Prenatal and Infant Health Indicators

Spurred by our National Educational Goals, as well as by the economic woes of school dropouts, wehave improved the conceptualization and measurement of high school dropout. Moreover, greater public andprivate resources have been devoted to development and dissemination of temporally and spatially comparableindicators of dropout. These may help improve both our understanding of school-leaving and the allocation ofresources to prevent or remediate the effects of dropout. At the same time, indicators of dropout are sometimesweak, misleading, or excessively aggregated. Recent changes in the Census concept of educational attainmenthave in some ways madeit more difficult to identify high school dropouts, and many indicator series fail toidentify proximate sources of school-leaving. At the national level, improvement should focus on timelymeasurements for major social and economic groups that will add theoretical understanding to indicator timeseries. At the state level, there are painful trade-offs between timeliness, specificity, and validity; model-basedestimates may lead to improvement over present practice. Model-based estimates also offer possibilities forcreation of risk-adjusted series.