucial Threats To Children Such As Poverty And Racism

Defending CP by pointing out that there are worse things that can happen to children is analogues to saying that we should focus on heart disease and not worry about broken legs because more people die of heart disease.  Actually, it is even more illogical because in the case of medical services and medical research, different health problems compete for the same set of resources.  But in the case of CP, pediatricians, psychologists, and parent educators do not even have the alternative of using their resources to eliminate poverty and racism.  Moreover, advising parents to not spank takes minimal resources, of which the most important is to bring it up when talking to parents.  There are also many other inexpensive and practical ways of getting a no-spanking message to parents, such as a no-spanking message on milk cartons, and “never spank” posters and leaflets in maternity wards and physician waiting rooms. These methods have worked in Sweden (Durrant 1999).   I think they will also work in the US. I suspect that a major reason for not using these approaches doing so is doubt that never-spank is good advice.




            This chapter examined a large number of defenses of CP and has shown that none of them stand up to scrutiny.   Nevertheless, ending CP faces many obstacles (see Chapter 1 and (Straus 2000a, Chapters 1 and 10).   I will conclude by illustrating the obstacles with one that is inherent in the behavior of toddlers.


            When toddlers are corrected for a misbehavior (such as hitting another child or disobeying), the “recidivism” rate is about 80% within the same day and about 50% within two hours.  For some children, and for almost all children some of the time, it is within two minutes (Larzelere et al. 1998; Larzelere, Schneider, Larson and Pike 1996).  Moreover, Larzelere found that these “time to failure” rates apply equally to corporal punishment and to other disciplinary strategies.  Consequently, on any given day, a parent is almost certain to find that so-called alternative disciplinary strategies such as explaining, deprivation of privileges and time out, “do not work.”   When that happens, they turn to spanking.   So, as pointed out several times, just about everyone (at least 94%) spanks toddlers.


            The difference between spanking and other disciplinary strategies is that, when spanking does not work, parents do not question its effectiveness.  The idea that spanking works when other methods do not is so ingrained in American culture that, when the child repeats the misbehavior an hour or two later (or sometimes a few minutes later) parents fail to perceive that spanking has the same high failure rate as other modes of discipline.  So they spank again, and for as many times as it takes to ultimately secure compliance.   Persistence in spanking despite the inevitable repeated failure is graphically illustrated in the study by (Bean and Roberts 1981) of parents who used spanking to secure compliance with the child remaining in “time out.”  The average number of spankings was 8.3 and the median was 3.5.  The median session lasted 22 minutes.   Thus, the children in this group were spanked once every 3 minutes until the child did comply.  


What needs to be realized is that it is equally necessary to repeat non-spanking modes of discipline, such as just placing the child back on the time out chair.   As in the case of spanking, it needs to be done over and over again until the child learns.  In short, persistence and consistency are critically important, but are often missing from everything except spanking.  There is, however, one tremendously important difference between repeating spanking and repeating non-violent modes of correction and control.   Non-violent methods do not have the harmful side effects of CP demonstrated in this book.   But to avoid spanking and its side-effects, it is critical for parents to be aware of the high short-term failure rates of all modes of correction and control, and to be committed to never spank when those failures occur.  Until that information and that commitment becomes part of the everyday perspective of parents, parents will continue to spank, including parents who “do not believe in spanking,” and American children will continue to suffer the harmful side-effects.








Anderson, Eliajh. 1999. Code of the street: Decency, violence, and the moral life of the inner city. New York: W.W. Norton.

Aronfreed, J. 1976. “Moral development from the standpoint of a general psychological theory.” in Moral development and behavior, edited by T. Lickona. New York: Holt.

Associated Press. 1997. “Runaway is slapped; mother faces trial.” in The New York Times. New York, New York.

Baumrind, Diana. 1996. “Parenting.  The Discipline of controversy revisited.” Family Relations 45:405-414.

Bean, Arthur W. and Mark W. Roberts. 1981. “The effect of time-out release contingencies on changes in child non-compliance.” Journal of Abnormal Child Psychology 91:95-105.

Bernal, ME., JS. Duryee, HL. Pruett, and BJ Burns. 1968. “Behavior modifications and the brat syndrome.” J Consult Clin Psychol 32:447-455.

Brezina, Timothy. 1999. “Teenage violence toward parents as an adaptation to family strain: Evidence from a national survey of male adolescents.” Youth & Society 30:416-444.

Bryan, Janice W. and Florence W. Freed. 1982. “Corporal punishment: Normative data and sociological and psychological correlates in a community college population.” Journal of Youth and Adolescence 11:77-87.

