We might default to Tier 1 in places where the requirements of higher tiered reporting would be impractical. We don’t always prioritise this as it can be a large workload for a small reward.
There is a huge variety in the rationale behind each decision to stay Tier 1, or to go higher, but one of the biggest reasons for these defaults is simply that it is harder to acquire all the data we would need to, given that we are a globally applicable tool.
A good example is for the fertiliser model, where the Tier depends on disaggregation per country specific factors and if we had to build a model for each of these factors, it would be a huge feat.
Of course - this is just one example, we do recognise that Tier 2 is not always just country-specific (for example, fertiliser type specific, or when we calculate the EF ourselves to make it specific to animal consumption) so you know it isn't purely the global aspect that makes it difficult.