Outcomes-based financing: Possibility and promise in global health


Outcomes-based financing: Possibility and promise in global health


Hitting kids: American parenting and physical punishment

When Minnesota Vikings’ running back Adrian Peterson was indicted for hitting his son with a ‘switch’ in September 2014, there was a public furor – with arguments on both sides, but a general sense that Peterson was in the wrong.[i] Quickly, however, the debate over corporal punishment, which was at a fever pitch only two months ago, died down:


What explains this turnaround? Perhaps America lost interest because most Americans hit their kids, and most think that that is the way it should be. More than 70% of Americans agreed in 2012 that, “it is sometimes necessary to discipline a child with a good, hard spanking.”[ii] Of course, there is a wide range in how people define ‘acceptable,’ both in terms of frequency and severity.

Why do adults hit children? Whichever euphemism is used – “spank,” “smack,” “pop,” “whup/whip”—the goal is typically the same: to correct or to punish a child’s behavior by causing physical pain. In terms of altering children’s behavior in the short run, physical punishment is mostly effective. But questions remain about its long term effects, some of which we address in this memo:

  1. What are the longer-term consequences of physical punishment in terms of behavior?
  2. What are the longer-term consequences of physical punishment in terms of skill development?
  3. Is physical punishment associated with stronger or weaker parenting?
  4. What stance do governments in the U.S. and elsewhere adopt with regard to physical punishment of children?

Spanking and Child Behavior

Children spanked frequently and/or severely are at higher risk for mental health problems, ranging from anxiety and depression to alcohol and drug abuse, according to some research studies. Children whose parents hit them regularly may also develop more distant parent-child relationships later on.

There is also robust evidence of an increased incidence of aggression among children who are regularly spanked. A 2002 meta-analysis of 27 studies across time periods, countries, and ages found a persistent association: children who are spanked regularly are more likely to be aggressive, both as a child and as an adult. Many parents spank their children to put an immediate stop to bad behavior (e.g., shoving another child, reaching for a hot stove, etc.). Being on the receiving end, children may learn to associate violence with power or getting one’s own way. Indeed, much of the aggressive behavior attributed to children who were spanked differentially tends to correspond to interactions where violence is used to exert power over another person—bullying, partner abuse, and so on.

Note, however, that these studies focus on regular and/or severe physical punishment in terms of associations with child behavior.

Spanking and Child Skills

Studies dating back to the early 1960s suggest a relationship between corporal punishment and decreased cognitive ability in early childhood. Recent research has added support to these findings. A 2009 study examined two cohorts of children within the National Longitudinal Study of Youth (NLSY) and found that, even controlling for other parenting behaviors and demographics, children of mothers who used little or no corporal punishment “gained cognitive ability faster than children who were spanked.” MacKenzie et al. (2013) show that father’s high-frequency spanking at age five was associated with lower child vocabulary scores at age nine. Other studies have shown corresponding effects on school achievement. Bodovski and Youn (2010)  find that the use of physical discipline in kindergarten is associated with lower fifth grade math achievement. Margolin et al. (2010) find that children who were spanked are at higher risk of academic failure in the fifth grade.

Emerging evidence suggests that non-cognitive skills may also be affected. In an experimental study, Talwar, Carlson, and Lee (2011) tested whether attendance in a punitive versus non-punitive school environment had any effect on West African children’s executive functioning (EF) skills.[iii] They measured children’s abilities using three EF tasks: delay of gratification; gift delay; and dimensional change card sort. Their results suggested that—starting in grade 1—children who were in a punitive environment performed significantly worse than their peers in non-punitive school environments.

If hitting children is associated with slower skill development or other behavioral problems, there may be implications for life chances and social mobility, especially since the prevalence or intensity of punishment varies across socio-economic groups. But we should be very careful about drawing any causal conclusions here, even when there are robust associations. It is very likely that there will be other factors associated with both spanking and child outcomes. If certain omitted variables are correlated with both, we may confound the two effects, that is, inappropriately attribute an effect to spanking. For example, parents who spank their children may be weaker parents overall, and spanking is simply one way in which this difference in parenting quality manifests itself.

Parenting Quality and Physical Punishment

So: are parents who spank their children different on other dimensions of parenting? We investigate the relationship between parenting and corporal punishment using data from the Bureau of Labor Statistics’ Children of the National Longitudinal Survey of Youth, 1979 (CNLSY). As in our previous Parenting Gap research, we employ the HOME-SF scale as our proxy for parenting quality, but limit our sample to children who were ages 3 to 5 in 1986, one of the survey’s largest cohorts, and for whom the HOME-SF scale information is available.

The HOME-SF scale for children aged 3 to 5 includes 26 items—each its own proxy for “good” parenting. There are two designated items for corporal punishment. One self-reported item indicates how many times, if any, the mother hit her child during the previous week. The other item indicates whether the mother hit her child during the home observation. Figure 2 shows the distribution of responses, where N,N refers to a mother who did not hit her child in the previous week or during the observation. Nearly two-thirds of mothers reported spanking their children at least once in the two-week period. As might be expected, very few (5%) hit or spanked their children during the in-home observation.


Scoring the HOME scale is straightforward. Each positive behavior earns the mother one point. For the purposes of the corporal punishment items: If the mother is not observed hitting her child, she gains a point. If she reports hitting her child no more than one time during the previous week, she gains a point. If both, she gains two points. If neither, her score is unchanged. Of course, given this mechanical relationship, it is inappropriate to compare raw HOME scores between mothers who hit their children and mothers who do not. To make a meaningful comparison requires removing any items on spanking from the scale.

