I guess this question has rarely been raised, but several studies (including my own) indicate that students after being retained one year show some benefits in math, compared to matched promoted students, but show less frequently benefits in languages.