In-group favoritism has selective advantages. What are the bases (genetic) for such favoritism. Two strong hypotheses:
1. A flexible central affiliation mechanism (CAM) that has access to the full range of perceptual information signifying group membership. Here the chosen in-group will be relatively arbitrarily assembled on the basis of context and will be updatable over time.
2. A set of essentialist features (such as kinship, ethnicity, or shared beliefs), as suggested by the strong-specificity hypothesis. This predicts that favoritism will be uncorrelated across domains, relatively hard to form from arbitrary cues, and relatively invariant to context.
The authors note that these are not necessarily mutually exclusive. It is also plausible that both a flexible CAM system and a set of essentialist mechanisms have evolved in humans.
The main contrast: If there is a single neural or mental mechanism for affiliation across multiple behavioral domains, this would constrain genetic and environmental influences on favoritism to be mediated through the CAM factor. However, if there are distinct, essentialist mechanisms (i.e., if each essential group-formation domain has its own evolved mechanism), different kinds of favoritism would have different sets of genetic and environmental influences.
NOTE: this kind of hypothesizing seems crazy to me. Of course they are both active. And this is what the results show. But that might actually be useful. There are essentialist features of how we identify others, but we are also flexible in how we form group affiliations.
Participants: 957 pairs of twins (MZ, DZ same-sex, DZ opposite-sex) were assessed for religious, ethnic, and racial in-group favoritism.
Measures of in-group favoritism: For each group type, we assessed three distinct elements of in-group favoritism: (a) strength of identification with the group, (b) preference for affiliating with in-group members, and (c) the importance placed on marrying within the group. For instance, religious in-group favoritism was assessed with the following items: “How closely do you identify with being a member of your religious group?” “How much do you prefer to be with other people who are the same religion as you?” and “How important do you think it is for people of your religion to marry other people who are the same religion?” The items for ethnic and racial in-group favoritism had the same wording as the items for religious in-group favoritism, except that the appropriate group type was substituted (e.g., “How closely do you identify with being a member of your racial group?”). All responses were made on 4-point Likert scales (from 1, very, to 4, not at all). Interitem correlations were high within each group type (ranging from .32 to .65, with all but one correlation greater than .40), and scale scores were calculated as the sum of item scores within a group type.
Results
MZ twins were significantly more similar to each other than were DZ twins for all three types of in-group favoritism, a pattern suggesting a genetic influence.
Their first result is that MZ twins are more strongly correlated across all three types of in-group favoritism (see Figure). Which makes sense, since MZ twins are correlated across many personality dimensions.
Then, they ran statistical models to predict the levels of correlations between twins using path analysis. They found that models that assumed either factors only for a central affiliation mechanism (i.e., general levels of in-group favoritism across domains) or completely independent factors for in-group favortitism in each domain (e.g., in-group favoritism in the racial domain would be uncorrelated with favoritism in the religious domain) were not as good fits for the data as a model that incorporated both.
This all seems incredibly obvious to me. There are broad personality tendencies that should correlated with tendencies to express in-group favoritism/out-group aversion across domains. At said, there are also differences in how people identify with different types of communities. Since twins share much of their environment, there should be overlap in how they view these domains that are separate from the correlations between general tendencies for in-group favoritism. There are also a lot of confounds, because they don't break-up the genetic and environmental components that well. The assumption that MZ twins share the same amount of their environment with DZ twins is always problematic, though I think comparing same-sex and opposite-sex DZ twins was a nice way to at least partially control for that.
The take away, I guess, is that people can express strong affiliations for different groups, and that the degree to which people do this may have genetic components, probably related to more general personalities.
