[A local governmental welfare function is specified to explore two of its central characteristics: the equity-productivity tradeoff and differential weights across neighborhoods. The model is estimated using service outputs (safety) in the welfare function, as opposed to publicly provided inputs (police), over neighborhoods. The equity-productivity tradeoff is found to be considerable, and not all neighborhoods are weighted equally. The estimation results raise several questions about accepted analysis of governmental behavior.]

A local governmental welfare function is specified to explore two of its central characteristics: the equity-productivity trade-off and differential weights across neighborhoods. The constrained maximization model is estimated using service outcomes (safety) in the welfare function, as opposed to publicly provided inputs (police), over neighborhoods. The equity-productivity trade-off is found to be considerable, and not all neighborhoods are weighted equally. The results show that inequality aversion and unequal concern by local government over service outcomes must be addressed explicitly to understand the observed distribution of publicly provided inputs, with important implications for standard analysis of local governmental behavior.

The World Bank and others maintain that the major mechanism for improving nutrition in poor communities is increases in income. Aggregate estimates of food expenditure are consistent with such a possibility, implying income/expenditure elasticities close to one. However, the high degree of aggregation at which such estimates are made means that the considerable increase in price per nutrient as income increases is ignored, and the nutrient elasticities are therefore overstated. Estimates for a rural south Indian sample indicate that this bias is considerable and that the true nutrient elasticities with respect to income may be close to zero.