Current calibration uses separate national gross targets for alimony_income and alimony_expense, but the strongest public source is IRS SOI alimony reported on tax returns, and after TCJA that series is a shrinking grandfathered residual rather than a clean current-population administrative total.
Why the current gross targets are weak:
- The current targets are explicitly survey-reported.
- IRS SOI is more authoritative than survey totals, but since divorce or separation instruments executed after 2018 generally made alimony non-deductible/non-includable, the tax series no longer measures a broad current-population flow.
- IRS therefore gives us a public return-based benchmark, but not a clean national administrative total for current alimony flows.
- Separate gross targets are therefore weak calibration anchors and can be circular.
Proposed change:
- Remove the separate gross calibration targets for
alimony_income and alimony_expense.
- Replace them with a single balance constraint or diagnostic:
sum(alimony_income) - sum(alimony_expense) = 0.
- Keep IRS SOI alimony aggregates only as a legacy external sanity check.
Primary sources:
Current calibration uses separate national gross targets for
alimony_incomeandalimony_expense, but the strongest public source is IRS SOI alimony reported on tax returns, and after TCJA that series is a shrinking grandfathered residual rather than a clean current-population administrative total.Why the current gross targets are weak:
Proposed change:
alimony_incomeandalimony_expense.sum(alimony_income) - sum(alimony_expense) = 0.Primary sources: