Is Fertility in sub-Saharan Africa That High, or is it an Artefact of Survey Eligibility Criteria?

Ashira Menashe-Oren , Tel Aviv University
Stéphane Helleringer, University of New York, Abu Dhabi

Nationally representative household surveys are relied on to estimate fertility, and have documented fertility stalls in many sub-Saharan African countries. Yet such survey estimates of fertility may be misleading if the survey design and management influences the fertility outcomes. It is possible that, in defining who to include in surveys, that is, who is eligible for interviews, distorts fertility estimates. Notably, interviewers may manipulate the age of respondents to reduce their burden, moving respondents outside of eligibility boundaries. We aim to assess whether age eligibility criteria in surveys distorts fertility estimates. We compare fertility across two types of surveys with different age eligibility criteria: Demographic and Health Surveys (DHS) where women are interviewed between ages 15-49, and Population-Based HIV Impact Assessment (PHIA) where women are interviewed from age 15-64 or 15+. Across five African countries these surveys were held within less than a year of each other, thus fertility estimates should be very similar across PHIA and DHS, unless survey design and management differed. However, we find that with stricter eligibility criteria, in the DHS, age manipulation is more prevalent, leading to distorted age structures and higher fertility rates. Robustness checks of comparing to other surveys, using alternative indirect estimates of fertility, and zooming in on a sub-population expected to report more accurately on ages and dates, all confirm our results. Lower fertility estimates from the PHIA surveys across all countries suggest that the projected population size of African countries may be much lower than currently anticipated.

See paper

 Presented in Session 68. Dealing with Incomplete or Deficient Data from Surveys and Censuses