[Flashtoo]

Very cool, thank you for your efforts!

You mention that it is hard to check whether information is up to date on the CIA website. I see that Wikidata includes both a "start" and "retrieved" timestamp - as many offices are appointed for a set term, do you think it could be helpful to also include an expected end of term date or would that just make it easier to draw the wrong conclusions from outdated data?

[pudo]

Interesting point! I wonder if https://www.wikidata.org/wiki/Property:P8554 would be a good model for this? Perhaps this is also a bit different between countries - in some, I think the government has to resign while in others the term change is sort of constitutional-automatic...

[KarlKemp]

I would use nature of statement (https://www.wikidata.org/wiki/Property:P5102) -> expected (https://www.wikidata.org/wiki/Q50376823) as a qualifier. That would make it easier to update in the common case where it does happen by just deleting the qualifier.

[Flashtoo]

I think not, because that would be a promise that the person would hold that office at least until a certain date. In reality, they may resign or pass away at any moment so you cannot make this claim. If there is such a property, "latest possible end date" might be better but that also does not work if an official can be appointed for another term. So really you can only claim some expected date as of this moment and you can question the value of that beyond being a signal that some piece of data is outdated.