NFL superstar Adrian Peterson was apparently raised to believe that spanking is good for children—indeed, that his grown-up success is partly due to the spankings he received as a child. Now he has been arrested and deactivated indefinitely by the Minnesota Vikings because of allegations that he disciplined his four-year-old son by spanking him with a tree branch. Peterson may well think that he did nothing wrong, feelings shared by people worldwide who believe that children learn self-discipline from a hand, belt, paddle or tree branch. Many surely agree with a teammate who said, “growing up, my mom, she disciplined me the same way and it got me to this point now. I'm in the NFL and I know how to behave.”
What’s the truth? Is spanking good or bad for children? The truth is that we don’t know and probably never will. There have been pseudo-scientific studies but they are little better than anecdotes. One study, for example, found that mothers who reported spanking their children also reported that their children were aggressive. The authors concluded that spanking makes children aggressive. Not so fast. Perhaps the causation runs in the other direction—aggressive children are more likely to be spanked. Or perhaps mothers reported the aggression in order to justify the spanking.
A more general problem—endemic to these kinds of studies—is that these are observational data, not experimental data. In a true scientific study, families would be randomly divided into two groups: spankers and no-spankers. The no-spank families would not be allowed to spank their children, no matter what. The spank families would be compelled to spank their children. After the children were grown, their character would be assessed by impartial observers who do not know which children were raised in which group.
Crazy? Of course. That’s why we have to make do with observational data—observing spankers and non-spankers. Unfortunately, observational data are plagued by self-selection bias in that people are not randomly assigned, but instead choose to do what they are doing. People who make different choices may be different, and the outcomes may reflect who they are rather than what they are doing.
On average, college graduates earn more than high school graduates, suggesting that the income difference measures the financial return from going to college. However, part of the reason college graduates earn more may be that they are brighter and more ambitious than those who do not to go to college.
A psychology professor observed students drinking at three bars near campus and found that, on average, drinkers who ordered beer by the pitcher consumed more than twice as much beer than did drinkers who ordered beer by the glass or bottle. He conclusion, reported nationwide, was that, “If we banned pitchers of beer we would have a significant impact on drinking.” There is self-selection bias here in that people who order pitchers of beer are surely planning to drink a lot and generally fulfill their objective. Big drinkers will be big drinkers even if they are forced to do so by the glass or bottle. Years later, the [rofessor admitted the obvious: many college students “intend to get intoxicated. . . . We have shown in several studies that their intentions influence their behavior. If they intend to get drunk, it’s difficult to stop that.”
If people who work on Wall Street are competitive, that may be why they chose to work on Wall Street. If people who play football are aggressive, that may be why they chose to play football. If parents choose to spank their children, it may be because (you fill in the blanks). Genetic differences, socioeconomic differences, home environment, parenting style. There is no shortage of possibilities. All we know for certain is that observational data will not tell us for certain whether spanking is good or bad for children.