» Events » Endogenous Diffusion Model of Entry into First Marriage: Theoretical and Empirical Progress in Nuptiality Models
Endogenous Diffusion Model of Entry into First Marriage: Theoretical and Empirical Progress in Nuptiality Models
Describing the age schedule of cohort-based first marriage process is an important demographic issue. The focus of this thesis is to develop and utilise nuptiality models for demographic analysis. The proposed nuptiality model is illuminated by Hernes’ (1972) seminal model of marriage diffusion and constructed in terms of ordinary differential equation(s). The modelling strategy adopts a social diffusion framework to describe the age schedule of first marriage from a one-sex and cohort-level perspective. The age-specific hazard rate of entry into first marriage is assumed to be governed by two opposite diffusion-driving forces named as “social pressure” and “marriageability”. The model’s main feature is the “endogeneity” property, which means that the hazard rate is completely specified by a functional of the first marriage process itself. This feature makes the theoretical model applicable to fit a variety of population’s first marriage experiences in reality, and the influences from some historical external factors can be eliminated due to the endogeneity property. The endogenous marriage diffusion model is both theoretically and empirically compared with the two classic nuptiality models – the Coale-McNeil model (1972) and the Hernes model (1972). The new model stands out for the reasons that it can both provide a theoretical implication on social behaviours with formal mathematical language and accurately account for observed data from a variety of real populations in different historical and geographic settings. By decomposing the observed process, the model can reveal the dynamic nature of the cohort’s first marriage behaviours and provide a new behavioural interpretation on this event. New concepts can be defined based on the model, such as the “intensity” of social pressure increase and the “degree” of marriageability depreciation. Moreover, the model can also provide the mathematical relationships between some vital statistics of first marriage (e.g., maximal first marriage frequency or maximal hazard rate of first marriage) and the model parameters.
Xiaoguang Jia is a PhD candidate in the School of Demography, with research interests in the decision-making processes of life course behaviours. His research experience is mainly on the mathematical demographic modelling, in particular, for marriage formation phenomena. Xiaoguang holds a Master of Science (in Operations Research and Cybernetics) and a Bachelor of Science (in Information and Computing Science) from Fudan University.