In this project we worked with the Thailand National Statistical Office (TNSO) and United Nations Population Fund (UNFPA) to provide technical assistance and training to TNSO staff on modelled population estimates.

As a global leader in developing bespoke statistical methods for modelled population estimates we provide support to various national statistical offices in training and co-production of high-resolution gridded population estimates from existing data sources (e.g., household surveys, building footprints, administrative records, census projections). 

Thailand National Statistical Office (TNSO) carries out a population and housing census every ten years, but due to the Covid-19 pandemic the population and housing censuses planned for 2020 are delayed until 2025 or 2026. Postponement posed unique challenges to data quality and census enumerators’ workloads and to address the difficulties we worked with the TNSO to strengthen their capacity in applying non-traditional, hybrid census methods.

Senior Research Fellow Dr Chris Nnanatu led the project which delivered the following key outputs:

• Context-specific Bayesian statistical methods for modelled population estimates.
• Technical notes on the implementation of the bespoke statistical population modelling techniques.
• Outline Plan for a sustainable technical support and partnership with the TNSO.
• A 5-day in-person workshop to build the capacity of TNSO staff to learn about modelled population estimates for hybrid census and develop a bespoke method for population estimates to meet their operational needs using existing data.
• Follow-up support with expert guidance for resolving complex technical issues that may arise.
• Training materials and technical notes introducing methods and providing hands-on experience.
• Technical report on the project outcomes.