The spatial-temporal patterns of fertility transition in Belgium (1886-1935)

Rafael Costa, Université Catholique de Louvain

The purpose of this study is to investigate the underlying mechanisms of spatial-temporal patterns of fertility transition in Belgium. The analyses lie in a new dataset compiled from Belgian historical sources, offering a very high level of spatial and temporal detail. In fact, the dataset provides yearly demographic indicators at the municipality level from 1886 to 1935, as well as cultural and economic information at the district and municipality levels. The yearly dataset allows us to treat municipalities as individuals followed up in time year by year, like a longitudinal dataset. Based on this idea, a macro-level discrete-time event-history model is constructed. The purpose of the model is to test the effect of several covariates at time t-1 on the probability, for a given municipality, of starting its fertility transition at time t. The dependent variable is the municipalities' year of onset of fertility transition, defined as the year of an irreversible 10% drop of Coale’s marital fertility index Ig. Covariates include economic, demographic and cultural indicators, as well as a ‘propagation’ variables accounting for the distance and commuters to pioneering municipalities. The use of longitudinal methods on the aggregate dataset has the advantage of taking into account both the temporal and the spatial dimensions of fertility transition.