In a series of articles, Benton Johnson has investigated the effects of ascetic Protestantism on political party preference. His findings indicate that among laymen exposed to fundamentalist teachings, religious involvement would vary directly with Republican party preference. However, among laymen exposed to liberal teachings, religious involvement would vary inversely with Republican identification. This present study shows that church involvement and political party identification are not significantly related. A refinement of Johnson's liberal-fundamentalist dichotomy and his church interaction index again resulted in non-significant findings. Finally, a weighted least-squares procedure was employed yielding a set of linear estimation equations that again showed no significant effect. Results such as these should make the social scientist wary of the dangers associated with the measurement of religion and the contemporary relevance of Weber's Protestant ethic.