This study uses empirical analysis of the school lunch and breakfast programs as it has been implemented since 1972 in 28 north Alabama school districts to seek answers to a number of questions concerning the effects of an administration proposal to provide block grants to the states to design and implement their own child nutrition programs. The analysis is guided by the hypothesis that local resources, social structure, and values can and do have significant supportive and perverse effects on participation in mandated federal programs. Multiple regression analysis relates class structure variables (middle class, coalition of poor, and power elite) with the covariates of poverty rate, family size, and total lunches served. Inflation and, in particular, increased food prices since 1972 underpin the increasing explanatory power of the class-structure variables. Most important for the benefit of the poor is the fact that participation corresponds more closely with poverty rates in 1974 than in 1972. (Author/IRT)