I believe these are incorrectly labeled. Coding for this variable should be:. lab list V105
V105:
0 capital, large city
1 city
2 town
3 countryside
4 abroad
However, in Zimbabwe 2015 the only categories in the questionnaire are Urban and Rural, which should have been recoded to codes 2 and 3 in this case (codes 0-2 being urban, and code 3 being rural). The good news for you is that code 1 means urban and code 2 means rural in the current dataset. This will probably be corrected in the next iteration of the dataset.