Where will a state’s population actually be in twenty years?
Client
Place
Practice
Services
Year
Schools, hospitals and roads are committed decades ahead of the people who will use them, and small areas are the hardest places to forecast.
Birth rates, migration patterns and economic conditions were modelled together to project population change across SA2 areas, local government areas and custom zones, with machine learning used to describe migration within the state.
Forecasting is hardest where it matters most: in the small areas that are growing fastest.
Working with demographer Dr Tom Wilson, Geografia delivered a Python-based forecasting tool that models the effect of economic conditions, housing development and infrastructure investment, and lets the department test scenarios for infrastructure, housing and economic planning across the state.
