A Customizable, Modular System

DYNAMIS-POP is a customizable and open source modular modeling platform. Modules can be adapted and extended, and new modules can be added.

Currently available modules:

• Starting Population: The model starts from a baseline population. Records can be weighted, and users can choose the size of the simulated population. All output is automatically scaled to the actual population size.
• Fertility: Model users can choose between two fertility modules. The base one corresponds to the widely used macro model DemProj with parameters and projection scenarios available online for most countries. The refined one models fertility separately by parity and includes a more detailed list of variables than just age. When using the refined module, the user can choose to align results to the macro model thereby reproducing the same aggregate results of existing projections while producing realistic life-courses concerning number, timing and spacing of births by characteristics like partnership status and education.
• Mortality: Again, we provide two model versions: a base version resembling DemProj which is based on projected life tables, and a refined model version. The refined model focuses on infant mortality by mother’s characteristics. When using the refined model, the user can align results to the base model to produce the same aggregated outcomes in the number of deaths for an initial year.
• Migration: The model includes modules for internal migration, emigration, and immigration. All modules currently follow typical macro approaches. Internal migration is based on age-specific origin-destination matrices. The model distinguishes three geographical layers: national, regional, and provincial.
• First union formation/marriage: First marriage is implemented by fitting parametric Coale & McNeil models by education attainment and birth cohort. The parameterization of such a model is very intuitive; parameters are the minimum age at first union formation, the average age, and the proportion of women who will eventually marry.
• Primary education: We provide two model versions. The base model is a simple “fate model” assigning the basic outcome (never / ever entering, graduation) - by birth cohort, sex and district. A refined version models the inter-generational transmission of education accounting for the difference in education outcomes by parental background. We also model the progression through the education system tracking students by grade including grade repetitions and interruptions of studies.
• Secondary education: The module follows a period approach parameterized by flow rates like take-up, repetition, and success rates.
• School resource planning: Tracking students through the education system by grade allows calculations like the number of required classrooms and teachers on a sub-national level. The model supports easy scenario creation by setting target student to classroom and student to teacher ratios.
• Ethnicity: Ethnicity is inherited from the mother (applying a transition matrix) and sampled from a distributional table (by sex and destination) for immigrants.
• Prenatal care: The model for the receipt of prenatal care accounts for education, regional differences, the effect of very young age of the mother, and trends.
• Immunization: Child vaccination is modeled based on individual characteristics including mother’s education and the receipt of prenatal care. The model also accounts for ethnicity, region, the effect of very young age of the mother and supports easy scenario creation.
• Human Capital Index: This module calculates the World Bank Human Capital Index (HCI) on a sub-national level and projects it over time. The module draws on the detailed modeling of child mortality and education and adds necessary components for calculating the index to the model, namely stunting and preschool attendance.