PREDICTING INDICATORS AT SMALL SCALE USING ENTROPY ECONOMETRICS

Statistical information for empirical analysis is very frequently available at a higher level of aggregation than it would be desired. Economic and social indicators by income classes, for example, are not always available for cross-country comparisons, and this problem aggravates when the geographical area of interest is sub-national (regions). In this paper we propose entropy-based methodologies that use all available information at each level of aggregation even if it is incomplete. This type of estimators have been studied before in the field of Ecological Inference. This research is related to a classical problem in geographical analysis called to modifiable area unit problem, where spatial data disaggregation may give inaccurate results due to spatial heterogeneity in the explanatory variables. An empirical application to Spanish data is also presented.