We use a twenty-one year panel of data to examine the role of past income and aid, and expectations of future income, in regressions explaining state and local education spending. We show that simple estimates of the elasticity of spending with respect to financial resources are not robust to...