The degree of disability in Indonesia 2010 census has three levels but the tabulation only gives two levels. How did you combined three levels into two groups?
It sounds like you are looking at the variable DISABLED in the Indonesian samples. There are four available codes shown; this is because DISABLED is available for samples from many countries, some of which make use of all or some of all of the codes. For example, with the 2010 Indonesia sample, the available codes are 0, 1, and 2, while for Burkina Faso’s 1985 sample, the available codes are 0, 1, 2, and 9.
Since DISABLED indicates whether the person reported a disability of any kind, it is based on several source variables. You can see how they relate either by adding either the harmonized disability variables or the relevant source variables to your extract.
I hope this helps, however, if I have misinterpreted your question, please let me know.