Summary
The enhanced CPS carries capital gain distributions only for filers who report them without Schedule D (non_sch_d_capital_gains = PUF E01100, ~$13.7B in 2026 across 4.0M tax units). Filers with a Schedule D report their capital gain distributions on Schedule D line 13, where they are folded into long_term_capital_gains (PUF P23250) and are no longer separable. That Schedule-D-routed slice is most of the dollars: ICI reports mutual fund shareholders reinvested $234B of capital gain distributions in 2023 alone (all account types), and ~23M households hold ~$7T of long-term mutual fund assets in nonretirement accounts.
Without separating this component, PolicyEngine cannot model reforms that treat fund distributions differently from other realized gains — currently requested for the GROWTH Act (H.R. 2089 / S. 1839), which would defer tax on automatically reinvested RIC capital gain dividends (see PolicyEngine/policyengine-us#8828 for the companion model fix).
Proposal
Add a capital_gain_distributions_in_ltcg (name TBD) imputed component of long_term_capital_gains:
- Donor/target source: SOI Sales of Capital Assets study publishes capital gain distribution totals and their share of net long-term gains by AGI class; SOI Individual Complete Report tables provide line-item aggregates for calibration targets.
- Imputation: predict the distribution share of each record's LTCG conditional on AGI, total LTCG, age, and filing status (microimpute), then calibrate to the SOI aggregates by AGI band (microcalibrate) — same pattern as existing PUF-variable imputations.
- RIC/REIT split: Schedule D line 13 includes REIT capital gain dividends, which are not RIC distributions (subchapter M part II vs part I). Apportion using ICI (fund industry distributions) vs NAREIT data; reforms like the GROWTH Act apply only to the RIC share.
- Consistency:
non_sch_d_capital_gains + the new component should reconcile against total capital gain distribution aggregates.
Note the base is inherently taxable-account-only (tax-deferred accounts do not pass distributions through to 1099-DIV), so no retirement-account carve-out is needed.
Motivation
Requested by The Tax Project for GROWTH Act analysis (revenue and household impacts of deferring reinvested capital gain distributions). Layer B (reinvestment/DRIP fraction, from ICI data as a sensitivity parameter) and Layer C (multi-year recapture timing) build on this but are out of scope here.
Summary
The enhanced CPS carries capital gain distributions only for filers who report them without Schedule D (
non_sch_d_capital_gains= PUFE01100, ~$13.7B in 2026 across 4.0M tax units). Filers with a Schedule D report their capital gain distributions on Schedule D line 13, where they are folded intolong_term_capital_gains(PUFP23250) and are no longer separable. That Schedule-D-routed slice is most of the dollars: ICI reports mutual fund shareholders reinvested $234B of capital gain distributions in 2023 alone (all account types), and ~23M households hold ~$7T of long-term mutual fund assets in nonretirement accounts.Without separating this component, PolicyEngine cannot model reforms that treat fund distributions differently from other realized gains — currently requested for the GROWTH Act (H.R. 2089 / S. 1839), which would defer tax on automatically reinvested RIC capital gain dividends (see PolicyEngine/policyengine-us#8828 for the companion model fix).
Proposal
Add a
capital_gain_distributions_in_ltcg(name TBD) imputed component oflong_term_capital_gains:non_sch_d_capital_gains+ the new component should reconcile against total capital gain distribution aggregates.Note the base is inherently taxable-account-only (tax-deferred accounts do not pass distributions through to 1099-DIV), so no retirement-account carve-out is needed.
Motivation
Requested by The Tax Project for GROWTH Act analysis (revenue and household impacts of deferring reinvested capital gain distributions). Layer B (reinvestment/DRIP fraction, from ICI data as a sensitivity parameter) and Layer C (multi-year recapture timing) build on this but are out of scope here.