Nevertheless, the report concludes, measuring changes in the counts at a given school could be a fairly accurate way of determining changes in the number of poor children. For example, if a school’s lunch count increased, its Census data could be adjusted upward.
In addition, federal lawmakers would have to incorporate into the Chapter 1 formula accurate measures of fiscal capacity and local differences in the cost of providing educational services, the report states.
The report was requested by leaders of the House Education and Labor Committee, who are exploring options for changing the formula when the program is reauthorized in 1993. (See Education Week, May 6, 1992.)
In analyzing existing Education Department data, the GAO found that schools with high numbers of poor children “have disproportionately more low achievers than schools with fewer children in poverty.’' For example, schools with 50 or more poor children had an average of 2.9 “low achievers’’ for every 10 poor children, while schools with 126 or more poor children had an average of 4.9 low achievers for every 10 poor children.
Therefore, the researchers concluded, basing Chapter 1 allocations on the number of poor children in a school district, as the formula now does, underestimates the need in areas with particularly high concentrations of poor children. They recommended building in factors that increase funding for such districts.
Expenditure Factor Faulted
The report also cited shortcomings in the policy of allocating more funds to states with higher average per-pupil expenditures. The rationale is that the Chapter 1 dollar buys less in areas where costs are higher.
“The formula does not differentiate between the reasons for differences in average state spending,’' the report states. “Instead, it allocates fewer funds to those states that either cannot or do not spend as much on education.’'
The GAO suggested replacing the expenditure factor with one based on counties’ ability to generate revenue, and provided an example using county-level income data.
Finally, the researchers addressed proposals to more frequently update counts of poor children, which are based on the decennial census. States with rapidly growing student populations have often proposed using data on the number of children eligible for reduced-price school lunches for this purpose, and the GAO. recommended a partial use of these data.
The researchers noted, however, that school-lunch counts are an imperfect measure because they are influenced by local participation patterns and could be manipulated.
Nevertheless, the report concludes, measuring changes in the counts at a given school could be a fairly accurate way of determining changes in the number of poor children. For example, if a school’s lunch count increased, its Census data could be adjusted upward.