This article from the Weekly Standard answers the question.

Excerpt:

A March 1996 study by the Bureau of Justice Statistics contains some interesting findings that indicate just how widespread the problem may be. In a nationally representative survey of state prisoners jailed for assaults against or murders of children, fully one-half of respondents reported the victim was a friend, acquaintance, or relative other than offspring. (All but 3 percent of those who committed violent crimes against children were men.) A close relationship between victim and victimizer is also suggested by the fact that three-quarters of all the crimes occurred in either the perpetrator’s home or the victim’s.

A 1994 paper published in the Journal of Comparative Family Studies looked at 32,000 documented cases of child abuse. Of the victims, only 28 percent lived with both biological parents (far fewer than the 68 percent of all children who live with both parents); 44 percent lived with their mother only (as do 25 percent of all children); and 18 percent lived with their mother and an unrelated adult (double the 9 percent of all children who live with their mother and an unrelated adult).

These findings mirror a 1993 British study by the Family Education Trust, which meticulously explored the relationship between family structure and child abuse. Using data on documented cases of abuse in Britain between 1982 and 1988, the report found a high correlation between child abuse and the marital status of the parents.

Specifically, the British study found that the incidence of abuse was an astounding 33 times higher in homes where the mother was cohabiting with an unrelated boyfriend than in stable nuclear families. Even when the boyfriend was the children’s biological father, the chances of abuse were twice as high.

These findings are consonant with those published a year earlier by Leslie Margolin of the University of Iowa in the journal Child Abuse and Neglect. Prof. Margolin found that boyfriends were 27 times more likely than natural parents to abuse a child. The next-riskiest group, siblings, were only twice as likely as parents to abuse a child.

More recently, a report by Dr. Michael Stiffman presented at the latest meeting of the American Academy of Pediatrics, in October, studied the 175 Missouri children under the age of 5 who were murdered between 1992 and 1994. It found that the risk of a child’s dying at the hands of an adult living in the child’s own household was eight times higher if the adult was biologically unrelated.

The Heritage Foundation’s Patrick Fagan discovered that the number of child-abuse cases appeared to rise in the 1980s along with the general societal acceptance of cohabitation before, or instead of, marriage. That runs counter to the radical-feminist view, which holds that marriage is an oppressive male institution of which violence is an integral feature. If that were true, then child abuse and domestic violence should have decreased along with the rise in cohabitation.

Heritage also found that in the case of very poor children (those in households earning less than $ 15,000 per year), 75 percent lived in a household where the biological father was absent. And 50 percent of adults with less than a high-school education lived in cohabitation arrangements. “This mix — poverty, lack of education, children, and cohabitation — is an incubator for violence,” Fagan says.

Why, then, do we ignore the problem? Fagan has a theory: “It is extremely politically incorrect to suggest that living together might not be the best living arrangement.”

The moral of the story is that it is a lot safer for children if we promote marriage as a way of attaching mothers and fathers to their children. Fathers who have a biological connection to children are a lot less likely to harm them. And a lot of social problems like child poverty, promiscuity and violence cannot be solved by replacing a father with a check from the government. We need to support fathers by empowering them in their traditional roles. Let the men lead.