In this chapter, we have presented an exhaustive review of the literature on parental warmth/rejection and violent behavior. While the findings on parental warmth and violence are fairly consistent, they are not especially strong. The findings on parental rejection and violence are more consistent and robust. The proportion of studies with a PoC that is in the right direction and statistically significant in the rejection category is much larger (66%) compared to the proportion of studies of warmth (25%). This conclusion stands if we look at higher quality multivariate studies where the evidence of an association between parental rejection and violence is quite strong (7 out of 9 studies), and the evidence for an effect of parental warmth is weak (2 out of 10). In particular, in analyses where a control for child effects was applied, associations between parental warmth and physical aggression were wanting, but the effects of parental rejection remained statistically significant.
The effect of parental rejection on violence among males is quite consistent, but it is surprisingly unclear whether parental warmth or parental rejection influence violence among females. There is a paucity of studies and the findings among them are unconvincing. In addition, the evidence reviewed here indicates that any broad generalizations about the value of parental warmth as protective against causes of violence are premature, though we can say that parental rejection appears to exacerbate a variety of risks. In keeping with a long history of research on caregiving, the effects of maternal warmth do not seem to differ from the effects of paternal warmth in their association with physically aggressive and violent behavior. Unfortunately, too few studies of “paternal rejection” are available to comment on whether rejection by fathers matters as much as rejection by mothers.
Finally, some mysteries remain regarding the differential etiology hypothesis. Both violent and nonviolent offending are associated with the parental warmth/ rejection dimension and studies providing estimates of both outcomes report similar patterns. However, in studies comparing violent to nonviolent offenders, the evidence leans in favor of our hypothesis, especially for parental rejection, leaving us to conclude that parental rejection is still a good prospect and deserves further study. In Chapter 2, we brought in the possibility that environments that depart very significantly from an “average expectable environment” may be most important in the development of severe pathology. From that point of view, we suggest that the pattern of findings is consistent with the idea that normal variation in parental warmth may not have the capacity to influence violent behavior very much, but at the extreme low end (rejection), it can cause significant, ongoing, developmental problems some of which would lead to severe antisocial behavior like violence.