ADDRESSING DATA GAPS AT LOWER GEOGRAPHIC LEVELS

Muthetho Nkwinika, Statistics South Africa

Fiscal constraints force national statistical offices (NSOs) to use higher geographic levels to stratify household-based sample surveys explicitly; however, service delivery happens at lower geographic levels. This makes it difficult for local and district municipalities to use service delivery indicators from national household-based sample surveys for policy formulation, planning, monitoring, and evaluation. To respond to the challenges of data needs/gaps at local geographic levels, NSOs may need to realign sample survey explicit stratification levels to produce statistically significant indicators at these levels. However, some local geographic levels may be too small to produce statistically significant sample sizes, which may require further rearrangement of these geographic levels like pooling similar geographic levels together to form bigger ones. Can the realignment of explicit sample survey strata at national geographic levels assist in generating reliable estimates that address data gaps at lower geographic levels? This research question can be answered by looking at the measures of dispersion at the lower geographic level from secondary data sources. This quantitative study will use secondary data sources to estimate key service delivery indicators at lower geographic levels. Identified service delivery indicators will further be used to locate geographic levels with no statistically significant indicators.

Keywords: Data and Methods, Small area estimation, Multi-level modeling , Geographic Information Systems (GIS)

See extended abstract.