The output is only as trustworthy as the scope and assumptions behind it.

An ordinary test proves the calculator still agrees with itself. These cases fail when it stops agreeing with the source. Nothing in them reads a value out of the calculator to build an expectation. Each source listed below records the date it was last retrieved and checked against the code, which is a staleness signal rather than a review attestation. Each pack is also run against deliberately wrong implementations of the rules it covers and is required to fail every one. That check uses a maintained set of known errors per pack, so it does not prove that every individual case listed here is independently load-bearing.

• The age/service decision between the 1.0% general multiplier and the 1.1% enhanced multiplier: both conditions are required, tested at separation, and a fraction of a year is not a completed year.

  Statutory — does not change with the calendar · 7 cases · re-check due 2027-07-01

  Not covered by this claim:

  • § 8415(i) other statutory conditions — retirement on a § 8412 annuity, and the excluded special categories (Congressional employees, law enforcement officers, air traffic controllers).
  • The full calculateFERSPension path, which layers service credit, proration and survivor reductions on top of the rate.
  • § 8415(g), the part-time computation rule.

  5 U.S.C. § 8415 — Computation of basic annuity

• The MRA+10 age reduction of five-twelfths of 1 percent for each FULL month by which the annuity commencement date precedes the 62nd anniversary, including that a partial month does not reduce and that postponing commencement shrinks the reduction.

  Statutory — does not change with the calendar · 6 cases · re-check due 2027-07-01

  Not covered by this claim:

  • Eligibility for an MRA+10 annuity itself (§ 8412(g)), and the deferred-annuity rules (§ 8413(b)).
  • Whether a given separation qualifies for postponement.

  5 U.S.C. § 8415(h)

• The FERS annuity supplement fraction: total years of service over 40, with a fraction of a year rounded to the nearest whole number (½ rounding up) and the numerator capped at 40; plus the helper’s age-62 cut-off.

  Statutory — does not change with the calendar · 7 cases · re-check due 2027-07-01

  Not covered by this claim:

  • The § 8421(a)(3)(B) MONTH-level cessation rule — the helper under test takes a whole-number age, so only its age-62 cut-off is certified.
  • The SRS earnings test and its exempt amounts.
  • Special-category (6C) provisions.

  5 U.S.C. § 8421 — Annuity supplement

• The 2025 and 2026 MFJ and single (unmarried) tax rate tables. Every bracket ceiling in all four ladders is asserted against the IRS’s own printed cumulative tax at that breakpoint, plus one case above the top breakpoint, that a surviving spouse uses the § 1(j)(2)(A) joint ladder, and that the two years’ tables are distinct.

  Tax years 2025, 2026 · 28 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The Head of Household and Married Filing Separately tables — the engine models only MFJ, qualifying surviving spouse and single.
  • The 2024 tables, and the projected tables the engine derives for years after 2026 from the scenario bracket-inflation assumption.
  • Capital-gains rates, the alternative minimum tax, credits, and every other part of the § 1 computation outside the ordinary-income ladder.

  IRS Revenue Procedure 2024-40, § 2.01 (2025 tax rate tables) · IRS Revenue Procedure 2025-32, § 4.01 (2026 tax rate tables)

• The 2025 and 2026 standard deductions for MFJ and single filers, and the § 63(f) additional deduction for the aged — including that the higher unmarried amount is denied to a qualifying surviving spouse, and that the 2025 general deduction and 2025 age addition come from different revenue procedures.

  Tax years 2025, 2026 · 11 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The additional deduction for BLINDNESS, which § 63(f) grants on the same terms but the engine does not model.
  • The Head of Household and Married Filing Separately amounts, and the § 63(c)(5) dependent limitation.
  • Itemized deductions — the engine always takes the standard deduction.

  IRS Revenue Procedure 2024-40, § 2.15(3) (2025 additional standard deduction for the aged) · IRS Revenue Procedure 2025-32, §§ 3.01 and 4.14 (OBBBA-amended 2025 standard deduction; 2026 standard and additional standard deductions)

• The 2025 enhanced senior deduction: $6,000 per eligible individual aged 65 or over, phased out at 6% of MAGI above $75,000 ($150,000 for MFJ), clamped at zero, counting a second person only on a joint return — computed exactly as Schedule 1-A Part V lays it out line by line.

  Tax years 2025 · 11 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Tax years 2026 through 2028. The engine models the deduction through 2028, but neither annual revenue procedure adjusts or restates it, so 2025 is the only year for which a primary source was read. The later years remain uncertified modeling.
  • The valid-SSN and must-file-jointly conditions in the form’s caption, which the engine does not model.
  • HOW AGE 65 IS DETERMINED. This claim certifies the AMOUNT and the PHASEOUT against Schedule 1-A; it takes the eligible-person count as given. The engine decides that count elsewhere: at the close of the year for a living taxpayer, and — for a decedent still on their final return — at the DATE OF DEATH, per Pub. 501, which states the rule and its one-day boundary under the enhanced deduction as well as under the higher standard deduction for age (“Your spouse was born on February 14, 1960, and died on February 13, 2025. Your spouse is considered age 65 at the time of death. However, if your spouse died on February 12, 2025, your spouse isn’t considered age 65.”). That determination lives in the projection adapter rather than in this calculator, so its evidence is the boundary tests in projection.test.ts (#358), not a case in this pack.
  • The other Schedule 1-A deductions (tips, overtime, vehicle loan interest), which the engine does not model at all.
  • The provision’s FIRST and LAST applicable years. Schedule 1-A evidences the 2025 amount and phaseout; it is not an effective-date provision, so neither the 2025 start nor the 2028 end is certified here. The engine’s year window is covered by ordinary tests in taxes.test.ts instead.

  IRS Schedule 1-A (Form 1040), 2025, Part V — Enhanced Deduction for Seniors

• The OASDI contribution and benefit base for 2025 ($176,100) and 2026 ($184,500), that a published base is returned exactly rather than scaled by the engine’s wage-growth assumption, the statutory 6.2% OASDI employee rate applied up to that base PER EARNER, and the uncapped 1.45% HI (Medicare) employee rate.

  Tax years 2025, 2026 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Calendar years after 2026, where the engine projects the base forward at a fixed built-in 3.5% wage-growth default — production FICA callers pass no rate, so this is not a user-configurable scenario input. Disclosed, not certified.
  • The 0.9% Additional Medicare Tax and its filing-status thresholds, which come from IRC § 3101(b)(2) and were not read for this pack.
  • The employer share, the self-employment rates, and SSA’s $300-increment rounding convention for setting the base.

  SSA Office of the Chief Actuary — Contribution and Benefit Base

• The § 402(g)(1) elective-deferral limit and both catch-up tiers for 2024, 2025 and 2026 — the § 414(v)(2)(B)(i) age-50 catch-up and the SECURE 2.0 § 414(v)(2)(E)(i) higher catch-up — including that the higher tier is chosen by the age ATTAINED IN THE CALENDAR YEAR, is inclusive at 63, and drops back to the ordinary catch-up at 64.

  Tax years 2024, 2025, 2026 · 17 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 415(c) annual additions limit, which caps employee plus agency contributions and is not modelled here.
  • The § 414(v)(7) Roth catch-up wage threshold, which forces catch-up contributions to be Roth above a prior-year wage level. The engine does not model that requirement.
  • The separate SIMPLE-plan limits under §§ 408(p) and 401(k)(11), which the same notices set and which do not apply to the TSP.
  • The EFFECTIVE DATE of the SECURE 2.0 higher catch-up. The 2024 cases assert the amounts Notice 2023-75 states; that the higher tier was unavailable for 2024 is inferred from the notice not mentioning § 414(v)(2)(E), which is not the same as an enacting provision read for this pack.
  • Calendar years outside 2024–2026. Beyond 2026 the engine HOLDS the latest published limits, and before 2024 it reuses the earliest — both are disclosed modelling choices with no IRS figure behind them, so they are covered by ordinary tests in projection.test.ts rather than claimed here.

  IRS Notice 2023-75 — 2024 Limitations Adjusted as Provided in Section 415(d) · IRS Notice 2024-80 — 2025 Amounts Relating to Retirement Plans and IRAs · IRS Notice 2025-67 — 2026 Amounts Relating to Retirement Plans and IRAs

• The § 219(b)(5)(A) IRA contribution limit and the § 219(b)(5)(B)(ii) age-50 catch-up for 2025 and 2026, from the same cost-of-living notices that carry the elective-deferral amounts — their titles read "Retirement Plans and IRAs" because they set both. Certified in three parts. FIRST, the amounts themselves: $7,500 for 2026 "increased from $7,000", with the age-50 catch-up "increased from $1,000 to $1,100". SECOND, that there is NO 60-63 tier for an IRA, which is an absence asserted rather than assumed — the higher SECURE 2.0 catch-up is § 414(v)(2)(E), which reaches only an "applicable employer plan", so an IRA holder aged 61 gets the ordinary age-50 catch-up and a rule reusing the elective-deferral table would grant $35,750 instead of $8,600. THIRD, that the two statutes are NOT interchangeable, pinned as a relationship between the two limits in the same year: this is the defect #440 corrected, where every non-federal account — savings, 401(k) or IRA alike — was held to the TSP’s own § 402(g) ceiling, permitting nearly four times the legal IRA contribution.

