Small Area Population Projections: Suitability of Hamilton-Perry Model in Undertaking Population Projections by Age and Sex for Selected Sub-Counties in Kenya.

Abstract

Background: Devolution of services in Kenya from the National Government to the 47 County Governments has led to a rise in demand for reliable age and sex disaggregated population statistics at lower levels (sub-county and ward levels) for effective segmented programming and planning. Objective: The study was undertaken to determine the suitability of Hamilton-Perry model in undertaking sub-county population projections by age and sex using case of selected sub-counties in Nyeri and Vihiga Counties, Kenya. Method: The study used the 1999, 2009 and 2019 Kenya Population and Housing Census data for the selected sub-counties.  Vihiga county was selected because  it had the highest net out migration rate among the 47 counties of 97 per 1,000 in 2019 census while Nyeri was selected due to its stable Total Fertility Rete (TFR) and low out migration rate. Results: The Mean Absolute Percentage Error (MAPE) for the 2019 population projections for Nyeri county, Kieni East and Kieni West Sub-counties were 8.5, 11.5 and 17.5 respectively. That of Vihiga County, Emuhaya Sub-county and Vihiga Sub-county were 9.6, 19.3 and 7.9 respectively.  This implies that the projection had very good precision for Nyeri County, Vihiga County and Vihiga Sub-county and had good precision for Kieni West, Kieni East and Emuhaya Sub-counties. Conclusions: The study results from the two sub-counties of Nyeri and Vihiga Counties, showed that the HP population projection model can be used reliably to undertake population projections for the sub-counties of Nyeri and Vihiga Counties, Kenya. The model produced population projections with very good precision in the Sub-counties with higher population size and its precision was shown to decrease with decrease in population size.