Methods: We use a linear mixed effects (LME) model to understand predictors of Title I and Title II allocations from 1998 to 2004. Independent variables included: concentration of AIDS cases in eligible metropolitan areas (EMAs), southeastern geography, poverty, race, and, as a measure of political power, number of Congressional delegates. A linear spline at the year 2000 controlled for the new Presidential administration and the year 2000 reauthorization legislation. Observations were weighted by the number of AIDS cases in each state.
Results: States' concentration of AIDS cases within urban areas was the dominant predictor of RWCA funding (p<0.001), though the impact of concentration declined after the year 2000 (p<0.001). Plots of state-specific trajectories and smoothed mean curves comparing states with and without EMAs illustrate this finding graphically. After adjusting for urbanicity of AIDS cases, southeastern geography, race, and poverty were jointly statistically insignificant in predicting funding allocations (p = 0.862). After the year 2000, there was a small redistribution of funds to states with fewer cases in urban centers (p<0.001). The separate funding streams to metropolitan areas and states, embodied in the dual Title I/Title II allocation, remains the dominant predictor of funding variation across jurisdictions.
Conclusions: Despite the complexity of RWCA funding algorithms, empirical analyses do not support contemporary arguments regarding the fairness of current allocations. We identified little systematic targeting of funds to achieve explicit goals beyond the allocations dictated by the pattern of AIDS cases within and across jurisdictions.