Answer

Possible Ways to Improve the Demographic Evolution Model

Population system – a third component/subsystem of urban systems, after transport and land use – is composed of three key elements; namely, households, social networks, and persons with their demographic attributes. Urban planners and other actors have put forth two approaches to incorporate these elements in population modelling. The two approaches are static population modelling, which simulates a virtual city and the population and facilities therein; and the dynamic population modelling, which simulates the manner in which a city’s population evolves over time. The first model under the dynamic population modelling category is the demographic evolution model, with the second one being the social network evolution. This discussion post offers a suggestion on how the demographic evolution can be improved, in terms of effectiveness and scope.

As the name implies, the Demographic Evolution model simulates various demographic transitions of persons and households through life stages. Under the model, we have ten sub-models that are interlinked: the age, income, birth, death, marriage, education, divorce, immigration, emigration and employment models. One of the ways that we can improve the demographic evolution model is by the making the income sub-model more holistic. Currently, the model only considers the income of agents (individuals) in formal employment then applies the Gaussian distribution to generate the change in income over a given time. To make the model more accurate and holistic, adjustments should be made such that it takes into account the income of the self-employed and small business owners or entrepreneurs. Another area where the demographic evolution model can be improved is the use of data. Currently, the model is based explicitly on the random selection of agents from a population, and then using the agents to forecast future trends in population. This makes the results of the model more superficial and inaccurate. To overcome this shortcoming and avoid overestimation, the model should be interfaced with real-world data sources such as the city’s registrar of persons, the census agencies, and tax collection department among other sources.