The IPUMS second-level geographic boundary codes are distinct from the ones used by the Colombian Statistics Agency, DANE. This is true for both time-consistent and year-specific boundaries. Does IPUMS have a cross-walk between their geographic codes and the official geographic codes used by the country statistics office? Thanks.
Yes, the original codes from DANE are available, but you’ll need to download the appropriate shapefiles and take the codes from the attribute table. For Colombia 2005, the DANE codes are available as MUNI2005 in the attribute table for the shapefile. If a few municipalities are merged together (see here) the code for the municipality with the highest population is listed. I’ve attached a CSV version of the attribute table for the year-specific geography for Colombia in 2005 so you can see what this looks like.
geo2_co2005.csv (26.9 KB)
Thanks very much! This answer was very helpful.
A remaining doubt that I have has to do with the variable MIGCO4 in Colombia2005, which records the previous municipality of residence. This code appears to be completely different from all other IPUMS codes listed in the set of geographic codes for Colombia. Is there any way to map this variable to the other geography codes? Many thanks.
Currently IPUMS-I does not have a crosswalk between the 2nd-level geography codes and migration codes. You should be able to match up the codes yourself with a bit of work by using the municipality names.