Demography, Vol. 28, No.2, May 1991
Individual-Level Sterility:
A New Method of Estimation with Application to
Sub-Saharan Africa
VIla Larsen
Department of Sociology
SUNY Stony Brook
Stony Brook, NY 11794-4356
Jane Menken
Department of Sociology and Population Studies Center
University of Pennsylvania
3718 Locust Walk
Philadelphia, PA 19104-6298
This paper extends work on measures of population proportions sterile to propose a
new estimator of an individual woman's age at sterility and consequently her sterility
status at given ages. Accuracy and reliability, examined in a simulation study, appear
satisfactory. From World Fertility Survey data for five African countries, the
proportions sterile by age estimated by the individual measure and by the population
estimator are almost identical. Cameroon and Kenya show substantial variation in
prevalence and incidence of sterility across ethnic groups and by number of
marriages. Unexpectedly, the evidence suggests that sterility increased from 1960 on
in Kenya and remained unchanged in Cameroon.
In earlier work (Larsen and Menken 1989) we developed a method for estimating
sterility in a population from a sample of women, not all of whom had reached the end of
their reproductive span. Neither that procedure nor those from which it was derived (Henry
1965; Leridon 1977; Trussell and Wilson 1985; Vincent 1950) measure the sterility status of
a particular woman; rather, they estimate the age-specific proportions sterile. Currently
available individual-level methods are unsatisfactory for a variety of reasons described in
our earlier paper. Yet if a woman's sterility status at any given age and her age at the onset
of sterility could be determined with reasonable accuracy, the study of the determinants and
the differentials in sterility both within and between populations would be feasible, and
sterility could be treated as a predictor that is related to, or affects, dependent variables of
interest. Therefore we attempt the extension of our earlier work to the individual level when
information is based either on complete reproductive histories (when all women are
followed until age 50) or on incomplete histories (such as those obtained from women of
reproductive age on a particular date when a survey is taken).
This effort is particularly timely; increasingly, data are becoming available from
populations showing evidence of high levels of sterility, such as those of sub-Saharan
Africa; there, for example, more than 20% of women aged 25 in Cameroon are sterile
(Larsen and Menken 1989), and childlessness is common in various populations, such as
those of Ethiopia (Mammo and Morgan 1986) and Zaire (Tabutin 1982). Surveys
Copyright © 1991 Population Association of America
229