Not sure what the status is on this, or who has been working on it, but I think we still use 2020 WDI from views2 data in our models. If this is the case, that data is likely not to be raw data but rather the result of various imputations and transforms, which we then apply anew in the current set-up for fatalities002.