Heterogeneous treatment effects: Merge IUCN categories to matching frames

[Jo-Schie]

No description provided.

[Jo-Schie]

as @yotaae noticed, boarders and PA overlaps might create problems via multiple assignments (see here). We should be aware of that and find a solution. Proposed solutions are:

• For categorical variables: Create new combined variable categories e.g. "IUCN II & IUCN IV". This could result in lots of new categories with few observation, so an alternative could be to classify all overlaps as "Multiple Usage Categories" (preferred one for me).
• For numerical variables like finance: If the two PAs are part of the same project, no need to adapt. If the two are part of different projects, then we could sum up the finance per area.
• For numerical variables like PA area: Probably the best would be to take the average of both.

[melvinhlwong]

I have created a variable IUCN_strict which is 1 for cat 1-4 (including Ia and Ib) and 0 for IUCN 5 and 6. They are NA otherwise. Currently, the variable is partially integrated in file 02 and I will make sure we have it in the final dataset.