Key points are not available for this paper at this time.
Approaches estimating wanted fertility typically depend upon responses to survey questions on desired family size, wanted status of recent births, and desire to continue childbearing. These responses and dependent approaches are, however, typically upwardly biased in measuring wanted fertility. The latter of these 3 approaches, women's desires to continue childbearing, is the least biased of standard preference measures, and is proposed as the basis of a new indirect method of estimating wanted fertility. Existing approaches are reviewed, followed by a description of this proposed model of securing more accurate estimates of both wanted and unwanted components of total fertility. 2 hypothetical applications are conducted to help explain the procedure's logic. The method is then applied to data from 35 World Fertility Survey and 13 Demographic and Health Survey developing countries, and finds results indicating an average 26% of fertility to be unwanted. With wanted fertility ranging from 98% in Senegal to 51% in Peru, average unwanted fertility is substantially higher than that estimated using other approaches. The proportion of unwanted fertility was also found to vary systematically over the course of fertility transition, with lowest levels at beginning and end, and highest among countries in mid-transition. Considering perfect birth control, the potential role of birth spacing to contribute to quantitative control is mentioned.
John Bongaarts (Sat,) studied this question.