  Tax years 2024, 2025, 2026 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • DEDUCTIBILITY, which § 219(g) phases out on income for someone covered by a workplace plan. That is a separate question from the contribution CEILING certified here, and it is moot in the engine today because no contribution reduces taxable income for anyone (#442). The § 219(g) phase-out ranges the same notice states are deliberately not modelled.
  • The Roth IRA contribution phase-out under § 408A(c)(3), and the saver’s credit under § 25B, both also set by these notices.
  • The SPOUSAL IRA rule of § 219(c), which lets a non-working spouse contribute against the other’s compensation. The engine derives a contribution from the person’s OWN salary, so a person with no wage contributes nothing regardless of household income.
  • Whether the account a user labels an IRA is Traditional or Roth. The engine adds every modelled contribution to the traditional balance, so a Roth IRA’s after-tax character is not represented. An ORDINARY TAXABLE account is not offered as a type at all: suppressing its contributions would not stop the balance taking required minimum distributions and being taxed as ordinary income, so it is declined rather than mis-modelled.
  • Calendar years outside 2024–2026, on the same holding and reuse policy as the elective-deferral limits above.

  IRS Notice 2023-75 — 2024 Limitations Adjusted as Provided in Section 415(d) · IRS Notice 2024-80 — 2025 Amounts Relating to Retirement Plans and IRAs · IRS Notice 2025-67 — 2026 Amounts Relating to Retirement Plans and IRAs

• A pre-tax elective deferral is EXCLUDED from federal taxable wages and INCLUDED in FICA wages, and the two halves are certified together because getting either one alone is a defect. IRS Notice 2025-67 names the first in its own words — § 402(g)(1) is a limitation “on the EXCLUSION for elective deferrals” — and the IRS’s retirement-plan contributions FAQ states both sides in one table: “Employee elective salary deferrals — pre-tax” are subject to Social Security and Medicare withholding, while Box 1 wages “don’t include pre-tax contributions made under a salary reduction agreement”. Before #442 FERSCalc did neither: the deferral was subtracted from CASH FLOW only, so the same dollars were taxed as income when contributed and again as ordinary income when withdrawn from the traditional balance. Also certified: that the exclusion is the EMPLOYEE’s deferral alone and not the agency automatic or matching contribution, which was never the employee’s income; that it follows the money actually contributed, so the statutory ceiling bounds the deduction as well as the contribution; and that it is not subtracted from spendable cash a second time, since `calculateNetIncome` already removes it — so against the SAME contribution under the old treatment net income rises by exactly the tax saved, while against making no contribution at all it still falls by the contribution less that saving.

  Tax years 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • ROTH (after-tax) elective deferrals. The engine models no Roth contribution at all — every modelled contribution lands in the traditional balance — so there is no after-tax deferral for this claim to be wrong about. A Roth contribution correctly reduces neither income tax nor FICA.
  • The § 219 IRA DEDUCTION ENTIRELY. An IRA contribution is not a salary-reduction deferral — it is made from pay that was taxed and recovered, if at all, through a deduction that phases out on income under § 219(g) for anyone also covered by a workplace plan. The engine cannot test that eligibility, so it does not grant the deduction at all: an IRA contribution reaches the balance and leaves taxable wages untouched, which OVERSTATES tax for a genuinely eligible saver. That is the conservative direction and is certified as a case rather than left implicit.
  • STATE conformity beyond Pennsylvania. The default is that a state beginning from federal AGI inherits the exclusion, which is true of most states and is a DISCLOSED ASSUMPTION rather than a verified survey — read the flag’s absence the way #429 taught us to read its declared list: the non-conforming states we have FOUND, not the ones that exist.
  • FUTA and any state unemployment wage base, which follow the FICA side rather than the income-tax side and are not modelled.
  • The § 414(v)(7) Roth catch-up requirement, which from its effective date forces catch-up contributions above a wage threshold to be Roth — and therefore NOT excluded. The engine does not model that requirement, so a high earner’s catch-up is excluded here when a real return may not exclude it.

  IRS Notice 2025-67 — 2026 Amounts Relating to Retirement Plans and IRAs, for the § 402(g)(1) limitation "on the exclusion for elective deferrals" · IRS — Retirement plan FAQs regarding contributions: are retirement plan contributions subject to withholding for FICA, Medicare or federal income tax? · PA DOR — Personal Income Tax Guide, Gross Compensation: the Federal Employee’s Thrift Savings Plan contributions rule

• The standard monthly Part B premium CMS published for 2024 ($174.70), 2025 ($185.00) and 2026 ($202.90), charged when MAGI is at or below the first IRMAA threshold, for both single and joint filers.

  Tax years 2024, 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The Part B IMMUNOSUPPRESSIVE-DRUG-ONLY schedule. Each fact sheet publishes a second, entirely separate premium and IRMAA table for people whose Medicare coverage ended 36 months after a kidney transplant; the engine does not model it.
  • The annual Part B DEDUCTIBLE and coinsurance, which the same fact sheets publish ($283 for 2026) and the engine does not model.
  • Part A premiums, and Part D and its separate IRMAA surcharge.
  • Calendar years after 2026, where the engine projects the premium forward at a disclosed 5.5% assumption rather than a CMS figure.

  CMS 2024 Medicare Parts A & B Premiums and Deductibles · CMS 2025 Medicare Parts A & B Premiums and Deductibles · CMS 2026 Medicare Parts A & B Premiums and Deductibles

• Every tier of the CMS "Full Part B Coverage" IRMAA schedule for 2024, 2025 and 2026, single and joint, asserted against the TOTAL MONTHLY PREMIUM column CMS prints rather than the surcharge column the engine stores.

  Tax years 2024, 2025, 2026 · 30 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The Part B IMMUNOSUPPRESSIVE-DRUG-ONLY schedule. Each fact sheet publishes a second, entirely separate premium and IRMAA table for people whose Medicare coverage ended 36 months after a kidney transplant; the engine does not model it.
  • The MARRIED-FILING-SEPARATELY schedule. CMS publishes its own single-column table for beneficiaries who lived with a spouse during the year but filed separately, and above the first tier it is not the single-filer table. The engine has only single and joint thresholds.
  • Part A premiums, and Part D and its separate IRMAA surcharge.
  • Calendar years after 2026, where surcharges are projected at a disclosed assumption.

  CMS 2024 Medicare Parts A & B Premiums and Deductibles · CMS 2025 Medicare Parts A & B Premiums and Deductibles · CMS 2026 Medicare Parts A & B Premiums and Deductibles

• The CMS band wording at EVERY boundary of the Full Part B schedule for 2024-2026, single and joint: a MAGI exactly equal to any lower threshold stays in the tier below ("less than or equal to") while one cent more moves up, and a MAGI exactly equal to the TOP threshold is already in the top tier ("greater than or equal to $500,000") — plus that neither ceiling helper reports a target that is itself inside the top tier, and that a MAGI already in the top tier gets no ceiling at all.

  Tax years 2024, 2025, 2026 · 36 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The MARRIED-FILING-SEPARATELY schedule. CMS publishes its own single-column table for beneficiaries who lived with a spouse during the year but filed separately, and above the first tier it is not the single-filer table. The engine has only single and joint thresholds.
  • The Part B IMMUNOSUPPRESSIVE-DRUG-ONLY schedule, whose separate table has its own boundaries.
  • Whether a real beneficiary lands on a boundary exactly; these certify the comparison, not its frequency.
  • The projected-year threshold VALUES beyond 2026, which depend on the engine’s indexing assumption — though the boundary comparison certified here applies to them too.

  CMS 2024 Medicare Parts A & B Premiums and Deductibles · CMS 2025 Medicare Parts A & B Premiums and Deductibles · CMS 2026 Medicare Parts A & B Premiums and Deductibles

• That a year’s Part B premium is priced on the MAGI of the second calendar year preceding it, per 42 U.S.C. § 1395r(i)(4)(B)(i); that the filing status is read from that same return rather than a different year; and that the first two projection years, where the statutory year predates the projection, are flagged as a proxy.

  Statutory — does not change with the calendar · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 1395r(i)(4)(B)(ii) fallback to the previous taxable year when Treasury lacks adequate data by October 15, and the (B)(iii) non-filer rules. Neither is modeled.
  • The § 1395r(i)(4)(C) life-changing-event procedure for using a more recent taxable year (SSA-44), already disclosed as not modeled.
  • The VALUE substituted for the first two projection years. Only the fact that it is MARKED a proxy is certified; the substituted figure is a modeling choice with no statutory basis.
  • The § 1395r(i)(4)(A) definition of MAGI itself. The engine builds MAGI in its income ledger and this claim does not certify that construction.

