AVOIDING BIAS IN ESTIMATES OF THE EFFECT OF DATA ERROR ON ALLOCATIONS OF PUBLIC FUNDS

Commonly used techniques for estimating the effect of data error on fund allocations are biased because they fail to account for numerous sources of data error and for misspecification of the allocation formula itself. Alternative techniquesfor estimating the effect of data error are presented and applied to estimate the effect of census undercount on general revenue sharing allocations. overnments use statistical allocation formulas to distributehuge amounts of funds. For example, the federal general revenue sharing (GRS) formulas allocated over $55 billion to 39,000 state and local governments during the period 1972-1980. The GRS formulas, which are still in use, calculate a government’s allocation on the basis of over a dozen data series such as population, per capita income, and state individual income tax collections. No data series is completely error free and errors in data give rise to errors in the allocations. Furthermore, the allocation formulas them-selves are often crude and imperfect. Recently, considerable effort has gone into determining the extent of the misallocation attributable to undercount in the decennial census (Maurice and Nathan, 1982