| Sumario: | Three alternative statistics for constructing a social index to distinguish geographical areas are considered. The relative merits of the signed χ2 statistic, the percentage, and the raw count are examined in relation to two distinct goals in resource allocation: identification of concentrations of need (geographic targeting) and estimation of local budgets (resource distribution). It is argued that the signed χ2 statistic overcomes disadvantages of the percentage when identifying high concentrations of need. Disadvantages of the signed χ2, relating to its construction, its dependence on the choice of a reference area, and its ability to deal with areas of low concentration of need, are identified and discussed. Alternative measures of geographical concentrations of need are outlined.
|
|---|