Using Census Data for Funds Allocations

Pursuant to a congressional request, GAO reviewed the potential effects of using census data, adjusted for undercounts, in the allocation of federal funds, focusing on whether jurisdictions would have gained an additional $56 per undercounted person, had the 1990 Census been adjusted to reflect those undercounts. The $56 estimate was suggested in an article by Professor Michael P. Murray in 1991.

GAO noted that: (1) citing a single figure to estimate the potential effects of census undercounts on federal funds distribution tends to oversimplify a complex subject; (2) census population data are rarely the only element in allocation formulas; (3) other data elements and a variety of provisions that affect the size of formula allotments have an important impact on the final amounts distributed to states; (4) it is important to recognize that many federal formula grants have a fixed pool of funds to distribute; (5) in separate studies completed in 1991, both GAO and Professor Murray estimated that adjusting census population data to reflect net undercounts in the 1990 Census would redistribute only a fraction of 1 percent of total net federal funds allocated through formula programs; (6) on the basis on his analysis, Professor Murray estimated that, for the 34 to 41 percent of states that might gain additional funding from an adjustment, such gains would average about $56 per miscounted person; (7) this estimate tends to oversimplify the effects because it is a composite average and applies only to the jurisdictions gaining funds and is not applicable to the remaining jurisdictions; (8) it does not fully consider the variety of potential changes in allocations for each federal formula program; (9) both GAO and Professor Murray's analyses were completed using estimates of 1990 net undercount rates that were available at the time the Secretary of Commerce made his decision on statistical adjustment; (10) the Bureau of the Census subsequently lowered those undercount estimates to correct for a computer coding error and to reflect other revisions made that resulted from further research; (11) this downward revision would lessen further the relative redistributional effects identified; (12) even when changes are relatively small, actual dollar amounts involved appear substantial, especially to the affected jurisdictions; (13) the overall magnitude of federal funding allocated on the basis of formulas using population data, in whole or in part, has increased dramatically since 1991; (14) GAO reported that 100 programs using population data in fiscal year (FY) 1991 had estimated obligations of over $116 billion; (15) by FY 1996, the 20 largest programs on GAO's list had estimate obligations of over $170 billion; and (16) the effect of changes in funds distribution would become more substantial when applied over an entire decade.