Cohen, Patricia. 1996. “Response:  How can generative theories of the effects of punishment be tested?” Pediatrics.  The short- and long-term consequences of corporal punishment (supplement) 98:834- 837.

Coopersmith, S. 1967. The antecedents of self-esteem. San Francisco, CA: WH Freeman & Co.

Day, Dan E.  and Mark W. Roberts. 1983. “An analysis of the physical punishment component of a parent training program.” Journal of Abnormal Child Psychology 11:141-152.

Deater-Deckard, Kirby, Kenneth A. Dodge, John E. Bates, and Gregory S. Pettit. 1996. “Physical discipline among African American and European American mothers: Links to children's externalizing behaviors.” Developmental Psychology 32:1065-1072.

DuRant, Richard H., Alan Getts, Chris Cadenhead, and S. Jean Emans. 1995. “Exposure to violence and victimization and depression, hopelessness, and purpose in life among adolescents living in and around public housing.” Journal of Deviant Behavior in Pediatrics 16:233-237.

DuRant, Robert H., Alan Getts, Chris Cadenhead, S Jean Emans, and Elizabeth R. Woods. 1995. “Exposure to violence and victimization and depression, hopelessness, and purpose in life among adolescents living in and around public housing.” Developmental and Behavioral Pediatrics 16:233-237.

Durrant, Joan E. 1999. “Evaluating the success of Sweden's corporal punishment ban.” Child Abuse & Neglect 23:435-448.

Edwards, C.P. 1980. “The comparative study of the development of moral judgment and reasoning.” in Handbook of cross-cultural human development, edited by R. L. Munroe, R. Munroe, and B. B. Whiting. New York: Garland.

Flynn, Clifton P. 1996. “Normative support for corporal punishment: Attitudes, correlates, and implications.” Child Abuse & Neglect :47-55.

Friedman, Stanford B., Kenneth Schonberg, and Mary Sharkey. 1996. “Introduction.” Pediatrics.  The short- and long-term consequences of corporal punishment 98:VI.

Gil, David G. 1970. Violence against children: Physical child abuse in the United States. Cambridge, Massachusetts: Harvard University Press.

Gunnoe, Marjorie Linder and Carrie Lea Mariner. 1997. “Toward a developmental- contextual model of the effects of parental spanking on children's aggression.” Archives in Pediatric Adolescent Medicine 151:768-775.

Hetherington, E. Marvis and Ross D. Parke. 1999. Child psychology: A contemporary viewpoint. New York: McGraw-Hill.

Hoffman, M.L. 1984. “Empathy, its limitations, and its role in a comprehensive moral theory.” in Morality, moral behavior and moral development, edited by W. M. Kurtines and J. L. Gewirtz. New York: Wiley.

Holmes, Snadra J.   and Lee N. Robins. 1987. “The influence of childhood disciplinary experiences on the development of alcoholism and depression.” Journal of Child Psychology and Psychiatry 28:399-415.

Kaufman, Joan   and Edward Zigler. 1987. “Do abused children become abusive parents.” American Journal of Orthopsychiatry 57:186-192.

Kessler, R. and W.J. Magee. 1994. “Childhood family violence and adult recurrent depression.” Journal of Health and Social Behavior 35:13-27.

Kochanska, G. 1995. “Children's temperament, mother's discipline, and security of attachment:  Multiple pathways to emerging internalization.” Child Development 66:597-615.

Kochanska, G. and R.A. Thompson. 1997. “The emergence and development of conscience in toddlerhood and early childhood.” Pp. 53-77 in Parenting and children's internalization of values, edited by J. E. Grusec and L. Kuczynski. New York: Wiley.

Larzelere, Robert E. 1986. “Moderate spanking:  Model or deterrent of children's aggression in the family?” Journal of Family Violence 1:27-36.

Larzelere, Robert E. 1993. “Response to oosterhuis: empirically justified uses of spanking: Toward a discriminating view of corporal punishment.” Journal of psychology and theology 21:142-147.

Larzelere, Robert E. 1994. “Should the use of corporal punishment by parents be considered child abuse? - No.” Pp. 204-209 in Debating Children's Lives: Current Controversies on Children and Adolescents, edited by M. Mason and E. Gambrill. Thousand Oaks, California: Sage Publications.

Larzelere, Robert E. 1996. “A review of the outcomes of parental use of nonabusive or customary physical punishment.” Pediatrics 98:824-831.

