Hmm.. The underlying studies (Baum, 2000; Eensoo et al., 2005) that found links between socioeconomic status and DUI/DWI did so by comparing those arrested against a control group – which I believe should deal with that concern?