Chapter 6: Does the Measure Matter?: The Sensitivity of Cost Function Estimates to the Choice of Measure of District Poverty

As policy changes and research findings have raised concern over the validity of free- and reduced-priced lunch (FRPL)-based measures of student disadvantage, questions have emerged about the validity of the existing adjustments in most state aid formulae for the share of disadvantaged students and other determinants of district costs. In light of these questions, this chapter explores the sensitivity of estimated cost adjustments in the cost function literature to the choice of measure of student hardship. We find that, while the pattern of estimates does not change as we change measures, the estimated cost differentials attributable only to the disadvantage measure vary substantially depending on the type of measure used to convey social or economic disadvantage. But we also observe, however, that when the cost differentials associated with the disadvantage measure are larger, the cost differentials associated with other determinants of cost are smaller. The reason why overall estimated cost differentials probably do not vary much when percent FRPL is replaced with an alternative measure is that percent FRPL may be proxying for other determinants of costs that contribute to material hardship. We argue that, as we move away from percent FRPL as the measure, it becomes that much more imperative to account for and adjust for a richer set of student attributes that influence costs.