Delinquency Balance: Revisiting Peer Influence*

An impressive number of inquiries across an array of methodological specifications has demonstrated that deviant peers are an important correlate of various criminological outcomes, which include withinindividual change and stability in offending behavior. Still, the causal mechanisms of peer influence arguably remain underdeveloped (Giordano, 2003; Warr, 2002). In an attempt to expand the dialogue on the nature of peer influence, this inquiry proposes that scholars would benefit from considering relative peer deviance in addition to exposure to deviant peers. Specifically, it argues that an imbalance in delinquency between friends helps to explain delinquency change/stability; therefore, exposure to deviant peers is not always risky and exposure to less deviant peers is not always protective. The analysis uses the AddHealth data to construct within-individual and across-individual (delinquency) difference scores and relies on self-reports rather than on perceptions for the best friends’ delinquency. The results provide support for the