Acknowledgments

This project, led by the World Bank Development Data Group, was funded by the World Bank Knowledge for Change Program (KCP III, grant number TF0A1095) and by the International Household Survey Network Trust Fund from the United Kingdom Departmenf for International Development (DFID, grant number TF011722 executed by the World Bank Development Data Group).

The project aimed to demonstrate how new approaches—synthetic data generation and dynamic micro-simulation—can take advantage and integrate data from multiple sources—in this case population censuses and household surveys—to inform development policies and research on development issues.

This report describes a micro-simulation population projection model applied to Mauritania. The model will be expanded and adapted to other countries. A specialized education module, more advanced than the education module currently available in the model, is being developed and will be integrated in the model. Solutions to exploit geo-spatial data will be developed and integrated in the model, allowing fine geographic disaggregation of the model’s input and output. And other specialized modules may be developed. This first application to Mauritania is thus the first of what we envision to become a collection of models to be developed and published as public goods under the generic name DYNAMIS (Dynamic Micro-Simulation).

The project was initiated and led by Olivier Dupriez, Lead Statistician at the World Bank Development Data Group. Martin Spielauer, expert in micro-simulation and consultant for the World Bank, is the main author of the model. Alexander Kowarik, statistician and World Bank consultant, contributed to the project as an expert in synthetic data generation. Staff from the Centre Mauritanien d’Analyse de Politiques (CMAP) and from the Office National de la Statistique de Mauritanie (ONS) participated in the country adaptation of the model, in the provision and analysis of input data, and in the analysis of the model output). The national counterparts in this research project included El Hassen Zein (Coordinator, CMAP), Cheikh Baye Beddy (Deputy Coordinator, CMAP), Mahjouba Habib (Junior expert, CMAP), Meimouna El Moustapha (Econometrician, CMAP), Yahya Menaya (Econometrician, CMAP), Oumelkhairat Abba (Computer expert, CMAP), Mohamed El Moctar Ahmed Sidi (General Manager, ONS), Taleb Mahjoub (Deputy General Manager, ONS), El Yass Didi (Director, Social and Demographic Statistics, ONS), Mariem Mohamed Saleh (Director, Dissemination and data processing, ONS), Boubekrine Mohamed Kabir (Statistician-demographer, ONS), and Hachim Mohamed Maaloum (Data processing specialist, ONS). Mehmood Asghar (World Bank) provided IT support. The report was copy-edited by Linda Klinger.

The model was developed using MODGEN (a freeware developed and published by Statistics Canada) and Microsoft Visual Studio. Data analysis for the calculation of the model parameters was done using Stata, and a synthetic population dataset (not used in the final version of the model) was produced using the simPop R package maintained by Matthias Templ. The html documentation of the model was built with Sphinx using a theme provided by Read the Docs.