Reading Fertility Stalls in Census Data: The Case of Kenya

Michel Garenne, Institut Pasteur
Clifford O. Odimegwu, University of the Witwatersrand
Robert McCaa, University of Minnesota
Sunday A. Adedini, University of the Witwatersrand and Obafemi Awolowo University
Garikayi Chemhaka, University of the Witwatersrand

Fertility stalls are defined as periods of several years without fertility decline after a period of steady decrease in fertility rates. Several cases were documented in the 1990’s and early 2000’s in Africa using DHS data, one of them being Kenya. When fertility rates stop declining, the size of birth cohorts tends to increase, leaving a trace in the age pyramid. Conversely, the size of birth cohorts in successive censuses allows one to reconstruct past trends in the General Fertility rate (GFR). This study will illustrate this point using three censuses from Kenya (1989, 1999, and 2009), all available as samples in the IPUMS-International database. The three censuses allow one to compute levels and trends in GFR for the 1980-2008 period, to identify periods of monotonic decline and periods of stalls. They will be compared with levels and trends in period estimates of TFR obtained from DHS data using maternity histories. All calculations will be done for urban and rural areas separately. The paper will illustrate the use and limitations of IPUMS samples of census data for identifying fertility trends and reversals in a long term longitudinal perspective.