  42 U.S.C. § 1395r(i) — Reduction in premium subsidy based on income

• The two statutory rules governing how IRMAA thresholds move: § 1395r(i)(5)(B) rounds every indexed amount to the nearest multiple of $1,000, and § 1395r(i)(5)(C) exempts the $500,000 amounts from indexing until a calendar year beginning after 2027 — so 2028 is the first year the top threshold moves, while the lower thresholds index normally before then.

  Statutory — does not change with the calendar · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The indexing RATE. § 1395r(i)(5)(A) indexes by CPI-U for the 12-month period ending with August of the preceding year; the engine applies a flat disclosed assumption, so projected threshold VALUES are not certified — only the rounding and freeze rules applied to them.
  • The § 1395r(i)(6) temporary 2011–2017 threshold adjustment, which is historical and not modeled.
  • The rounding DIRECTION at an exact half-thousand. The statute says "nearest" without specifying; the engine rounds half up.

  42 U.S.C. § 1395r(i) — Reduction in premium subsidy based on income

• The Virginia rate schedule of § 58.1-320 — 2% to $3,000, 3% to $5,000, 5% to $17,000 and 5.75% above — asserted at every bracket ceiling and above the top.

  Tax years 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Credits of every kind, including the low-income credit and the credit for tax paid to another state.
  • The pre-1990 rate schedules the section also carries, which are historical and unreachable.

  Code of Virginia § 58.1-320 — Rates of tax

• That Social Security never enters the Virginia base at all, per the § 58.1-322.02(3) subtraction — so adding a benefit changes the tax by nothing, a benefit-only household owes nothing, and a benefit cannot push a household into the age-deduction phase-out.

  Tax years 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Tier 2 and other Railroad Retirement benefits, which the Department treats separately and the engine does not model.
  • Every subtraction in § 58.1-322.02 other than Social Security — U.S. and Virginia obligation interest, disability income, and the rest. Note § 58.1-322.02(4) makes the disability-income subtraction mutually exclusive with the age deduction; the engine models neither.

  Code of Virginia § 58.1-322.02 — Virginia taxable income; subtractions

• The year-by-year standard deduction schedule of § 58.1-322.03(1)(b), asserted at every period in the subdivision from 2021 through the scheduled 2030 reversion to $3,000/$6,000, including that the married amount is twice the single amount and that the reversion falls in 2030 rather than 2029.

  Tax years 2021, 2022, 2023, 2024, 2025, 2026, 2027, 2028, 2029, 2030 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Itemized deductions. § 58.1-322.03(1)(a) allows them where the taxpayer itemizes federally; the engine always takes the standard deduction.
  • The § 58.1-322.03(1)(b) dependent limitation, which caps the deduction to earned income for someone claimable on another return.
  • Whether the 2030 reversion survives. It is current law and is modeled as such; Virginia has revised these amounts before, so this is a provision to re-check rather than trust indefinitely.

  Code of Virginia § 58.1-322.03 — Virginia taxable income; deductions

• The § 58.1-322.03(5)(b) age deduction as the Department’s Form 760 Age Deduction Worksheet computes it: $12,000 per claiming taxpayer POOLED, then reduced ONCE by the amount AFAGI exceeds $50,000 single or $75,000 married, clamped at zero, measured against income BEFORE the standard deduction, and unavailable below 65. Reducing each spouse separately — the natural reading of the statute alone — is wrong and understates the deduction.

  Tax years 2025, 2026 · 12 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 58.1-322.03(5)(a) FLAT $12,000 branch for individuals born on or before 1 January 1939, which takes no AFAGI reduction. It is UNMODELLED, not unreachable, and since #347 it is not blocked by missing data either — `applyTax` now receives the birth date it would need, so the branch is simply not implemented (#349). Everyone in it is treated under the (5)(b) income-based branch, which under-deducts once AFAGI passes the threshold. Anyone in that branch is 87 or older in 2026, but that is a judgement about who uses the calculator, not a statement about the statute.
  • Worksheet Line 15, which splits the household deduction between spouses. It matters only for married-filing-separately returns, which the engine does not model.
  • That claiming the age deduction BARS the disability-income subtraction, the Credit for Low-Income Individuals, and the Virginia (and Refundable Virginia) Earned Income Credit. The Form 760 instructions make them mutually exclusive; the engine models none of them, so it cannot detect the conflict.
  • The Department’s conformity additions and subtractions, which enter the worksheet’s AFAGI at Lines 3 and 5 and are not modelled.
  • Married filing separately, which § 58.1-322.03(5)(b) gives its own combined-AFAGI rule. The engine models only single and joint.
  • Every subtraction in § 58.1-322.02 other than Social Security — U.S. and Virginia obligation interest, disability income, and the rest. Note § 58.1-322.02(4) makes the disability-income subtraction mutually exclusive with the age deduction; the engine models neither.

  Virginia Department of Taxation — 2025 Form 760 instructions: "Deceased Taxpayers" (the deceased spouse’s Date of Birth goes on the return) and the Age 65 and Older Deduction Worksheet ("taxpayers born on or between January 2, 1939, and January 1, 1961") · Code of Virginia § 58.1-322.03 — Virginia taxable income; deductions · Virginia Department of Taxation — Subtractions and Deductions

• That Virginia routes a federal qualifying surviving spouse to FILING STATUS 1 — single — so the single standard deduction and the single AFAGI age-deduction threshold apply, exactly as for any unmarried filer. The claim is about the STATUS MAPPING; it says nothing about the total number of exemptions, because a surviving spouse may also claim the qualifying child that made them a QSS and the engine models no dependants.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household, which the same instruction also sends to Filing Status 1 but which the engine does not model as a status at all.
  • Filing Status 3 (married filing separately) and Filing Status 4 (married, spouse has no income), neither of which the engine models.
  • Whether a surviving spouse could benefit from the Spouse Tax Adjustment, which is available only to Filing Status 2 and is not modelled.
  • Dependants. Filing Status 1 does not stop a surviving spouse claiming the qualifying child that made them a QSS federally; § 58.1-322.03(2) would give an exemption for that dependant, and the engine models no dependants at all — so the modelled exemption is one short, overstating tax.
  • Which SLOT the survivor occupies is no longer an exclusion. It WAS one: when person A died first the live survivor was person B, whom Virginia’s age deduction and aged exemption never read — charging $1,773.40 where $1,037.40 was due on a $45,000 pension at 68. #356 fixed that at the tax boundary, so this claim now holds for a survivor in EITHER slot. The evidence is the slot-invariance property in projection.test.ts, not a case in this pack.

  Virginia Department of Taxation — Filing Status: "If your filing status on your federal return was Single, Head of Household, or Qualifying Widow(er), you must use Filing Status 1" · Code of Virginia § 58.1-324 — Married individuals (the only route to joint treatment, and it requires marriage)

• The § 58.1-322.03(2) personal exemptions: $930 per filer, plus a further $800 for each filer who has attained 65 (§ 63(f) "aged"), counted per person rather than per return.

  Tax years 2025, 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Dependents. The personal-exemption count is self plus spouse only, so a household with dependents is understated by $930 each.
  • Blindness. § 58.1-322.03(2)(b) grants the same $800 addition for blind taxpayers as for aged ones; blindness is not an engine input.

  Code of Virginia § 58.1-322.03 — Virginia taxable income; deductions

• The seven-band D-40 Calculation I rate schedule, asserted at every band ceiling, including that the 8.5% band ends at $250,000 rather than $350,000 and that the top rate is 10.75% rather than 9.25%. Each case reproduces by arithmetic the cumulative constant the NEXT line of the published schedule prints — $400, $2,200, $3,500, $19,650, $42,775 and $91,525 — so a wrong rate or threshold could not land on the printed figure. The schedule is identical in the 2025 D-40 and 2026 D-40ES booklets, so one table serves both years.

  Tax years 2025, 2026 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • DC’s rounded tax TABLES, which the booklet directs a filer to use at taxable income of $100,000 or less. These cases compute the Calculation I rate schedule exactly, so a real filed return can differ from them by a few dollars.
  • Head of household, which the engine cannot express — it carries single, joint and qualified surviving spouse only. DC gives a head of household a $22,500 standard deduction for 2025, so a HoH filer is treated as single here and OVERCHARGED. Identifying the status would need a question the calculator does not ask.

  DC Office of Tax and Revenue — 2025 D-40 booklet (Calculation I rate schedule, standard deduction table, additional standard deduction worksheet, Schedule I Calculation B lines 10 and 12) · DC Office of Tax and Revenue — 2026 D-40ES booklet (2026 standard deduction and additional amounts; the same rate schedule)

• The standard deduction and its additional amount for age, for BOTH published years: $15,000 single and $30,000 joint for 2025, rising to $16,100 and $32,200 for 2026, plus $1,600 per person aged 65 or over — $2,000 where the filer is unmarried and not a surviving spouse — rising to $1,650 and $2,050. Also that a qualified surviving spouse takes the JOINT base amount with only ONE addition at the LOWER rate, which is neither the single nor the joint treatment. Unlike Maryland (#354) both years are published, so absolute 2026 totals are certified rather than a year-over-year difference. AGE ITSELF IS MEASURED BY BIRTH DATE, UNDER IRC § 63(f) (#360): Calculation G-1’s own asterisk deems anyone "born before January 2, 1961" to be age 65 at the end of 2025 — the identical day-before-birthday convention Virginia’s age deduction already reads (#347) — not the plain year-end age the rule previously compared. That is a defect independent of death: a LIVING filer born 1 January of the cutoff year was wrongly denied the addition, because their plain calendar age has already advanced by 31 December where the deemed test has not. A DECEDENT keeps the addition too, on the SAME birth-date test and with no age-at-death condition at all — the deceased-taxpayer instruction confirms it ("Do not adjust the deceased’s income, exemptions or deductions to reflect the date of death").

