Simulating Sudden Refugee Influx and Its Impact on Demographic Structure: the Korean Case

Many developed countries of the world such as Japan, Korea, Singapore, Germany etc., are facing the issue of decreasing birth rates and increasing aged populations. Better healthcare services result in increased life expectancies, and change in preferences of younger generation lead to delayed marriages and childbirths. This problem could lead to building up of tremendous pressure on ever decreasing work force to support the ever increasing aged population. One of the potential solutions for this problem is liberal immigration and refugee laws. Korea has stringent immigration laws and most of the immigration into the country is temporary in nature. However, in the recent years, there has been noteworthy increase in international im-migration into Korea and there has been an increasing demand to relax the immigration laws. Due to wars and political circumstances in the middle-east, we have witnessed exodus from these countries to many European countries. Such a phenomenon could have lasting impacts on the host country. Following this cue, we simulated the rapid influx of people seeking refuge in Korea. In this particular simulation test case, we observed the change in demographic distribution of Korea.

We built a general purpose agent based model for Korean demography, using AnyLogic. This simulation test case is one of the modules of the bigger model. We made use of real census data from Korean Statistical Information Service. Using this model, we simulated a steady influx of a total of one million immigrants into Korea over a course of two years. All the agents in the simulation have characteristics like education, age, gender, marital status, desired and current number of children, income, etc. It must be noted that the proposed model is fully dynamic in nature with all the agent characteristics evolving with every simulation tick. A simulation tick corresponds to one year and simulation is run for 70 years (from 2005). Fertility and immigration events add agents, whereas agents are removed by mortality event. All the demographic activities are governed by real data and are not assumed randomly. The agents in the simulation also interact with each other and influence behavior (partic-ularly desired education, desired children). These interactions thus also influence immigrants’ behavior and that of the native population too. This whole functioning of model delivers us interesting insights and results.


Young, Working and Aged Population in Demographic Simulation Model

Dashboard in the patient flow simulation model

Related posts