Ghana second-level geography boundaries incorrect


The second-level geography for Ghana is said to be harmonized at the district-level for the period 2000-2010, i.e. the two population censuses.

However, in the respective datasets, only the variable geo2_gh2000, label “Ghana, District 2000 [Level 2, GIS]” contains the correct number of districts in the year 2000, which is 110. The variable geo2_gh, label “Ghana, District 2000 - 2010 [Level 2; consistent boundaries, GIS” only contains 102 districts. Unfortunately for my research, also the downloadable boundary map of Ghana shows only the incorrect 102 districts.

Is a correction of the map already in the works? I have recently worked on tracking the changes in Ghanaian districts from 1988 to 2012 and could potentially provide some help.



IPUMS International has significantly enhanced geography this year, adding maps for both harmonized geographic units as well as year-specific geographic units. The shapefiles for harmonized vs. year-specific variables are found on different pages within the geographic variable documentation.

Find year-specific 2nd-level shapefiles (e.g., Ghana 2000 districts, 110 units) here:…

Creating harmonized geographic units (with consistent spatial boundaries across time periods) requires the merging of some units in which boundaries have changed. These variables and shapefiles can be useful to researchers conducting comparative research across time.

Find shapefiles to match harmonized 2nd level geographic units here:…

Find more documentation about the full set of geographic variables and corresponding shapefiles here:…

Hope this is helpful.



Thanks, I wasn’t aware of the country-specific shapefiles before. They contain the correct number of districts in the respective year.