PURPOSE. This study aims to describe a social process that demonstrates the relationship between youth runaway behaviors and delinquency. Although many researchers have reported strong linkage between runaway behaviors and delinquency, the dynamics of relationships have not been precisely investigated. Drawing upon the symbolic interactionism, a process model was developed to test mediating factors, such as street adversity, social stigma, and deviant networks, between the runaway duration and their delinquency. A causal model is established based upon the hypotheses: 1) the longer runway youth stay away from home, the more likely they will experience an adverse situation, 2) experiencing an adverse situation will lead youth to perceive they are being stigmatized and to form deviant networks, 3) Perceiving social stigma and engaging with deviant networks will increase youths' involvement in delinquent behaviors, and 4) the runaway experience itself will increase youths' perception as socially stigmatized and expose them to deviant networks.

METHOD. 334 runaway youth (161 male, 173 female) in South Korea participated in a self-administered survey. The runaway duration, street adversity, perceived social stigma, deviant networks, and delinquency were measured. Structural Equation Modeling was employed to test the measurement models and the hypothesized causal model using LISREL 8.54. A two-step approach was used to test the conceptual model, after validating the measurement models. Full Information Maximum Likelihood (FIML) was utilized as an estimation method that takes into account missing data based on the assumption of multivariate normality.

RESULTS. The results showed that runaway duration contributes to increased exposure to street adversity (.465) and deviant networks (.291). The street adversity is a significant mediating factor that explains almost half (.220) of the total effect (.481) of runaway duration on delinquency. The path from street adversity to delinquency was mediated by perceived social stigma (.094) and, mostly, by deviant network (.379). The deviant network is affected by runaway duration, directly as well as indirectly through experiencing street adversity, and reveals the greatest direct effect on delinquency (.899). Results indicate the model fit the data well (chi-square=4.081, df=2, p=0.130, RMSEA=0.0559). The squared multiple correlation for the structural equation of delinquency was .935, indicating that the estimated variance of delinquent behaviors is mostly accounted for by runaway duration with the model.

IMPLICATIONS. Findings suggest that delinquent behaviors among runaway youth are explained in the context of social situations. The failure of society in responding to youth who are in the most need of support and help is emphasized as a focal point to understand their limited chance of legitimate behaviors. Youth appear to adapt delinquent behaviors with perception of being stigmatized from mainstream society, and by achieving resources for survival from their deviant networks. Clinically, social stigma and deviant networks become the targets of intervention for delinquent youth to empower youth and to provide prosocial peer cultures. There is a need for increased awareness of and systems of care for runaway youth in Korea that should be enhanced in order to stop the process of social exclusion.