Around two-thirds of Australia’s population growth can be attributed to the growth of the capital cities: Sydney, Melbourne, Brisbane, Adelaide, Perth, Hobart, Darwin, and the nation’s capital, the Australian Capital Territory (ACT). Australia’s high level of urban living, low fertility, and distinctive urbanisation experience compared to other developed countries make Australia’s capital cities interesting and important to examine. How much do we know about the population of Australia’s capital cities and their components? What are the drivers of fertility and mortality of Australia’s capital cities? Have they changed over time and how do they differ between cities?

This thesis examines the role of socioeconomic structure on fertility and mortality of Australia’s capital cities using Bayesian approach. Socioeconomic structure in this thesis refers to the profile of the population based on age, education, employment, household characteristics, income, marital status, and occupation. This thesis explores the use of Bayesian approach in explaining total fertility rates and standardised death rates of Australia’s capital cities using socioeconomic data obtained from the 2006, 2011, and 2016 population censuses of Australia.

This thesis makes an original contribution to the literature on Australia’s capital cities by revisiting the demographic changes in these cities using geographically comparable macro-level and hierarchical data. Most studies conducted on related topic are at national level and rarely at sub-national level or used macro-level data. In addition, previous studies use data that are not based on the same geographic classification. This thesis uses the Australian Statistical Geography Standard (ASGS) developed by the Australian Bureau of Statistics making the data comparable over time. The findings from the Bayesian regression models show that socioeconomic structure has a key role in fertility and mortality of Australia’s capital cities (except for Hobart (fertility and mortality) and Darwin (mortality), and its role varies between cities and over time.

Estimating fertility and mortality using socioeconomic structure data from population census has merit, especially in countries where the quality and coverage of vital statistics are not good. Bayesian mixed regression model fits the data in estimating total fertility rate (in larger cities), while Bayesian OLS regression model is sufficient in estimating standardised death rate. Furthermore, the use of Bayesian approach in producing demographic estimates and projections with attached probabilities have benefits and is highly recommended.

Tita Tabije is a part-time PhD Candidate at the School of Demography, holds a degree in Master of Arts in Demography from the ANU and Bachelor of Science in Statistics from the University of the Philippines. She worked as Senior Statistical/Data Analyst, Demographic Specialist, SAS Programmer/Developer at different government offices, including Australian Bureau of Statistics (17 years), Philippine National Statistics Office (13 years), Australian Electoral Commission (3) years, Department of Human Services, Calvary Public Hospital.

She also taught as an Assistant Casual Sessional Academic at the School of Sociology at ANU and Assistant Professor at the Sociology Department of Dela Salle University. Tita has been a registered migration agent in Australia since 2007 after completing the Graduate Certificate in Australian Migration Law from the ANU.