ABSTRACT
We estimate age-province-specific birth rates for Cambodia in 2010. The 2008 census provides large cell sizes even after disaggregating by province and age of mother. However, it suggests implausibly low levels of fertility. The Demographic and Health Survey (DHS) produces data of high quality. The DHS for 2010, however, has a sample size only 100th as large as the 2008 Census. We show how the framework of Part IV can be used to estimate birth rates for Cambodia, capitalizing on the large sample size of the 2008 census and the accuracy of the 2010 DHS.
Our overall model contains a system model to describe demographic behaviour, and data models to describe the measurement processes. The system model relates true birth counts to exposure, where exposure is treated as known. The first data model describes the relationship between the true births counts and the census data. The second data model describes the relationship between the true births counts and the DHS data.
We jointly estimate the true birth counts, birth rates, and census coverage ratios. We illustrate how to use information on demographic rates and coverage ratios, plus some demographic intuition, to iteratively check and improve our models.