  Tax years 2025, 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household, which the engine cannot express — it carries single, joint and qualified surviving spouse only. DC gives a head of household a $22,500 standard deduction for 2025, so a HoH filer is treated as single here and OVERCHARGED. Identifying the status would need a question the calculator does not ask.
  • The blindness addition. D-40 lines 2d and 2f add the same $1,600 or $2,000 again for a blind filer and a blind spouse. Blindness is not an engine input, so the addition is never granted and tax is overstated for a blind filer.
  • Itemized deductions. DC requires the same election as the federal return and computes its own modified itemized amount; the engine models the standard deduction only.
  • Years after 2026. The engine holds the 2026 amounts forward as a labelled projection baseline; no case certifies a later year.

  DC Office of Tax and Revenue — 2025 D-40 booklet (Calculation I rate schedule, standard deduction table, additional standard deduction worksheet, Schedule I Calculation B lines 10 and 12) · DC Office of Tax and Revenue — 2026 D-40ES booklet (2026 standard deduction and additional amounts; the same rate schedule) · DC Office of Tax and Revenue — 2025 D-40 booklet, Schedule S Calculation G-1 (standard deduction worksheet, asterisked "born before January 2, 1961" deeming note) and the "Filing for a deceased taxpayer" / "Deceased Taxpayer" instructions

• That § 47-1803.02(a)(2)(L) excludes Social Security from District gross income entirely, so adding benefits to a household changes no DC figure and a household living on benefits alone owes nothing. The engine previously added the GROSS benefit to the base, charging DC tax on 100% of it where even a federal return includes at most 85%.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Tier 1 Railroad Retirement specifically. (a)(2)(L) names it alongside Social Security and the treatment is identical, but the engine models no Railroad Retirement input, so the exclusion cannot be exercised against one.

  D.C. Code § 47-1803.02 — Gross income; deductions · DC Office of Tax and Revenue — 2025 D-40 booklet (Calculation I rate schedule, standard deduction table, additional standard deduction worksheet, Schedule I Calculation B lines 10 and 12)

• That § 47-1803.02(a)(2)(N)(ii) excludes survivor benefits received from the District or the federal government once the RECIPIENT is 62 by the end of the taxable year — gated on that person’s own age rather than on the return, reaching the survivor annuity only and not the recipient’s own pension, and taken at GROSS as Schedule I Calculation B line 12 directs ("enter the total survivor benefits"). There is no cap and no expiry: the "taxable years beginning before January 1, 2015" proviso sits inside sub-subparagraph (i) and does not reach (ii). A DECEDENT RECIPIENT KEEPS THIS EXCLUSION TOO (#360), gated on the PLAIN age they reached at death — Line 12’s own words are "62 years of age or older AS OF DECEMBER 31, 2025", with no deeming rule of the kind Calculation G-1 carries, so this is deliberately NOT the same § 63(f) test the standard-deduction claim above certifies. Under-granting a decedent who was genuinely under 62 is the disclosed, safer direction for an exclusion with no cap.

  Tax years 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Social Security survivor benefits, which line 12 expressly tells the filer not to include because line 10 has already removed them. They never reach this field in any case: Social Security flows through ssIncomePersonA/B.
  • Survivor benefits from a District or federal source OTHER than a FERS survivor annuity. The engine models no other survivor benefit, so the exclusion cannot wrongly reach one.

  D.C. Code § 47-1803.02 — Gross income; deductions · DC Office of Tax and Revenue — 2025 D-40 booklet (Calculation I rate schedule, standard deduction table, additional standard deduction worksheet, Schedule I Calculation B lines 10 and 12) · DC Office of Tax and Revenue — 2025 D-40 booklet, "Filing for a deceased taxpayer" / "Deceased Taxpayer" instructions

• That an ordinary FERS annuity and a TSP withdrawal are FULLY taxable by the District, and specifically that a 62-year-old receives no $3,000 pension exclusion. § 47-1803.02(a)(2)(N)(i) granted one, but its proviso limits it to "taxable years beginning before January 1, 2015" — it is EXPIRED, so full taxation is the CORRECT treatment for every year this calculator projects. Certified so that a later pass reading (N)(i) without its proviso cannot restore it.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Credits and the remaining DC Schedule I subtractions — the Schedule H homeowner and renter property credit, which can be material to a DC retiree but turns on residence, rent or property tax and household facts the calculator does not collect; the disability exclusion of up to $10,000; the DC EITC; and the sourcing distinction for interest on U.S. obligations and in-District municipal bonds.

  D.C. Code § 47-1803.02 — Gross income; deductions · DC Office of Tax and Revenue — 2025 D-40 booklet (Calculation I rate schedule, standard deduction table, additional standard deduction worksheet, Schedule I Calculation B lines 10 and 12)

• The TWO rate schedules of § 10-105 — (a)(1) for single and married-filing-separately, (a)(2) for joint, surviving spouse and head of household — asserted at band ceilings on each, including the 5.75% band and the fact that the two schedules give DIFFERENT tax on the same taxable income.

  Tax years 2025, 2026 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.
  • The § 10-105(a)(3) capital-gains surtax, which applies only where federal adjusted gross income exceeds $350,000 and the return includes net capital gain, subject to the subsection’s own exclusions. The engine models no capital gains at all.
  • Maryland’s mandatory tax TABLES, which round within $50 income steps for taxable income under $100,000. These cases certify the rate SCHEDULE computed exactly, so a real filed return can differ from them by a few dollars.

  Md. Code Tax-General § 10-105 — State income tax rates

• The § 10-217(b) standard deduction of $3,350 for single and married-filing-separately and $6,700 for a joint return, head of household and qualifying surviving spouse, as stated in Instruction 16 of the 2025 Resident Booklet.

  Tax years 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 10-217(c) inflation indexing, which applies for taxable years beginning after 31 December 2025. The Comptroller’s 2026 withholding guide gives $3,400 for the single amount, but it sets WITHHOLDING and publishes one figure for every filing status, so the joint amount is unavailable — and (c) rounds each amount DOWN to $50 separately, so it cannot be inferred by doubling. The engine holds the 2025 amounts forward as a labelled projection baseline and NO case certifies a 2026 deduction (#354).
  • Itemized deductions under § 10-218, including the reduction that applies from tax year 2025 to filers above $200,000 of federal AGI. The engine models the standard deduction only.
  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.

  Md. Code Tax-General § 10-217 — standard deduction

• The § 10-211 exemptions — $3,200 per filer, an additional $1,000 for each filer aged 65 or over, and the federal-AGI phase-down of the BASE exemption through $1,600 and $800 to $0 — INCLUDING that the phase-down does not reach the aged addition.

  Tax years 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Dependants, and the additional $3,200 for a dependant aged 65 or over. The breakdown carries no dependants.
  • The § 10-211(b)(4) blindness exemption, which is also outside the phase-down. Blindness is not an engine input.
  • The § 10-211(a)(2) exemption for a spouse on a separate return. The engine models only single and joint.
  • The exemption adjustment for interest from U.S. obligations, noted at code hh of Instruction 13.
  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.

  Comptroller of Maryland — 2025 Resident Booklet, Instruction 10 (additional exemptions, tested "on the last day of the taxable year") and Instruction 27 (FILING RETURN OF DECEASED TAXPAYER) · Md. Code Tax-General § 10-211 — exemptions · Comptroller of Maryland — 2025 Resident Booklet (Instruction 16, Exemption Amount Chart 10A, Pension Exclusion Computation Worksheet 13A)

• The § 10-209 pension exclusion as Worksheet 13A computes it: a separate column per spouse, capped at the Comptroller’s annual maximum ($41,200 for 2025, $40,600 for 2026), reduced dollar-for-dollar by that person’s OWN GROSS Social Security and Railroad Retirement benefits, limited to their own qualifying income, gated on being 65 or older at the close of the year, and covering TSP withdrawals because the TSP is a § 401(a) plan.

