Interactions between survey estimates and federal funding formulas

Formula allocations of United States federal government funds to states and localities are determined jointly by formula provisions and the data sources and estimation procedures used to derive formula inputs. Their interactions can have unanticipated consequences that are inconsistent with the policy goals of a program. Such interactions occur particularly when estimates subject to error are used with nonlinear estimation procedures or with nonlinear formula provisions such as a hold harmless or a threshold for funding. Introducing a new estimation procedure, such as model-based small area estimation, or a new data source, such as the planned American Community Survey, can also affect the allocations. Through simulations, we illustrate some of the paradoxical effects of interactions among formula provisions, data sources, and estimation procedures. We also describe features of the allocation formulas used by programs to improve elementary and secondary education and maternal and child nutrition for families in poverty. (January 2002)