Summary: In response to a congressional request, GAO examined a Department of Agriculture (USDA) demonstration project that tested the distribution of cash payments and commodity letters of credit (CLOC) in lieu of donated USDA commodities at selected school districts participating in the National School Lunch Program (NSLP). Specifically, GAO determined the: (1) appropriateness of the project methodology USDA used; (2) reliability of data supporting the project's findings; and (3) cost-effectiveness of implementing procedures.
GAO noted that the final USDA reports of the demonstration project: (1) compared the cash, CLOC, and commodity systems in areas of interest related to the project, such as operating costs, student participation, and the nutritional value of school lunches; and (2) indicated that there were no or minor differences in the ways that the three systems affected the areas of interest. GAO found that: (1) the methodology that USDA used to select school districts to participate in the demonstration project limited the validity of the study results and did not provide for highly confident statistical generalizations; (2) USDA may have overstated labor and storage costs for cash and CLOC participants; and (3) the USDA decision to distribute commodities to cash and CLOC participants made a year of the project's data unusable and caused a 1-year extension of the project. GAO concluded that, since the project had limitations and weaknesses, USDA should use any nationwide generalization of the project's results with caution.
