Context
policyengine_uk_data/parameters/take_up/marriage_allowance.yaml is set to a placeholder 1.0 (100% take-up). HMRC outturn shows ~2.1m claimants of ~4.2m eligible couples, i.e. roughly 50% take-up. The model overstates Marriage Allowance cost by ~£500m/year as a result.
This file is the load-bearing one: datasets/frs.py reads it via load_take_up_rate("marriage_allowance", year) to draw would_claim_marriage_allowance stochastically per person at FRS build time.
Ask
Update the parameter to 0.5 with the HMRC reference. PR incoming.
Companion change
The matching documentation parameter in policyengine-uk is being updated in PolicyEngine/policyengine-uk#1636 — these two need to stay in sync (or eventually be deduplicated by having uk-data read from the policyengine-uk parameter).
Source issue
Originally raised in PolicyEngine/policyengine-uk#623.
Context
policyengine_uk_data/parameters/take_up/marriage_allowance.yamlis set to a placeholder1.0(100% take-up). HMRC outturn shows ~2.1m claimants of ~4.2m eligible couples, i.e. roughly 50% take-up. The model overstates Marriage Allowance cost by ~£500m/year as a result.This file is the load-bearing one:
datasets/frs.pyreads it viaload_take_up_rate("marriage_allowance", year)to drawwould_claim_marriage_allowancestochastically per person at FRS build time.Ask
Update the parameter to
0.5with the HMRC reference. PR incoming.Companion change
The matching documentation parameter in
policyengine-ukis being updated in PolicyEngine/policyengine-uk#1636 — these two need to stay in sync (or eventually be deduplicated by having uk-data read from the policyengine-uk parameter).Source issue
Originally raised in PolicyEngine/policyengine-uk#623.