Bad argument for a spanking ban

In an essay advocating making benign spanking a criminal offense, Emily Bazelon writes:

Is the rise in kids beating on each other evidence of undisciplined, unruly child mobs, or the result of other unrelated forces? The data don't tell us, so take your pick. A similar split exists in the American social-science literature. In a 2000 article in the Clinical Child and Family Psychology Review, Dr. Robert Lazelere (who approves of spanking if it's ``conditional'' and not abusive) reviewed 38 studies and found spanking posed no harm to kids under the age of 7, and reduced misbehavior when deployed alongside milder punishments like scolding and timeouts. By contrast, a 2002 article in Psychology Bulletin by Dr. Elizabeth Gershoff (not a spanking fan) reviewed 88 studies and found an association between corporal punishment and a higher level of childhood aggression and a greater risk of physical abuse. This is the sort of research impasse that leaves advocates free to argue what they will -- and parents without much guidance.

This is not really a research impasse. It is entirely possible that the more aggressive kids exhibit behavior that causes them to be spanked more, and that the spanking is beneficial. Bazelon even admits that there is a study that indicates exactly that:

An effort by University of California-Berkeley psychologist Diana Baumrind to tease out the effects of occasional spanking compared with frequent spanking and no spanking at all. Baumrind tracked about 100 white, middle-class families in the East Bay from 1968 to 1980. The children who were hit frequently were more likely to be maladjusted. The ones who were occasionally spanked had slightly higher misbehavior scores than those who were not spanked at all. But this difference largely disappeared when Baumrind accounted for the children's poor behavior at a younger age. In other words, the kids who acted out as toddlers and preschoolers were more likely to act out later, whether they were spanked occasionally or not at all. Lots of spanking was bad for kids. A little didn't seem to matter.

Yes, there is a correlation between corporal punishment and physical abuse. That is like saying that there is a correlation between owning a car and drunk driving. It is just an artifact of the definitions, and has no policy implications.