Since Bradlaugh & Heather have mentioned Judith Rich Harris, I would recommend both of her books, The Nurture Assumption & No Two Alike to any reader who wishes be introduced to behavior genetics. You can also check out my interview of Harris from a few years back. For the more politically inclined, a summary of a discussion in The Corner about Ms. Harris’ ideas.
The core of Judith Rich Harris’ reason for offering her thesis is simple; for decades behavioral geneticists have found that on a wide range of traits human population level variation is predicted by the following components:
50% genetic variation
10% shared environment variation
40% other environment variation
You can dispute the details about the specific number, but, the short of it is that “shared environment,” the environment we experience growing up as children within the family, seems to have a much smaller longer term predictive effect than genetic variation and other environmental variation. The “other environmental variation” is somewhat squirrely. It could for example be reducible down to biological components, such as gene-gene interactions, or developmental noise. Judith Rich Harris argues that that other 40% are our peers, or, specifically peer culture.
When I say that 50% of the variation in the population can be accounted for by genes, I do not mean that the trait is “50% genetic.” I mean that if you look at genetic variation, from that you can already predict 50% of the variation of the distribution of the trait across the population. To make it more concrete, consider that genetics accounts for 80-90% of the variation in height in developed nations, but less in developing nations. How to interpret this? In developed nations nutritional inputs are maximized so that variation in height is not predicted by what you eat, because everyone receives a sufficient amount of calories for optimal development. In contrast in many developing nations there is variation in caloric intake so that some people are short simply due to a lack of nutritional inputs. So more of the variation can be predicted by environmental variation in a society where there is more environmental variation. In developed nations there is very little operational nurtitional variation when it comes to inputs which determine height; beyond a certain number of calories humans increase in size horizontally as opposed to vertically. Naturally the only variance that remains is genetic, ergo, height is more heritable in developed nations.
What does this have to do with personality and behavior? In a society which is fragmented into various subcultures one assumes that predicting from genes to trait value will be dicier because all things are not controlled. In other words, cultural framework matters, but within that framework variance still remains. Judith Rich Harris makes the provocative argument that selection of peers are more efficacious in affecting life outcomes in offspring than direct parental modelling. Perhaps being from a broken home is less important than your peers being from broken homes. Within families this naturally results in much of the variance in sibling outcomes being due to their varied genetic endowments because they likely share peer environment.
How does this affect politics and public policy? I believe it makes the model of human psychological and social development more complex. It also suggests that there will be limits on what environment can do to eliminate inequality or variation. It also implies that a certain level of goodness or badness may be “baked in the cake,” so to speak.