Wisconsin school districts are increasingly reliant on revenue garnered via referendums. The frequency, passage rates, and dollar amounts of school referendums in Wisconsin have all reached levels that are unprecedented in Wisconsin school finance history, even in White, working-class, rural communities that are growing more politically conservative. I review 60 factors that are associated with referendum outcomes. Rather than relying on old adages-football championships do not, in fact, improve the chances of referendum passage-I use methods of predictive regression to generate a model that best predicts post-Act 10, rural referendums in Wisconsin. This model produced particularly high sensitivity and excellent discrimination. The model also demonstrated that conservative, rural, White, working-class Wisconsin communities cannot and should not be written off as places that are antagonistic to referendum passage. I then present case studies focusing on two deviant rural referendum sites, one whose referendum passed when I predicted it would fail and one whose referendum failed when I predicted it would pass. I interviewed 53 people across both communities. My findings suggest that variables outside of school boards' and administrators' direct control, e.g. median income, are still "campaign-able." Independent campaign committees must answer three questions to persuade voters: where is the referendum money going, why is that project needed, and why is that project needed now? In order to make an impact, answers to these questions must be sent through an amenable social, cultural, political, and economic environment in which the district is situated. This method of campaign work is persuasive because of directed giving through voluntary taxation (DGVT). DGVT can be invoked via altruism, warm glow, preference elicitation, common identity reinforcements, and making individuals the face of the campaign.