I wrote before about the problems people have mistaking correlation and causation, mistaking a pair of events as cause and effect simply because they occur together. However, I was struck today by another such assumption, and, unfortunately, one dear to the hearts of many conservatives.

That is the assumption that criminality is significantly higher in single parent households due to the lack of a father.

Now, I am not going to argue about the significance of a father in a child's life. I am not saying that children do not suffer for lack of a father. In fact, my entire argument says nothing one way or another about the concept that a father is essential to raising a child. My only point is this, the arguments sued are hopelessly flawed.

How so?

Well, the most basic problem is with the sort of families which are predominantly female headed. By and large, such families are among the poorest of urban families, of all ethnic groups. The children raised in such household are also raised in an environment which is conducive to criminality. Whether or not one has both parents, the environment and social pressures are such that more criminals are produced than in other environments.

But that is just the first problem.

The second problem is the question of why a household would be female headed. Some may be single through choice, but many fall into two categories. First, women who are promiscuous by nature, either as a profession or simply from their acculturation, both of which also tend to correlate with the same socioeconomic groups which produce high rates of criminals. Second, women whose boyfriends or husbands are either in jail or dead. As should be obvious form this description, children born into such surroundings would also be prone to adopt a criminal lifestyle*. Their absent father may not be pressing them into crime, but the fact that their mother accepted him indicates that the family has no problem with criminal behavior. So it is much more likely such children will adopt a criminal course in life.

Of course, one could adjust for this, pick middle children from similar socioeconomic backgrounds, with families having similar criminal backgrounds, but a far as I know such studies have never been done. Even if they had, it would be hard to control for all the factors which lead tot eh creation of single parent households.

About the best I could imagine would be a study of the children of soldiers whose fathers had died compared to other military children. In that case not only would we have a fairly uniform pool, but we would also have a pool of absentee fathers whose absence is not due to criminality or irresponsibility, and whose mothers had no expectation of being a single parent.

But I have never seen such a study done. By and large, the studies I have seen have been of two types. Those trying to maximize the importance of a father, comparing single parents from poor and criminal circumstances to more affluent and responsible "controls" or those trying to minimize the father, which compare families from equally criminal backgrounds**. Neither provides a good basis on which to form judgments.

But that doe snot stop either side. I have heard, over and over, both sides of the argument touting these defective studies as "clear proof" of the importance of fathers. And that, in the end, helps no one. Though some may feel they gain by convincing others of their point, even with defective evidence, in truth, we all benefit when an argument is evaluated fairly and the truth is revealed. Using faked evidence to "win" does nothing to advance the truth.

---------------------------------------------------------------

* Whether one accepts that promiscuous groups are also more prone to criminality, or whether promiscuity correlates with groups prone to more criminality, it should be clear professional prostitutes are a group likely to have little objection to criminal activity. Similarly the wives and girlfriends of jailed or dead criminals are not likely to be strongly opposed to crime. So at least three of the four possible categories are clearly more prone to criminality.

** These sound like valid studies, where both groups are equally criminal and poor, but they are not. Picking groups in which the father is as likely to encourage bad behavior as discourage it is hardly a valid way to study whether a father is important under normal circumstances. No one doubts that fathers can be a force for ill, what needs to be established in whether, in a properly functioning family, a single mother, or a mother and father, provides a better environment. Picking disfunctional families and finding them equally destructive with or without a father does nothing to prove either side of the argument.

------------------------------------------------------------

POSTSCRIPT

I confess I am not up to date on studies of the role of fathers in families. My knowledge is several years old and based on a layman's dabbling in this area, not a comprehensive and life long review of all the literature. However, as I have often heard arguments from both sides, and seen what they consider the "best" evidence, and found it wanting, I feel comfortable assuming there is little truly sound evidence out there.

But I admit I may be wrong. Perhaps both sides overlooked some better evidence. Or the good evidence is too inconclusive to appeal to either side. If there is a better study out there, please let me know, as I am curious to see where the evidence points.

POSTSCRIPT II

My arguments about misuse of evidence were most recently presented in the post "The Devil is in the Definitions (And Assumptions)". A complete list of writing on the topic can be found in the postscript to "Another Example".