  Tax years 2025, 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The "totally disabled" route into the exclusion, and the separate § 10-209 exclusion for retired forest, park and wildlife rangers aged 55 to 64 (Worksheet 13E). Disability is not an engine input.
  • The military-retirement and public-safety subtractions (code letters u and v), which Worksheet 13A Line 1 tells the filer to keep out of qualifying income.
  • Railroad Retirement Tier I and Tier II specifically. The offset is asserted using Social Security; the rule adds no separate Railroad Retirement input.
  • That IRAs, Roth IRAs, SEPs, Keogh plans and ineligible deferred compensation do NOT qualify. The engine models none of them, so the exclusion cannot wrongly reach them.
  • Any ABSOLUTE 2026 tax figure. The 2026 cap of $40,600 is sourced and certified, but the 2026 standard deduction is not (#354), so the 2026 case asserts the year-over-year DIFFERENCE — in which the deduction cancels — rather than a total.
  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.

  Comptroller of Maryland — 2025 Resident Booklet, Instruction 10 (additional exemptions, tested "on the last day of the taxable year") and Instruction 27 (FILING RETURN OF DECEASED TAXPAYER) · Md. Code Tax-General § 10-209 — subtraction for retirement income · Comptroller of Maryland — 2025 Resident Booklet (Instruction 16, Exemption Amount Chart 10A, Pension Exclusion Computation Worksheet 13A) · Comptroller of Maryland — Maryland Pension Exclusion (KB0010012)

• That § 10-207(j) keeps Social Security and Railroad Retirement out of the Maryland base entirely, so benefits alone are untaxed and adding a benefit raises no Maryland tax through the base — while the SAME figure reduces the § 10-209 exclusion, which is certified separately.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Every other subtraction in § 10-207 — the two-income subtraction of up to $1,200 on Form 502 line 14, U.S. obligation interest, the pickup-contribution subtraction of § 10-207(l), and the rest. None is modelled.
  • Additions under § 10-204, including the three new for tax year 2025.
  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.

  Md. Code Tax-General § 10-207 — subtractions from federal adjusted gross income

• That a qualified surviving spouse — the Comptroller’s Filing Status 5, and the IRC § 2(a) person the projection attests to — takes the JOINT rate schedule of § 10-105(a)(2), the $6,700 standard deduction of § 10-217(b)(2) and the joint exemption phase-down bands of § 10-211(c)(2), while the DECEASED SPOUSE generates no exemption of their own, no second aged addition and no second Worksheet 13A column, because § 10-211(a) allows a spouse exemption only where that spouse is living. This is a claim about the spouse, NOT about the total: a real Filing Status 5 return also claims the § 10-211(a)(3) exemption for the qualifying child the status requires, and the engine models no dependants in any filing status.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • HEAD OF HOUSEHOLD, whom the same three provisions treat identically. The engine has no head-of-household filing status at all, so there is nothing to map — an unmodelled status rather than a mis-mapped one.
  • The § 10-211(a)(3) exemption for the dependent child that Filing Status 5 requires the survivor to have. Dependants are not an engine input, so the modelled exemption is one short for every filer in this status — understating their deduction and overstating their tax.
  • Maryland Filing Status 3 (married filing separately) and Status 6 (dependent taxpayer), neither of which the engine models.
  • Which SLOT the survivor occupies is no longer an exclusion. It WAS one: the projection zeroes a deceased person’s age and birth date, so when person A died first the live survivor was person B and every age-gated provision here read person A — losing the pension exclusion and the aged exemption entirely, $6,155.25 against the correct $2,884.75 on a $90,000 pension at 70. #356 fixed that at the tax boundary rather than in this mapping, so these cases now hold for a survivor in EITHER slot. The evidence is the slot-invariance property in projection.test.ts, not a case in this pack.
  • The DEATH YEAR under the projection’s `immediate` filing-status mode, which labels that year `surviving_spouse`. IRC § 2(a) gives qualified surviving spouse status for the two years AFTER the year of death; the death year itself is normally a joint return. That label is a deliberate, disclosed modelling choice in the projection rather than a statutory reading, and Maryland now maps it to Filing Status 5 — so the death year is modelled conservatively, not certified.
  • The § 10-106 COUNTY income tax, which this model does not levy at all. Each of Maryland’s 23 counties and Baltimore City sets its own rate between 2.25% and 3.30%, and may levy it on brackets; FERSCalc resolves tax to the STATE level and does not ask for a county in any jurisdiction. Every figure in these cases is Maryland STATE tax alone, so Maryland tax is UNDERSTATED by the whole county share — disclosed on the public state page as Maryland’s largest gap rather than approximated by a default rate. See #351.

  Md. Code Tax-General § 10-105 — State income tax rates · Md. Code Tax-General § 10-211 — exemptions · Comptroller of Maryland — 2025 Resident Booklet, FILING STATUS chart (Filing Status 5) and Instruction 16

• Maryland FAMLI’s employee half, the eighth state payroll charge FERSCalc models and the SECOND with a start year (#429). Certified in five parts. FIRST, that DEDUCTIONS BEGIN IN JANUARY 2027 and nothing is charged before — an ABSENT programme rather than a zero rate, pinned on both sides of the boundary. Maryland’s own declaration recorded this shape in #429’s first phase, when it was the state that proved a rate table needs a start year at all; Minnesota implemented it first and this is the case it was written for. SECOND, the rate: "0.9% of wages up to the Social Security wage cap", "split equally between employers and employees (0.45% each)", and the employee half is what is stored, because Maryland publishes it as a rate in its own right rather than as a share of the total. THIRD, that the WAGE is capped at the Social Security wage cap in Maryland’s own words — a figure Maryland does not set, so the rule reads the engine’s published SSA table; 2027 is past the last published base, so the case asserts that the cap BINDS rather than a dollar amount, which two very different wages above it confirm. FOURTH, that the base is WAGES and the cap is applied to each earner’s own. FIFTH, that federal employment is outside it: Maryland requires every employer with an employee in the State to register, "no exceptions under state law", and a state cannot reach the federal government.

  Tax years 2027 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE START DATE ITSELF, which Maryland has moved more than once. The figures here were re-checked at the agency for this change rather than carried forward, and the start year is a named constant so a further delay is a one-line edit — but a projection produced today assumes 2027 holds. If it slips again this model charges the contribution EARLY.
  • THE PROJECTED SOCIAL SECURITY CAP. 2027 is beyond the last base SSA has published, so the ceiling is the engine’s projection and is disclosed as an estimate, exactly as it is on the FICA path.
  • WHETHER THE EMPLOYMENT IS MARYLAND EMPLOYMENT. Maryland states there are no exceptions for an employer with an employee in the State, so the uncertainty here is narrower than in most states — but the model has no way to place a spouse’s workplace, and charges the contribution for a Maryland household.
  • DEDUCTIBILITY ON A FEDERAL RETURN by an itemizing taxpayer; see the note on `StatePayrollWithholding`.

  Maryland FAMLI — Make contributions: "Starting January 2027, you will collect employee contributions through payroll deductions"; "0.9% of wages up to the Social Security wage cap"; "split equally between employers and employees (0.45% each)" · Maryland FAMLI — For employers: "If you have at least one employee in Maryland, you are required to register with FAMLI. There are no exceptions under state law."

• Both 2025 rate schedules, asserted at every band ceiling. TWO OF THOSE CEILINGS ESTABLISH THE FLOOR RATHER THAN THE RATE: at the 1% ceiling of each schedule the exemption credits ($153 single, $306 joint) exceed the tax ($110.79 and $221.58), so the return is zero and any schedule tax below the credit would give the same answer. Those two cases certify nonrefundability; the 1% band itself is established by the next case in each schedule, whose printed constant has the first band built into it. Schedule X and Schedule Y each print the cumulative tax at every band floor, so each case reproduces by arithmetic the constant the NEXT line of the published schedule prints — $110.79, $414.49, $1,022.01, $1,987.41, $3,201.97, $30,986.19, $38,638.27 and $72,219.84 for X; $221.58, $828.98, $2,044.02, $3,974.82, $6,403.94, $61,972.37, $77,276.52 and $144,439.65 for Y. A wrong rate or threshold could not land on the printed figure. This replaces California’s 2023 tables, which the rule had been running: the 2023 booklet prints exactly the $10,756 and $21,512 first-band ceilings it held. Includes the joint schedule’s 11.3% band running to $1,485,906 rather than stopping at $1,000,000, which was not staleness but a conflation of the rate schedule with the Behavioral Health Services Tax threshold.

  Tax years 2025 · 18 cases · re-check due 2026-12-01

  Not covered by this claim:

  • California’s rounded tax TABLES, which the booklet directs a filer to use at taxable income of $100,000 or less. These cases compute the rate schedules exactly, so a real filed return can differ from them by a few dollars.
  • Schedule Z, head of household. The engine carries single, joint and qualifying surviving spouse only; a head of household is treated as single here and OVERCHARGED, since Schedule Z is wider than Schedule X at every band. Identifying the status would need a question the calculator does not ask.
  • THE ROUNDING REGIME IS NOT CERTIFIED, and these cases are not filed Form 540 line 31 amounts. The FTB directs a filer to round the marginal step to cents, ADD the printed cumulative constant, and enter a whole-dollar result; this model integrates the bands exactly and never rounds. So the asserted $72,219.827 is the schedule’s own arithmetic where the booklet prints $72,219.84 and a return would carry $72,220. What is certified is the TABLE — the rates and the breakpoints — which is what the cases discriminate; the arithmetic regime on top of it is excluded, alongside the tax TABLES the booklet requires below $100,000 of taxable income.

