The child indexing method appears to be not applying to the 4th child onwards (affecting 1% of households)
Actually, a misdiagnosis. It turns out that the FRS data isn't correctly being adjusted for the fact that the child limit applies only to children born after 2017. Given that we can't just increase the ages of children across survey years (we'd end up with no 6 year olds or younger in 2023), we should adjust the 2017 year parameter downwards within the formula itself for FRS microsimulations as a partial fix.
The child indexing method appears to be not applying to the 4th child onwards (affecting 1% of households)Actually, a misdiagnosis. It turns out that the FRS data isn't correctly being adjusted for the fact that the child limit applies only to children born after 2017. Given that we can't just increase the ages of children across survey years (we'd end up with no 6 year olds or younger in 2023), we should adjust the 2017 year parameter downwards within the formula itself for FRS microsimulations as a partial fix.