Can Spatial Diffusion Process explain Family Changes? An Analysis of Non-Marital Births over 50 years in Belgium

Yoann Doignon, French National Centre for Scientific Research (CNRS)
Adrita Banerjee, French National Centre for Scientific Research (CNRS)
Amira Hamieh, French National Centre for Scientific Research (CNRS)
Oumar Mbath, French National Centre for Scientific Research (CNRS)

Over the decades, many countries in Europe has witnessed significant family changes, marked by declining marriage rates, increasing cohabitation, delayed childbearing and persistently low fertility. The Second Demographic Transition (SDT) has been a key framework for understanding these shifts, with alternative theories focusing on economic, cultural, and gender-related factors. Despite extensive research, the spatial dimension of these changes remains understudied. This paper addresses this gap by examining the spatial dynamics of non-marital births in Belgium over the past 50 years at the municipal level. We investigate whether these changes can be attributed solely to socio-economic factors or if a spatial diffusion effect also plays a significant role. Using a Dynamic Spatial Durbin Model, we find that non-marital birth rates in a municipality are influenced by rates in neighbouring municipalities from previous periods, indicating a significant spatial diffusion effect. This finding highlights the importance of geographical proximity as a key determinant of family change, alongside socio-economic factors. Furthermore, the influence of explanatory variables shifts over time, suggesting that different theoretical frameworks are applicable at various stages of the diffusion process. Our results emphasize the need to integrate multiple theoretical perspectives to better understand the family changes.

Keywords: Spatial Demography, Neighbourhood/contextual effect analysis, Econometrics , Fertility

See extended abstract.