  FTB — 2025 Form 540 booklet, 2025 California Tax Rate Schedules X and Y

• The 2025 Standard Deduction Chart for Most People — $5,706 for a single filer and $11,412 for a joint return — replacing the $5,363 / $10,726 the rule held, which the 2023 booklet prints verbatim. Includes that filing status 5, qualifying surviving spouse, takes the JOINT amount and Schedule Y rather than a single filer’s, which on $60,000 of annuity income is the difference between $444.18 and $1,639.53.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Itemized deductions, which California allows on its own modified basis and this model does not attempt.
  • The Standard Deduction Worksheet for Dependents, which applies where someone else can claim the filer.
  • Head of household, whose $11,412 amount this rule would give correctly but whose Schedule Z it would not.

  FTB — 2025 Form 540 booklet, California Standard Deduction Chart for Most People (Form 540 line 18)

• The Form 540 exemption credits, which the rule did not model AT ALL — $153 per personal exemption (line 7) and $153 for each spouse aged 65 or older (line 9), so a retired couple both over 65 was overcharged by UP TO $612 every year — the full amount wherever pre-credit tax exceeded it and the credits had not phased out, and less or nothing where it did not. Certified per person, not per return: neither-65, one-65 and both-65 give three different answers on identical income. Includes two readings that come from the FORM rather than the booklet or the Code — that a QUALIFYING SURVIVING SPOUSE claims TWO personal credits though one person is on the return ("If you checked box 2 or 5, enter 2 in the box"), and that the senior credit uses the IRC § 63(f) day-before-the-birthday convention, which the line 9 footnote states in its own words. Also certifies the line 32 AGI Limitation Worksheet: the $2,500 steps round UP, so one dollar of federal AGI over $252,203 (single) or $504,411 (joint) costs a full $6 per exemption; the reduction scales with the exemption COUNT rather than the dollar total; and the credit floors at zero rather than becoming a surcharge.

  Tax years 2025 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Blind exemptions (line 8) and dependent exemptions (line 10), neither of which the calculator collects. The AGI limitation implemented here therefore runs the worksheet’s lines a-i and not its dependent branch, lines j-m.
  • THE SENIOR CREDIT FOR A DECEDENT STILL ON THE RETURN. The § 63(f) reading certified here is the LIVING one. California’s own final-return instructions were read — the 2025 booklet’s "Deceased Taxpayers" section — and they say nothing about the age a decedent is treated as attaining for line 9: they cover who files, what to print beside the name, and Form 1310. So the rule deliberately does not call `federalAgeAtEndOfTaxableYear`, whose other branch would grant an age at death, and a decedent falls through to the plain age instead. #360, #388 and #393 each found that jurisdictions genuinely diverge here, so inheriting another state’s reading on no California authority would be the mistake those issues were about.
  • The head-of-household threshold of $378,310, unreachable because the status is not modelled.
  • WHOLE-DOLLAR SEMANTICS. The worksheet says “Use whole dollars only”, and this rule compares the raw Decimal federal AGI instead, so a FRACTIONAL amount a cent above $252,203 starts the phase-out where a filed return, carrying whole dollars, would not. Every certified boundary case here uses integers and so cannot see it. Projections do produce fractional dollars, so this is reachable; it is engine-wide rather than Californian — no state rule rounds to the return’s own unit — and is filed rather than fixed inside a one-jurisdiction change.
  • The RDP recalculation of federal AGI on a pro forma return, which FTB Pub. 737 governs and the engine has no basis to perform.

  FTB — 2025 Form 540, lines 7 through 10 (the preprinted $153 and the line 7 exemption count) · FTB — 2025 Form 540 booklet, line 9 senior exemption instruction and the line 32 AGI Limitation Worksheet

• That Social Security never enters the California TAXABLE-INCOME BASE — Schedule CA (540) Part I line 6, "California excludes U.S. social security benefits or equivalent Tier 1 railroad retirement benefits from taxable income" — and that the FERS annuity and TSP withdrawals receive NO CALIFORNIA RETIREMENT-INCOME EXCLUSION at any age, and are treated identically to each other, because lines 4a/4b and 5a/5b say "Generally, no adjustments are made on this line". THE WORDING IS DELIBERATELY NOT "taxed in full": what Schedule CA declines to adjust is the FEDERALLY TAXABLE amount, and production hands this rule the annuity NET of the Simplified Method cost recovery (#348) and only the taxable part of a TSP withdrawal. California taxes what the federal return already taxed, no more and no less — it simply subtracts nothing further for being retirement income. The Social Security half confirms behaviour the code always had, against a header comment that until `e8d22b1` asserted the opposite in the largest jurisdiction the calculator models.

  Tax years 2025 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE WORD "BASE" IS LOAD-BEARING, and this claim must not be read as "Social Security cannot change California tax". It cannot enter line 19. But the line 32 AGI Limitation Worksheet keys on FEDERAL adjusted gross income, which carries the § 86 taxable portion of benefits, so benefits CAN phase out the exemption credits and so raise California tax — a route that runs through line 32, not line 19. No case here reaches it: the certified phase-out cases carry no Social Security, so they cannot distinguish an adapter that passes real federal AGI from one that reconstructs it without benefits. Production passes the real figure (the income ledger adds the taxable portion), which is what makes the narrow wording the accurate one.
  • The military retirement exclusion new for 2025 (R&TC §§ 17132.9/17132.10, up to $20,000, capped at $125,000/$250,000 of federal AGI). Deliberately not modelled: no input carries military retired pay or a DoD Survivor Benefit Plan annuity, so there is nothing to exclude. Recorded so a later pass does not "fix" it in without the input existing.
  • The three-year annuity rule for annuities commencing between 1 July 1986 and 1 January 1987, and the IRA basis restatement for a former nonresident (FTB Pub. 1005 / 1100).
  • The form FTB 3805P additional tax on a distribution taken before 59½.

  FTB — 2025 Schedule CA (540) instructions, Part I lines 4a/4b, 5a/5b and 6

• State Disability Insurance — which INCLUDES Paid Family Leave, per DE 231EE’s own footnote, so one rate covers both. Certified in five parts. FIRST, that it is not a Form 540 liability: it is returned by `applyPayrollWithholding` and never by `applyTax`, so $60,000 of wages costs exactly what $60,000 of annuity income costs — including when a covered non-federal spouse is present and genuinely owes the withholding. Before #409 the rule added 1.1% of wages PLUS other income to the income tax itself, charging interest as payroll; that conflation cannot return without failing a case. SECOND, that a federal employee is exempt, now by DERIVATION from `isFederalPersonA/B` rather than by the charge’s absence — "Federal employees are exempt from UI, ETT, and SDI" (DE 231EE, Rev. 9 (9-19)) — asserted at two wage levels. THIRD, the RATE and its wage limit: "The SDI withholding rate for 2026 is 1.3 percent" and "for 2025 is 1.2 percent", so the rate is a function of the year; there is NO taxable wage limit, "Effective January 1, 2024, Senate Bill 951 removes the taxable wage limit and maximum withholdings", proved at $500,000 of wages where a capped rule would stop near $1,600; and beyond the last published year the latest rate holds. FOURTH, that the BASE IS WAGES and nothing else: $600,000 of pension, TSP, Social Security and interest beside $100,000 of wages costs $1,300 and not a cent more, and a household with no wages at all owes nothing however large its retirement income — the case that stops the pre-#409 `wages + other_income` base returning. FIFTH, that the charge is PER PERSON and not per household — a federal earner at $100,000 beside a non-federal one at $100,000 owes $1,300, where a household total would give $2,600 or nothing — and that an exempt-employment assertion removes one person’s charge without touching the other’s. THIS FIXES A REAL UNDERSTATEMENT: a dual-income California household owed this every working year and the model charged nothing, in the direction that flatters a retirement plan (#412).

  Tax years 2025, 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • WHETHER A NON-FEDERAL JOB IS SDI-COVERED AT ALL, which is not derivable and is not asked. Under CUIC § 2606(a), service for a public entity as defined by § 605 — the State, a county, a city, a school or community college district — is "employment" for the Disability Compensation part ONLY as a Health and Safety Code Division 23 hospital district, a public housing agency, a state employee under § 2781 (which is Nonindustrial Disability Insurance, not SDI withholding), or under an elective coverage agreement; a public school employer or public agency may elect in under §§ 709, 710.4 and 710.5, and many have. DE 231EE adds family employment (spouse employed by spouse) and church employment. `employmentType` is only federal or non-federal, so none of these is visible. The default CHARGES, which overstates the withholding for such a household by the full amount, and the exempt branch is priced on the Summary tab rather than collected — docs/INPUT_BURDEN_POLICY.md step 3, with the register entry and the branch test there.
  • THE WAGE BASE IS THE ENGINE’S, NOT EDD’S DEFINITION OF PIT WAGES. `wagesPersonA/B` is salary plus post-retirement earned income — the same figure FICA is charged on — and the model does not represent self-employment income, which SDI reaches only through Disability Insurance Elective Coverage. Nor does it model a Voluntary Plan substituted for SDI under CUIC § 3251, whose employee contribution may differ.
  • ELECTIVE COVERAGE generally (DE 231SC), by which a person in otherwise exempt employment can opt into SDI. It would make an exempt person covered, and nothing collected could reveal it.
  • Every other California payroll item — UI and ETT are employer taxes, and the calculator models neither.
  • RATE YEARS BEYOND 2026. EDD sets the rate annually and has moved it in each of the last several years; later projection rows hold 1.3%, so a long projection asserts only "at today’s rate". Stated in the rule’s notes and certified as the hold, not as a forecast.