Larzelere, Robert E., Diana Baumrind, and Kenneth Polite. 1998. “The pediatric forum: Two emerging perspectives of parental spanking from two 1996 conferences.” Archives of Pediatrics and Adolescent Medicine 152:303.

Larzelere, Robert E., Paul R. Sather, William N. Schneider, David B. Larson, and Patricia L. Pike. 1998. “Punishment enhances reasoning's effectiveness as a disciplinary response to toddlers.” Journal of Marriage and the Family 60:388-403.

Larzelere, Robert E., William N. Schneider, David B. Larson, and Patricia L. Pike. 1996. “The effects of discipline responses in delaying toddler misbehavior recurrences.” Child and Family Therapy 18:35-37.

LaVoie, Joseph C. 1974. “Type of punishment as a determination of resistance to deviation.” Developmental Psychology 10:181-189.

Lemonick, Michael D. and Alice Park 1997. “Spare the rod? Maybe.” Time, August 25, 1997, pp. 65.

MacMillan, Harriet L., Michael H. Boyle, Maria Y-Y. Wong, Eric K. Duku, Jan E. Fleming, and Christine A. Walsh. 1999. “Slapping, spanking and lifetime psychiatric disorder in a community sample of Ontario residents.” Canadian Medical Association Journal 161:805-809.

Matteson, Margaret E., Earl S.    Pollack, and Joseph W. Cullen. 1987. “What are the odds that smoking will kill you?” American Journal of Public Health 77:425-431.

McCord, Joan. 1991. “Questioning the value of punishment.” Social Problems 38:167-179.

Messner, Steven F., Lawrence E. Raffalovich, and Richard McMillian  2000.  "Economic deprivation and changes in homicide arrest rates for white and black youths, 1958-1997; A national time-series analysis"  Paper presented at the Eastern Sociological Society Meeting,   March 3.

Montague, Ashley. 1978. “Learning non-aggression: The experience of non-literate societies.” . New York: Oxford University Press.

Parke, R.D. 1977. “Punishment in children:  Effects, side effects and alternative strategies.” in Psychological processes in early education, vol. 71-97, edited by H. Hom and P. Robinson. New York: Academic.

Patterson, Gerald R., B. D. DeBaryshe, and E. Ramsey. 1989. “A developmental perspective on antisocial behavior.” American Psychologist 44:329-335.

Polite, Kenneth. 1996. “Response:  The medium/ the message:  Corporal punishment, an empirical critique.” Pediatrics.  The short- and long-term consequences of corporal punishment 98:849-852.

Rheingold, H.L. 1982. “Little children's participation in the work of adults, a nascent  prosocial behavior.” Child Development 53:114-125.

Roberts, Mark W. 1988. “Enforcing timeouts with room timeouts.” Behavior Modifications 4:353-370.

Roberts, Mark W. and Scott W. Powers. 1990. “Adjusting chair timeout enforcement procedures for oppositional children.” Behavior Therapy 21:257-271.

Rohner, Ronald P., Shana L. Bourque, and Carlos A. Elordi. 1996. “Children's perspectives of corporal punishment, caretaker acceptance, and psychological adjustment in a poor, biracial southern community.” Journal of Marriage and the Family 58:842-852.

Rohner, Ronald P., Kevin J.  Kean, and David E. Cournoyer. 1991. “Effects of corporal punishment, perceived caretaker warmth, and cultural beliefs on the psychological adjustment of children in St. Kitts, West Indies.” Journal of Marriage and Family 53:681-69.

Rose, Geoffrey. 1985. “sick individuals and sick populations.” International Journal of Epidemiology 14:32-38.

Rosenthal, R. 1984. Meta-analytic procedures for social research. Newbury Park, CA: Sage.

Ross, Susan M. 1996. “Risk of Physical Abuse to Children of Spouse Abusing Parents.” Child Abuse & Neglect 20:589-598.

Sampson, Robert J. and John H. Laub. 1993a. Crime in the making: Pathways and turning points through life. Cambridge: Harvard University Press.

Sampson, Robert J. and John N. Laub. 1993b. Crime in the making: Pathways and turning points through life. Cambridge, MA: Harvard University Press.

Sampson, Robert J., Stephen W. Raudenbush, and Felton Earls. 1997. “Neighborhoods and violent crime:  A multilevel study of collective efficacy.” Science 277:918-924.

Sears, Robert R., Eleanor C.  Maccoby, and Harry Levin. 1957. Patterns of child rearing. New York, New York: Harper & Row.

