Our results provide more consistent evidence than previous cross-lagged analyses on the predictive negative association of early drinking behaviors on later academic achievement. Crosnoe (2006) used a large and representative sample and found that academic failure predicted alcohol use in adolescence more strongly than vice versa, thus contrasting our findings. Several differences between the two studies could explain the discrepancy. Crosnoe (2006) used the number of classes failed as an indicator of academic failure whereas we studied teacher-reported GPA and the level of ongoing or attained education. While these measures are correlated, GPA is likely to capture more variation in school performance than the academic failure index. Naturally, differences in the educational systems and in the general drinking culture between Finland and the U.S. may also explain the findings. Specifically, adolescent alcohol use may be regarded as more deviant in the U.S., whereas experimenting with alcohol is relatively common and even socially normative among Finnish adolescents (Currie et al., 2012).