  EDD — Information Sheet: Exempt Employment (DE 231EE), Rev. 9 (9-19): "Federal employees are exempt from UI, ETT, and SDI", and the footnote "* Includes Paid Family Leave (PFL)". Read as rendered page images, not by text extraction. · EDD — Rates and Withholding: "The SDI withholding rate for 2026 is 1.3 percent"; "The SDI withholding rate for 2025 is 1.2 percent"; "Effective January 1, 2024, Senate Bill 951 removes the taxable wage limit and maximum withholdings for each employee subject to SDI contributions." · California Unemployment Insurance Code § 2606(a) — the definition of "employment" for the Disability Compensation part, which reaches public-entity service only in four listed cases (with §§ 605, 709, 710.4, 710.5 and 2781)

• Form 540 line 62 — 1% of taxable income over $1,000,000, which the rule did not model at all. Certified as to WHERE the threshold sits (nothing at exactly $1,000,000, and a case that fails if it moves), at 1% of the excess, and as keying on TAXABLE income rather than gross: the same $1,008,000 of income owes the tax as a single filer and does not as a joint one, decided by nothing but the standard deduction. Renamed from the Mental Health Services Tax for taxable years beginning on or after 1 January 2025, when the Mental Health Services Act became the Behavioral Health Services Act; the tax is unchanged.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE ORDERING AGAINST THE EXEMPTION CREDITS IS NOT CERTIFIED, and cannot be. Form 540 puts this tax at line 62, after line 33 has subtracted the credits, so the credits cannot reduce it — and the rule implements that. But the tax needs taxable income over $1,000,000 while the credits are fully phased out above $566,911 of joint AGI, so the two provisions have no reachable overlap and a case asserting the order would pass whichever order the code used. Stated here rather than tested inertly.
  • THE SAME WHOLE-DOLLAR GAP APPLIES HERE. Line 62 also says “Use whole dollars only”, and this rule compares raw Decimal taxable income, so a fractional cent above $1,000,000 charges the tax where a filed return would show exactly $1,000,000 and owe nothing. Reachable, engine-wide, filed as #411 rather than fixed here.
  • NOR IS STRICT-VERSUS-INCLUSIVE AT THE THRESHOLD ITSELF, for the same reason and discovered the same way. "More than $1,000,000" and "at least $1,000,000" are the SAME FUNCTION, because at exactly $1,000,000 the excess is zero and 1% of nothing is nothing. A mutation flipping the comparison SURVIVED the pack, correctly — it was testing nothing — and was replaced by one that moves the threshold, which the case does catch. What "not charged at exactly $1,000,000" certifies is where the threshold sits, not which side of it the boundary falls on.

  FTB — 2025 Form 540 booklet, line 62 Behavioral Health Services Tax computation and the Behavioral Health Services Act renaming note

• The flat 4.95% rate (Form IL-1040 line 12, "Multiply Line 11 by 4.95% (.0495)") and the $2,850-per-person basic exemption for 2025 (line 10a chart; 35 ILCS 5/204(b)(7), $2,050 plus the cost-of-living adjustment), doubled on a joint return. Illinois has NO standard deduction — the Section 204 exemption allowance is the whole of it — and line 11 net income "may not be less than zero".

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The dependent branch of the line 10a chart, which turns on the Step 1 Line C box ("someone can claim you as a dependent") and is not modelled.
  • Indexing beyond 2025. § 204(b)(7) adjusts the basic amount for cost of living each year through 2028; the 2025 figure is held constant for later years, which understates the exemption and so overstates tax — the conservative direction.

  Illinois Department of Revenue — 2025 Form IL-1040 (lines 10a-10d, line 11, line 12) · 35 ILCS 5/204 — Standard exemption

• The $1,000 additional exemption for each spouse aged 65 or older — Form IL-1040 line 10b and 35 ILCS 5/204(d)(1)(A) — which the rule did not model AT ALL. Certified per person: neither-65, one-65 and both-65 give three answers $49.50 apart. ALSO CERTIFIED IS THE AGE CONVENTION, which is Illinois’s own and is NOT IRC § 63(f): the statute says "attained the age of 65 before the end of the taxable year" and the form carries no birthday proviso, so a 1 January 1961 birthday gets nothing extra for 2025 — the opposite of California, whose line 9 footnote deems exactly that case into the prior year (#409). A case pins the difference so neither state’s convention can drift into the other. The provision is reached through `other_income` rather than wages, because Illinois subtracts every retirement stream and interest is the only base income an Illinois retiree ordinarily has.

  Tax years 2025 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The blind additional exemption of § 204(d)(2), the same $1,000 again per person, which the calculator does not collect.
  • THE DECEDENT READING IS A READING, not a quoted instruction. § 204(d)(1) conditions the addition on the person having "attained the age of 65 before the end of the taxable year", and a decedent’s taxable year ends at death, so the rule grants it via `ageAtEndOfTaxableYear`. The IL-1040’s "Filing a decedent’s return" expressly provides for the joint return and places no restriction on the exemption allowance — but it does not address the allowance either way. Granting it reads the statutory condition as being about the person; denying it would be an inference too, and a worse one, since it would withhold a statutory exemption on no authority.

  Illinois Department of Revenue — 2025 Form IL-1040 line 10b · 35 ILCS 5/204(d)(1) — additional exemption for taxpayer or spouse 65 years of age or older

• 35 ILCS 5/204(g), absent from the rule before #413: "for taxable years beginning on or after January 1, 2017, no taxpayer may claim an exemption under this Section if the taxpayer’s adjusted gross income for the taxable year exceeds (i) $500,000, in the case of spouses filing a joint federal tax return or (ii) $250,000, in the case of all other taxpayers." Certified as a CLIFF rather than a taper (one dollar over costs $381.20 on a joint return), as keying on FEDERAL adjusted gross income rather than Illinois base income, and as taking the age-65 additions with it — "an exemption under this Section" is the whole of Section 204, so a household above the threshold keeps nothing rather than keeping the $1,000 additions.

  Tax years 2025 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The thresholds are NOT indexed by the statute and are held as enacted; if Illinois amends them this claim is what has to move.

  35 ILCS 5/204(g) — denial of the standard exemption above $500,000 / $250,000 of adjusted gross income · Illinois Department of Revenue — 2025 IL-1040 Instructions, the "Income Exceptions" box at line 10

• That the FERS annuity, TSP withdrawals and Social Security all leave Illinois base income, AND THAT THEY DO SO UNDER TWO DIFFERENT PARAGRAPHS. 35 ILCS 5/203(a)(2)(F) subtracts "distributions under the provisions of any retirement or disability plan for employees of any governmental agency or unit" plus amounts under IRC §§ 402(a), 402(c), 403(a), 403(b), 406(a), 407(a) and 408 — precise for the annuity and the TSP, and silent on benefits. Social Security is § 203(a)(2)(L), "all social security benefits and railroad retirement benefits included in such total pursuant to Sections 72(r) and 86". The rule’s public notes cited (F) for all four streams; one paragraph was doing the work of two, and the wrong one was on the public page.

  Tax years 2025 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE CITATION CORRECTION IS NOT FIXTURE-BACKED, and cannot be. Both paragraphs produce the same answer — the income leaves the base either way — so no case can distinguish a rule that cites (F) for Social Security from one that cites (L). That is exactly why the wrong citation survived. The cases pin the exemptions; this statement carries the correction.
  • Illinois taxes some retirement-adjacent income these cases do not reach, including the beneficiary share of certain retirement plans reported on Schedule K-1-T and retirement payments to retired partners.

  35 ILCS 5/203(a)(2)(F) and (L) — retirement plan and Social Security subtractions · Illinois Department of Revenue — 2025 IL-1040 Instructions, line 5 "Social Security benefits and certain retirement plans"

• The year-scheduled rate — 5.39% (2024), 5.19% (2025), 4.99% (2026) — and the year-scheduled standard deduction — $12,000 single / $24,000 joint (2024-2025), $15,000 single / $30,000 joint (2026) — per O.C.G.A. §§ 48-7-20(a.1) and 48-7-27(a)(1)(B) as amended by HB 463, holding both flat for 2027 and later rather than assuming the conditional further reductions/increases HB 463 only gestures at.

