Do Small States Get More Federal Monies? Myth and Reality about the US Senate Malapportionment

We analyze the relationship between senate malapportionment and the allocation of the US federal budget to the states during the period 1978-2002. A substantial literature originating from the influential paper by Atlas et al (1995, using a within estimation methodology finds that small and overrepresented states get significantly larger shares of federal funds. Revisiting the econometric specification used by the current empirical research, we show that the number of senators percapita is inappropriate to capture malapportionement in regressions using broad federal programs, and that the results obtained with this indicator are extremely non-robust to reasonable specification changes. In particular, senators percapita have a significant impact on federal spending only in regressions containing state fixed effects. Furthermore, the coefficients estimated using the within methodology are statistically different across states and, therefore, cannot be used to assess spending differentials between states. The magnitude and significance of those coefficients suggest a within state-specific inverse relationship between broad spending categories and population which is not systematically related to the size of the states and seems more compatible with incrementalist theories of budget allocation.