Using data collected in 95 Ohio school districts by Brock Hunt, this study reanalyzed those data, examined the reasons for Hunt's inability to predict a school board member incumbent's defeat, and developed a model that would have predicted those Ohio school board elections. Hunt attempted to develop a statistical model of socioeconomic-political indicators that would predict the defeat of an incumbent, but failed, as his model could only account for such defeats after the fact. It is suggested in this paper that the effect of time distorted the prediction model developed by Hunt because the model did not take account of the concept of "shrinkage." In the present study, variables which are stable and consistent across time were obtained and a dissatisfaction factor--based on the Dissatisfaction Theory of American Democracy--that took into account the number of challengers in the election was developed. It was found that, without the dissatisfaction factor, election outcomes could not be predicted, thus supporting the Dissatisfaction Theory. Many tables are supplied, and 13 references are listed. (DCS)