Figure 3 shows the results of such an exercise. The height of the bars indicates the raw HOME score, pre-adjustment; the darker blue height indicates the raw HOME score, post-adjustment. As expected, the gaps between mothers who hit and do not hit decrease by about one for both items.[iv] But the resulting gaps are miniscule—just over one-half of a point—and fall well within one standard deviation of the HOME score distribution.


Most studies suggest, however, that spanking becomes problematic with increased frequency and/or intensity. After all, there is a big difference between spanking your child once a month and spanking him or her twice a day, or spanking lightly with an open hand versus aggressively with a belt.

While we cannot observe spanking intensity, we do observe frequency in the data. Therefore, we replicate the above exercise in two ways—by frequency of spanking in the previous week (Figure 4) and by frequency across the two weeks (Figure 5).


Although sample size limitations prevent us from looking at mothers who reported hitting their children more than five times in the previous week, it is clear that—at least up until five—there is little evidence of any relationship between spanking and HOME score, even taking frequency in to account. At most, there is a one point gap between mothers who did not report hitting their children in the past week and those who reported hitting them at least five times, but this result is swamped by the corresponding standard deviations. Looking across weeks, the conclusion is the same.

Taken together, these results suggest that spanking is not a good predictor of parenting quality. That is, spanking is not systematically associated with other “negative” parenting behaviors.[v] There are some important caveats, however. We do not capture hitting by fathers or any other adults; nor do we have a measure of intensity of the hitting.  Moreover, the highest number of physical discipline incidents that we look at—five incidents over the span of a week—is a low threshold and as such, our analysis may not capture negative parenting skills associated with daily, repeated punishment. These may well be big factors.

But our overall finding is that spanking (by mothers, with no measure of intensity) tells us little about overall parenting skills. This contrasts with other parenting behaviors which have well-documented ‘spillover’ effects, such as reading books to young children.  Replicating the approach taken above for the reading item of the HOME scale, we document significant differences in raw scores between mothers who read to their kids more than once a week and those who do not (Figure 6). Even after adjustment, there is a two-point gap in HOME scores, which is large in terms of potential impacts on child development.


Hitting Children: The International Picture

It is also worth noting that the U.S. is relatively unusual in terms of attitudes, prevalence, and legal sanctions. Hitting children is more culturally acceptable in American than in many other nations – not only by parents, but by teachers (corporal punishment in schools is still permitted in 19 states). In many nations, physical punishment of children has now been outlawed, even for parents. In the table below, we summarize the legal position with regard to hitting children in a selection of counties.


Allowed in Home Allowed in School Conditions

Selected Evidence on Prevalence**

(from poll or study)

North America
United States Yes Depends Unlawful in 31 states 81% of parents say that spanking their children is sometimes appropriate.
Canada Yes No  “Force does not exceed what is reasonable under circumstances” 35% of children experienced some form of corporal punishment at least once per year.
Mexico Yes Yes “Right to correct” 26% of men 18-59 reported having been spanked or slapped by parent as a child.
South America
Brazil No No Banned as of June 2014 N/A
Colombia Yes No “Correct and sanction moderately” 61% of women report hitting, beating, spanking, or slapping their children.
Argentina No No Banned effective January 2016 N/A
Germany No No Banned as of 2000 N/A
UK Yes No “Reasonable punishment” 41.6% of parents physically punished or “smacked” child in the past year.
Sweden No No Banned as of 1979 N/A
China Yes No “Strict discipline” 50-60% of parents reported using mild corporal punishment in the last month.
India Yes Depends Unlawful in some states 65% of parents said they hit their children.
Russia Yes No “Reasonable chastisement” N/A
Nigeria Yes Yes “Correct…for misconduct or disobedience” 91% of children experienced “violent discipline” in their homes.
Ethiopia Yes No “For the purposes of proper upbringing” Only 1% of children reported never having experienced any type of corporal punishment.
Egypt Yes Yes “Right to discipline” 81% of children experienced moderate physical discipline.
Australia Yes Yes “Reasonable chastisement” 85% of parents admit to smacking their kids.
New Zealand No No Banned as of 2007 N/A

** Evidence is not comparable across countries. Each represents a finding of a specific study or poll with a unique sample and timeframe. For further information, please click the corresponding hyperlink.

Source for columns 1-3: Global Initiative to End All Corporal Punishment for Children.


In terms of parenting, our findings suggest that the immediate focus of U.S. policy-makers who want to improve our nation’s parenting should be on promoting positive behavior such as reading, rather than on large-scale efforts to prevent spanking (at least at the milder end of the spectrum).

But it is worth adding that most experts in child development believe that alternatives to spanking can be just as effective in terms of regulating behavior and that physical punishment of children is corrosive of long-term emotional development. The international trend is towards a growing opposition to the use of physical punishment for children. Within the U.S. there are signs of a slow trend in the same direction. On balance this must be seen as good news.

[i] After a plea deal, Peterson ended up with 80 hours of community service and a $4,000 fine for misdemeanor assault.

[ii] Smith, Tom W, Peter Marsden, Michael Hout, and Jibum Kim. General Social Surveys, 1972-2012 [machine-readable data file] Principal Investigator, Tom W. Smith; Co-Principal Investigator, Peter V. Marsden; Co-Principal Investigator, Michael Hout; Sponsored by National Science Foundation. NORC ed. Chicago: National Opinion Research Center [producer]; Storrs, CT: The Roper Center for Public Opinion Research, University of Connecticut [distributor], 2013.

[iii] Punitive schools employed discipline methods including beating with a stick, slapping on the head, and pinching. Non-punitive schools used time-outs, verbal reprimands, and trips to the principal’s office.

[iv] Women who reported spanking their child once in the previous week were still given one point. Therefore, the adjustment affected both groups.

[v] We also tested whether parents who spank their children ‘compensate’ with other positive parenting behaviors—like reading extra books to their children or helping them with school lessons, and found no evidence of such behavior.