Comparison of the 1990 Census Low Income Count
to Department of Human Services Medicaid / Food Stamp Count

We have obtained the new poverty measure from the Department of Human Services (DHS). Of the nearly 478,000 cases, 85.9% were matched to the Tiger Line files by address. Another 3.9% were matched to 9-digit Zip Code and 2.9% to 7-digit Zip Codes. Nearly 35,000 (7.3%) could only be matched to 5-digit Zip Codes. Of those 35,000 cases 26,900 were either post office boxes or rural routes. The count was provided to us by school district by age of recipient. Included in the count are recipients of Food Stamps ,TANF, Kid Care and Medicaid. Since the 1990 Census identifies children ages 5 thru 17, we used the same age groupings for the DHS count. Following are general comparisons of the DHS count to the 1990 Census (currently used in the calculation of the GSA poverty grant).

Table I shows that 198 districts have a smaller DHS count than Census count. This implies that if the DHS count were simply substituted for the Census count in the current poverty grant formula, 198 districts would receive reduced funding under that mechanism. Table II provides the distributional effects of substituting the DHS count for the Census count.

Table II shows the total cost of directly substituting the DHS count for the Census count to be $427.6 M ($437.1 - $9.5). 198 districts would lose a total of $9.5 M through the grant. Table III provides the distributional effects if districts were allowed to use the higher of the DHS or Census counts in the poverty formula.

Since the increased cost of substituting the DHS count will result in an increased cost to the State in excess of $400 Million, the alternatives below are offered for discussion purposes.

1. Use the greater of the Census or DHS counts but reduce the current poverty grant amounts. For example assume poverty grant amounts, currently $355, $675, $1,190, $1,333, $1,680 and $2,080, were adjusted downward to $150, $300, $500, $700, $850 and $1,050. The cost to the State would be $37.9 M. 433 districts would gain $63.8 M and 460 districts would lose $38.0 M.

2. Use the greater of the Census or DHS counts but prorate the grant amount down to an acceptable level. For example if it is determined that only $400 M would be available for the grant (an increase of only $25 M), the grant could be prorated to cost only $400 M. If this were done, for example, 466 districts would gain $61.6 M and 427 districts would lose $36.9 M.

3. Use the greater of a percentage of the Census or a percentage of the DHS count and reduce the percentage of Census and increase the percentage of DHS over time. For example, if in the first year the greater of 100% of the Census or 65% of the DHS were used and the balance of the DHS count were funded at 5% of the grant category, the cost to the State would be $54.1 M. No districts would lose funding under the poverty grant portion in the first year.

4. Use the greater of the Census or DHS counts but limit dollar increases to a certain percentage for a certain number of years. For example, if increases were limited to 20% per year for five years, the cost to the State in the first year would be $87.4 M. No districts would lose funding under the poverty grant portion of General State Aid. Lower percentage increases would lower the State cost.

5. Use the DHS count and change the formula to a minimum amount per student plus an additional amount based on the district concentration. Let districts use the greater of their current poverty grant or the new calculated grant. For example, if increases were limited to 20% per year for five years, the cost to the State in the first year would be $27.7 M. No districts would lose funding under the poverty grant portion of General State Aid. The use of the current grant could be phased out over time.

6. Since Chicago dominates the poverty count and grant, the last alternative is to somehow treat Chicago separately. This could mean either providing them with a separate dollar amount for each of the DHS count or simply providing a set grant amount.