[] Nathan wrote: > Now, the big theoretical point is whether unwanted children (from any > social class) are treated worse by their parents and are thus more likely > to be > involved in crime. I don't know the literature on this one well, but a > lot > of folks who fight child abuse have argued precisely this, and again, this > does seem like a more reasonable limited claim. > [] Child abuse is related to low socio-economic status. The point that I made was not that this wasn't the case, but that the elevated risks are not up to the task in terms of explaining a sizable portion of the criminal statistics. Other confounders, like alcohol, weapon availability, current (not past) socioeconomic status indicators such as income, occupation, and education, and a variety of situational variables that involve the victim, dilute the potential effect of unwantedness. There is also a problem with the definition of child abuse, which typically includes child neglect. Most poor parents cannot avoid the latter, since it is defined largely in terms available resources or living conditions. Another important issue is that (indexed, interpersonal non-economic) criminal acts are relatively rare. That's why statistical analyses using the national crime statistics (like the national crime victimization survey) will typically depend on several years' data to conduct statistical analyses. When we get down to the nitty-gritty in terms of counting and analyzing the numbers of reported criminal acts, there aren't a lot of data points to begin with. When crimes are examined by perpetrator characteristics, particularly one so narrow in scope, the ability to draw conclusions simply evaporates. In my opinion, instead of studying a very narrow and minor question because a dataset presented itself, or because of a desire to show that some social policy has unintended benefits (a policy that I agree with, by the way), it would have been nice if the authors had done some prior assessment of the real impact such knowledge would have. In actuality, their results really do nothing to improve social conditions. Instead, it seems to give more weight to the arguments of the Social Darwinists. Jeff