Simons, Ronald L., Kuei-Hsiu Lin, and Leslie C. Gordon. 1998. “Socialization in the Family of origin and male dating violence:  A prospective study.” Journal of Marriage and the Family 60:467-478.

Strassberg, Zvi, Kenneth A. Dodge, Gregory S. Petit, and John E. Bates. 1994. “Spanking in the home and children's subsequent aggression toward kindergarten peers.” Development and Psychopathy 6:445-461.

Straus, Murray A.  1992.  "Children as witnesses to marital violence: A risk factor for lifelong problems among a nationally representative sample of American men and women"  Paper presented at the Twenty-third Ross Roundtable on Critical Approaches to Common Pediatric Problems,  Ohio:  Ross Laboratories, 

Straus, Murray A. 1994. “Corporal punishment of children and depression and suicide in adulthood.” in Coercion and punishment in long term perspective, edited by J. McCord. New York: Cambridge University Press.

Straus, Murray A. 2000a. Beating the Devil out of Them: Corporal Punishment in American Families And Its Effects on Children, 2nd Edition. New Brunswick, NJ: Transaction Publishers.

Straus, Murray A. 2000b. “The benefits of never spanking: New and more definitive evidence.” in Murray A. Straus, 2nd Edition.  Beating the devil out of them: Corporal punishment in American Families And Its Effects on Children. New Brunswick, NJ: Transaction Publications.

Straus, Murray A. and Denise. Donnelly. 1994. “The fusion of sex and violence.” Pp. 121-136 in Beating the devil out of them: Corporal punishment in American families, edited by M. A. Straus. New York: Lexington Books.

Straus, Murray A., Richard J.  Gelles, and Suzanne K. Steinmetz. 1980. Behind closed doors: Violence in the American family. NY, New York: Doubleday/Anchor.

Straus, Murray A. and Holly S. Gimpel. 1994. “Alienation and reduced income.” Pp. 137-146 in Beating the devil out of them: Corporal punishmetn in American families, edited by M. A. Straus. New York: Jossey-Bass.

Straus, Murray A. and Glenda Kaufman Kantor. 1994. “Corporal punishment of adolescents by parents: A risk factor in the epidemiology of depression, suicide, alcohol abuse, child abuse, and wife beating.” Adolescence 29:543-562.

Straus, Murray A.  and Sean Lauer. 1992. “Corporal punishment and crime in ethnic group context.  Paper presented at the 1992 Annual Meeting of the American Society of Criminology, New Orleans.  To be published in Murray A. Straus  Corporal Punishment by parents in social context.  In Preparation.” . Durham, New Hampshire: Family Research Laboratory, University of New Hampshire.

Straus, Murray A. and Anitia K. Mathur. 1996. “Social change and change in approval of corporal punishment by parents from 1968 to 1994.” Pp. 91-105 in Family violence against children:  A challenge for society., edited by D. Frehsee, W. Horn, and K.-D. Bussmann. New York: Walter deGruyter.

Straus, Murray A. and Mallie J. Paschall  1999.  "Corporal punishment by mothers and children's cognitive development: A logitudinal study of two age cohorts"  Paper presented at the 6th International Family Violence Research Conference,  Durham, NH: Family Research Laboratory, University of New Hampshire,  27 July 1999.

Straus, Murray A., David B. Sugarman, and Jean. Giles-Sims. 1997. “Spanking by parents and subsequent antisocial behavior of children.” Archives of pediatric and adolescent medicine 151:761-767.

Thompson, Elizabeth E. In press. “The short- and long- term effects of corporal punishment on children: A meta-analytic review.” Psychological Bulletin .

Turner, Heather A. and David Finkelhor. 1996. “Corporal punishment as a stressor among youth.” Journal of Marriage and the Family 58:155-166.

Webb, Eugene J., Donald T. Campbell, Richard D. Schwartz, Lee Sechrest, and Janet Belew Grove. 1981. Nonreactive measures in the social sciences. Boston, MA: Houghton Mifflin Company.

Wells, L. Edward and Joseph H. Rankin. 1991. “Families and delinquency: A meta-analysis of the impact of broken homes.” Social problems 38:71-93.



* This chapter is a publication of the Family Research Laboratory, University of New Hampshire, Durham, NH 03824.  See the Laboratory web page ( for a program description and publications list.  It is a pleasure to express appreciation to members of the 1999-2000 Family Research Laboratory Seminar for valuable comments and suggestions. The work was supported by National Institute of Mental Health grant T32MH15161 and the University of New Hampshire.