摘要：That higher levels of national income per capita are associated with better health conditions and lower fertility is more than evident; recent analyses provide quantification of the magnitude of the effects and adduce strong evidence for its causal nature (at least for health). This paper extends prel'ious analyses by assessing regional and temporal differences in the income - health/fertility relationship (differences are pronounced) and by using deviations from this estimated relationship it constructs measures of performance at five year intervals from 1960 to 1990 for 19 countries of Latin America and the Caribbean. Performance for each country on under-5 mortality and total fertility rates. relative to other Latill American countries and controlling for income levels. Changes in performance, from one five-year period to the next, are highlighted in order to pose questions about the response of performance to changes in policy. Some results are expected (e.g. the relatively good performance of Colombia and Costa Rica for much of the period); other results are surprising (e.g. the sharp deterioration in performance of Argentina and Chile in the late 1980s). Health expenditures are then used to predict performance levels in 1990; for under-5 mortality rates, public sector expenditures on health are found to have a significant beneficial effect. The magnitude of the estimated effect is substantial: Over 10,000 child deaths in 1990 would have been averted in Latin America if public sector expenditures on health increased from 3.5% to 4.0% of CDP while holding total expenditures constant.
