# 8. Conclusion

;; 8.01
This paper measures what every maintained distribution table leaves out for the one asset where market prices permit the measurement. Municipal bondholders bear an implicit tax through depressed pre-tax yields; the tables record their burden as zero. Combining SCF portfolios, TAXSIM statutory rates, and a market-identified implicit tax rate, the paper produces corrected effective federal income tax rate schedules by percentile and reports the burden and the inframarginal rent separately.

;; 8.02
Three statements survive every construction. Top-percentile effective rates are understated — by {{N:top1_delta_pp_headline:.2f}} percentage points at the identified market rate, {{N:fed_component_top1_delta_pp:.2f}} at its federal component, never less than {{N:grid_min_top1_delta_pp:.2f}} anywhere on the paper's grid of market-implied rates — with an aggregate unrecorded burden between ${{N:burden_current_spread_bn}} and ${{N:burden_current_anchored_bn}} billion a year at current stocks and yields. The size of the correction is a maturity question, because the term structure of the market-implied rate is genuinely open. And at the short end the exemption delivered essentially no aggregate windfall in 2021: under the two pairings whose counterfactual rates are computed on the full taxable block, the rent is within a billion dollars of zero either side, and across every consistent pairing it stays small against the burden. What that establishes is where the rent is not. Estimated rent is near zero at the short-end anchor and rises as the assumed rate falls toward the levels observed further out the curve — which is a statement about the paper's scenarios, not a measurement of what long-maturity holders were actually paid in 2021, a quantity the household data cannot locate.

;; 8.03
The correction carries its own bias. Correcting municipal bonds alone tilts measured progressivity upward. The housing preferences are eight times larger in scored expenditure and sit in the middle of the distribution; capitalized even partially, they push the other way. The equity margin has reference points at zero and at an amount larger than the entire municipal correction, with no identification that would place the truth between them. The headline is not usable without those two neighbors stated alongside it.

;; 8.04
The forward design is concrete. Author-constructed curves from the MSRB transaction feed and the MMD scale history will replace the display curves and extend the identified record [WRDS robustness to be added]. The restricted SCF's geography will turn expected state rates into actual ones. Holdings-weighted maturity composition will replace the term-structure scenarios with a rate matched to what households actually hold. The institution-side table — the transfer to tax-exempt lenders under each incidence stance — is the natural companion paper. And the full program is the one Galper and Toder began four decades ago: every capitalized preference, both sides of the ledger, with market identification where markets exist.

;; 8.05
The distribution tables are not wrong to count remittances; they are incomplete in a way that market prices can now partly repair. This paper makes the repair where the market permits it, states the direction and the scale of what it cannot repair, and reports both the burden the tables hide and the windfall they hide — so that the next argument about who pays can start from a schedule that includes what the municipal bondholder actually gives up.
