Page 60 - IJPS-11-3

International Journal of
Population Studies Age patterns of fertility in Ethiopia

Demographic and Health Surveys (EDHS) conducted in for single-year ages of women aged 15 – 49 across all four
2000, 2005, 2011, and 2016. Due to the absence of a reliable EDHS surveys reveal considerable age preference in the
vital statistics system and recent census data in Ethiopia, terminal digits 0 and 5.
many researchers rely heavily on EDHS data. However, as Among the indicators provided by the EDHS surveys at
EDHS surveys depend on various sampling procedures, the regional level, TFRs are observed as plausible estimates
their fertility and mortality indicators are subjected to by many researchers. However, ASFRs obtained from the
sampling errors and other statistical biases (Central EDHS are sometimes under-estimated or over-estimated.
Statistical Agency [Ethiopia] & ICF, 2012, 2016; Central In addition, ASFRs for the extreme age groups of 15 – 19
Statistical Authority [Ethiopia] & ORC Macro, 2001, 2006). or 45 – 49 are occasionally reported as zero due to the
TFR, a widely used summary measure of fertility, small sample size of live birth data. This limitation leads
represents the average number of children per woman and many researchers to restrict their studies to the extent
is calculated by simply summing the age-specific fertility possible. Observing these drawbacks, this study employs
rates (ASFRs) of various age groups. ASFRs are generally an innovative approach to derive ASFRs from a model
reported for seven successive age groups: 15 – 19, 20 – 24, fertility table (MFT) developed following the study of
25 – 9, 30 – 34, 35 – 39, 40 – 44, …, and 45 – 49. In some Mitra (1965).
African countries, ASFRs may include younger (10 – 14) MFTs by Mitra provide ASFRs corresponding to general
or older (50 – 54) age groups due to the observed births in fertility rate (GFRs), which were developed using a set of
these ranges. Births below age 15 or above age 50 are often simple linear regressions between ASFRs and GFRs. These
added to adjacent groups for the calculation of ASFRs. tables are observed to be similar to the model mortality
For example, births taking place in women above age 50 tables. Given the GFR, one can determine the ASFRs for
are added to the 45 – 49 age group for ASFRs calculation. different age groups. Thus, seven ASFRs can be obtained
While ASFRs provide a detailed fertility profile, TFR offers from a given GFR for a specific population at a particular
a concise summary. Whenever feasible, it is preferable to time (Mitra, 1965).
analyze ASFRs instead of TFR or any other indicator. The
calculation of the Net Reproduction Rate further requires Over the years, researchers have made numerous
information on survivors from the relevant life tables. In attempts to indirectly estimate various fertility and
addition, ASFRs can be reported for individuals ages 15, mortality indicators at both national and regional levels to
16, 17, …, and 54, although this is a rare practice nowadays understand the levels, trends, variations, and determinants
of fertility in Ethiopia and its regions. Censuses and survey
(Siegel & Swanson, 2004).
data from the past few years served as the main data
Ethiopia has shown a decline in TFR from 6.4 in 1990 sources in all these attempts. However, their use and policy
to 4.6 in 2016 (Central Statistical Agency [Ethiopia] & implications are observed to be limited nowadays (See:
ICF, 2016; Central Statistical Authority [Ethiopia] & ORC Teklu et al., 2013).
Macro, 2001). Despite this declining trend, fertility remains
high and exhibits significant regional variation. According 1.2. Modeling ASFRs
to the 2016 EDHS, for instance, the TFR varies from 1.8 in Historical evidence shows that ASFRs are observed to be
Addis Ababa to 7.2 in the Somali region (Central Statistical low in the 15 – 19 and 20 – 24 age groups, comparatively
Agency [Ethiopia] & ICF, 2016). The same report also high in the 25 – 29, 30 – 34, and 35 – 39 age groups, and
indicate that the Afar and Somali regions experienced an then decrease in the 40 – 44 and 45 – 49 age groups.
increase in TFR during the period from 2000 to 2016. This pattern is observed universally across populations
The TFR derived from EDHS is susceptible to both due to the influence of biological, socio-economic,
sampling and non-sampling errors, a common issue in environmental, and other background factors on women’s
many developing countries. The TFR sampling errors for fertility. Theories such as Davis-Blake’s theory of fertility
Ethiopia and its regions are detailed in the “estimates of and Bongaarts’ proximate determinants theory provide
sampling errors” part in Appendix B of the EDHS reports, substantial evidence supporting this trend (Bongaarts,
highlighting precision issues in TFR estimates. Notably, the 1978, 1982, 2015; Bongaarts & Potter, 1983; Davis & Blake,
standard errors at the regional level are generally higher 1956; Mitra, 1967; United Nations, 1963).
than at the national level, primarily due to smaller sample ASFRs exhibit a particular shape and pattern, similar
sizes. The quality of fertility data related to age preference to age-specific death rates, and are often modeled by
and birth date reporting (affected by non-sampling errors) researchers using various mathematical curves (Gaire
is provided in the “data quality tables” in Appendix C of et al., 2022; Gogoi & Deka, 2023; Islam & Ali, 2004; Islam,
the EDHS reports. Simple line graphs (not reported here) 2009; Jena et al., 2023; Mishra et al., 2017; Wani et al.,

Volume 11 Issue 3 (2025) 54 https://doi.org/10.36922/ijps.4086