“With Friends Like These: Aggression from Amity and Equivalence”, 2020-11-01 (; similar):
Some teenagers are willing to bully, harass, and torment their schoolmates in order to achieve popularity and other goals. But whom do they bully? Here, we extend the logic of instrumental aggression to answer this question. To the extent that friendships are the currency of social status, we should expect social aspirants to target their own friends, their friends’ friends, and other structurally equivalent schoolmates. This tendency, we argue, extends beyond what would be explained by propinquity, and we expect that victimization by friends will be particularly distressing. We test these hypotheses using panel social network data from 14 middle and high schools at two time points during a school year. Findings from temporal exponential random graph models suggest that our expectations are correct: the tendency to be cruel to friends is not substantially influenced by propinquity, and victimization by friends has adverse consequences for mental health.
…We both heed this warning and expand on it, by challenging a core assumption in balance theory and in most network research: that positive and negative ties are mutually exclusive. Thus, our goal here is not to test balance—an impossibility if friends are also enemies—but instead to propose a theory of “frenemies.” Overlap between positive and negative networks is rarely if ever examined in the small empirical literature on negative tie networks, as it would seem strange to dislike a friend or to avoid eating lunch with a classmate you would nominate for student council (2013; 2017). But it is not incomprehensible for people to be cruel to their friends, or their friends’ friends. Indeed, there are good reasons to expect them to do so.
In contrast to both balance theory and much of the empirical literature on bullying, which concludes that victims are isolated or marginal and thus sit at relatively large social distances from their tormentors, we extend the logic of instrumental aggression to anticipate higher rates of aggression at low social distances, between friends and among structurally equivalent schoolmates. This is not because they spend more time with one another, but because they compete for the same social positions and relationships. We test these hypotheses using temporal exponential random graph models (TERGMs) of networks of aggression from 14 middle schools and high schools over two time points during one school year. We further anticipate that betrayal by friends is acutely painful relative to harassment by others, and so we also examine the consequences of each source of victimization for well-being. And thus, we are not so sanguine as Lincoln in asking, Where do our enemies come from? The answer, we conclude, is that they are close by.