But I’m confused on how to proceed. The generated series hiu_hhspov_guide, adj_inc, hiu_totval, and hiu_povuniv are all numerical. How do I sort them out based on 0 - 138% of FPG?
Second, in my dataset I have duplicate HIU-IDs, which I assume are members of the same family. Do I need to collapse the other variables by HIU-ID?
If you are using the SHADAC generated SAS or STATA code (found on the last few pages of the document you linked to), the varaible hiu_hhspov [= 100*(hiu_totval/hiu_hhspov_guide)] gives the percentage of FPG for the Health Insurance Unit.
The duplicate HIUIDs do indicate members of the same Health Insurance Unit, which is the primary unit for all HIU variables. If you follow the instructions on the document you linked to, the values for each HIU will be attached to each person from that HIU. So you can collapse other variables to HIU if you want HIUs to be your unit of analysis, but it is important to understand that the values of HIU variables are the same for each member of the Health Insurance Unit.