My dissertation examines how rural Southern farmers responded to exogenous changes to household incomes during the early twentieth century. I use a range of statistical approaches: Ordinary and Two Stage Least Squares, Probit, and instrumental variable Probit. Critically, the proxy for household incomes, cotton yields, is predicted using weather fluctuations. In chapter one, I find credit constrained households have lower school attendance rates following negative income shocks. In chapter two, I find the probability of farm wage work is negatively correlated with incomes in credit constrained households. In both chapters, I use black and tenant farmers as proxies for credit constrained households. In the final chapter, I find the probability of a lynching occurring in the local community increases after household incomes fall.