There is an established research literature that finds increasing adolescent religiosity is associated with decreasing forms of delinquent behaviors, including drug use and underage drinking, interpersonal violence, and non-violent crime.  However, there is little research examining religion and changes in delinquent involvement.  If religion is negatively related to overall delinquent involvement, it should protect against the initiation or intermittent involvement in deviant behavior.  There are reasons to also expect religiosity to foster desistence.  Once individuals become involved in deviant behavior, a strong religious background, or an increase in religious involvement may help steer them away deviant behavior. Therefore, in the present study published in Deviant Behavior: Religious Involvement and Dynamics of MarijuanaUse: Initiation, Persistence, and Desistence, we examine the effect of early adolescent religion, and then later changes in religiosity, on the initiation of, persistence in, and desistence from marijuana use in later adolescence and early adulthood.

 We used data from the National Longitudinal Study of Adolescent Health (Add Health).  For Add Health, a cluster sampling design with systematic sampling methods and implicit stratification was used to select a representative sample of 80 high schools, along with 52 ‘‘feeder schools’’ (i.e., junior high=middle schools) for those high schools that did not span grades 7 through 12.  A nationally representative ‘‘core sample’’ of 12,105 7th through 12th graders and ‘‘special oversamples’’ of 15,454 adolescents were drawn from school rosters. For each adolescent respondent a parent was also interviewed at Wave I.  Interviews at Wave II were conducted with 14,738 adolescents from April through August 1996.  Wave III data were collected five years later. After attrition between waves and listwise deletion of cases with missing data, the total number of cases for our analysis is 7,331.

 We find that (1) religious youth are more likely to never use marijuana, and are less likely to engage in persistent marijuana use; (2) adolescents who believe in a literal interpretation of religious scriptures are also more likely to never use marijuana; (3) adolescents’ religious involvement does not have a significant effect on any of the desistence comparisons; adolescents who believe in a literal interpretation of the sacred scriptures of their religion are more likely to desist from marijuana use than to persist; and (4) adolescents who decreased their religious involvement over time, compared to adolescents with stable religious involvement, are significantly more likely to initiate marijuana use and to persistently use marijuana than they are to never use.

 As we hypothesized, adolescent religiosity effects on marijuana use occur primarily because religiosity is associated with increased attachment to parents, greater self control, and fewer associations with delinquent peers.  These findings support the important argument that the effects of religion on behavior are unique in that they are not wholly attributable to non-religious social or psychological factors.

Jeffrey T. Ulmer and Scott A. Desmond are non-resident research fellows at Baylor’s Institute for Studies of Religion.  Sung Joon Jang is associate professor of sociology and a resident scholar at ISR.  Byron Johnson is Distinguished Professor of the Social Sciences at Baylor University, where he directs the Baylor Institute for Studies of Religion as well as the Program on Prosocial Behavior.  The Baylor Institute for Studies of Religion is an academic think-tank that specializes in social research and public policy analysis on religion.