  Tax years 2024, 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • HB 463’s further annual rate reduction toward 3.99% and further standard-deduction increase toward $18,000/$36,000, both beginning January 1, 2027. Each STEP’S AMOUNT is fixed by the statute — 0.125 percentage points on the rate, $375/$750 on the deduction — but WHETHER a given year’s step actually occurs is not: it is delayed a year for any year the Office of Planning and Budget finds a December 1 revenue trigger unmet (§§ 2-1, 2-3), so which of the fixed amounts applies to any post-2026 year cannot be derived from the statute alone. The engine holds the 2026 figures flat rather than assume the schedule proceeds on its default track — the same treatment as Massachusetts’s and Hawaii’s own unpublished forward years.
  • The § 48-7-26(b) dependent exemption ($4,000, rising to $5,000 in 2026 under the same Act), which the engine does not model at all — FERSCalc models no dependants in any state.
  • Itemized deductions under § 48-7-27(a)(1)(A), which the engine never elects.
  • HB 463’s new §§ (16)-(17) of § 48-7-27(a): up to $1,750 of qualified overtime compensation and up to $1,750 of cash tips, both excluded for taxable years 2026 through 2028. The engine has no way to identify either category within wages, so it taxes both in full — an OVERSTATEMENT for a filer who earns either, most likely a federal retiree working an hourly post-retirement job.

  House Bill 463 (2025-2026 Regular Session), "AS PASSED HOUSE AND SENATE" — §§ 2-1 (rate), 2-3 (standard deduction), 5-1 (general applicability, "taxable years beginning on or after January 1, 2026"). Signed by the Governor 11 May 2026 · Georgia Department of Revenue — 2025 IT-511 Individual Income Tax Instructions Booklet ("the income tax rate is 5.19%"; Form 500 Line 11, "Enter $12,000 if the filing status ... is A, C, or D. If the filing status is B, enter $24,000") · Georgia Department of Revenue — 2024 IT-511 Individual Income Tax Instructions Booklet ("the income tax rate is 5.39%"; same $12,000/$24,000 standard deduction)

• The O.C.G.A. § 48-7-27(a)(5) retirement income exclusion AS APPLIED TO THE PENSION, TSP AND WAGE-DERIVED-EARNED-INCOME STREAMS THE ENGINE MODELS: none under 62 for a non-disabled filer, $35,000 for 62-64, $65,000 for 65+ through 2026, $70,000 for 65+ from 2027, of which at most $5,000 per person may be earned income (division (E)(i)) — and that the exclusion is granted PER PERSON against that person’s own income and that person’s own age BAND VALUE, never pooled and never a value leaked from the other spouse’s band, per (a)(5)(B), "each spouse shall if otherwise qualified be individually entitled to exclude retirement income received by that spouse up to the exclusion amount." The 62-64 band ($35,000) is unchanged by HB 463 in every certified year, including 2027.

  Tax years 2024, 2025, 2026, 2027 · 11 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The under-62 permanent-disability route, (a)(5)(D)(ii). Georgia requires its own medically demonstrable standard and Form 500 asks for a disability date; the engine collects neither, so a disabled filer under 62 is under-granted. Tracked for a future priced comparison at #371.
  • The separate military-retirement exclusion — an entirely different provision from the (a)(5) civilian exclusion this claim certifies, and one that itself changes mid-range: § 48-7-27(a)(5.1) grants up to $17,500 under 62 (doubled with sufficient Georgia earned income) through 2026, but HB 266 (2025-2026 Regular Session) raises this to $65,000 for taxpayers under 65, without the earned-income condition and barred from stacking with the (a)(5) exclusion, beginning January 1, 2027 — the same year this claim’s own (a)(5) certification reaches. FERSCalc does not model military retirement pay in any state under either version, so this is out of scope by construction rather than a Georgia-specific gap.
  • § 48-7-27(a)(5)(E)(ii), income a surviving family member receives based on a deceased veteran’s military service record, excluded "without regard to the age of the surviving family member." Out of scope for the same reason.
  • § 48-7-27(a)(5)(E)(i)’s BROADER statutory definition of "retirement income" — TAXABLE interest, dividend, net rental, capital gains and royalty income all qualify for this exclusion by statute, in addition to pension/annuity income and the capped earned-income share. (TAX-EXEMPT interest is a separate matter, correctly outside this gap: IT-511’s own worksheet instructs that it not be included in the retirement-exclusion calculation at all, so it needs no allocation here.) `gaPersonExclusion` in state-taxes.ts only ever sums pension and TSP income as a person’s "unearned" retirement income; the engine has no per-person allocation of taxable interest, dividends, rental income, capital gains or royalties, so a retiree with any of those income types MAY get less exclusion than the statute allows, and so MAY be overstated — but only if that person’s own cap is not already exhausted by pension/TSP income alone, which several certified cases above show it often is.
  • Whether the 2027-and-later further escalation of the standard deduction (see the rate-and-deduction claim) interacts with this exclusion’s own dollar figures — it does not, by statute; the two provisions move independently.

  Georgia Department of Revenue — 2025 IT-511 Individual Income Tax Instructions Booklet, "Retirement Income Exclusion": the $35,000/$65,000 bands, the age bands, and "each must qualify on a separate basis" · House Bill 463 (2025-2026 Regular Session), "AS PASSED HOUSE AND SENATE" — § 2-3, amending O.C.G.A. § 48-7-27(a)(5)(A) divisions (xiii)-(xiv): $65,000 "ending on or before December 31, 2026" and $70,000 "beginning on or after January 1, 2027," both against an unchanged $35,000 62-64 band · O.C.G.A. § 48-7-27(a)(5)(B), (D), (E), (5.1) — cited to Justia’s 2024 Georgia Code. Georgia DOES have a free official portal (lexisnexis.com/hottopics/gacode, confirmed by direct navigation to open the current Official Code of Georgia Annotated at no cost), but it loads results into a session-scoped Lexis Advance URL rather than a stable per-section address, so Justia is used for its citable, human-readable URL — not because no free portal exists, which an earlier draft of this pack wrongly concluded after one check happened to bounce through a sign-in redirect. The (D) age-eligibility test and (E) earned-income sub-cap are NOT touched by HB 463, which amends only the dollar-amount divisions of (a)(5)(A)

• That Social Security never enters the Georgia base and never enters the retirement-exclusion calculation either. § 48-7-27(a)(7) SUBTRACTS "Social security benefits and tier 1 railroad retirement benefits, to the extent included in federal taxable income" — a deduction Georgia takes, not an absence from the federal AGI § 48-7-27(a) starts from: the federally TAXABLE portion of Social Security is already part of federal AGI, and Georgia removes it separately under (a)(7). IT-511’s own retirement-exclusion worksheet independently confirms Social Security is excluded from that calculation too. Adding Social Security to a household therefore changes the tax by nothing, and a Social-Security-only household owes nothing however large the benefit.

  Tax years 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Every § 48-7-27(a) subtraction other than (a)(7) — the federal-jobs-tax-credit and mortgage-interest-credit adjustments of paragraph (3), which the engine does not model.

  Georgia Department of Revenue — 2025 IT-511 Individual Income Tax Instructions Booklet, Retirement Income Exclusion worksheet ("Social Security and Railroad Retirement paid by the Railroad Retirement Board, exempt interest, or other income that is not taxable to Georgia should not be included in the retirement income exclusion calculation") · O.C.G.A. § 48-7-27(a)(7) — "Social security benefits and tier 1 railroad retirement benefits, to the extent included in federal taxable income" are subtracted, via Justia’s 2024 Georgia Code (mirror; see the retirement-exclusion claim’s sourcing note)

• That a decedent keeps their own age on Georgia’s final return, per § 48-7-27(a)(5)(D)’s any-moment-in-the-year eligibility test — satisfied outright by anyone who reached the qualifying age before dying — read PER PERSON as (a)(5)(B) already requires, so on a joint return a decedent’s age sets only their own exclusion band and never reaches the surviving spouse’s. Both certified cases reach 65, so the operative text is division (iii), "65 years of age or older DURING ANY PART OF THE YEAR" — one word shorter than division (i)’s 62-64 test, "during any part of the TAXABLE year."

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether a decedent’s final return may itself become a qualifying-surviving-spouse return the FOLLOWING year — out of scope for this claim, which is about the final-year return alone.

  O.C.G.A. § 48-7-27(a)(5)(D)(iii) — "Is 65 years of age or older during any part of the year" — via Justia’s 2024 Georgia Code (mirror; see the retirement-exclusion claim’s sourcing note) · Georgia Department of Revenue — 2025 IT-511 Individual Income Tax Instructions Booklet, "Filing for Deceased Taxpayers": "The surviving spouse, administrator, or executor may file a return on behalf of a taxpayer who dies during the taxable year"

• That a qualifying surviving spouse files Georgia Filing Status D and takes the SINGLE $12,000/$15,000 standard deduction, never the joint $24,000/$30,000 one — Form 500’s own filing-status key lists "D - Head of household or Qualifying surviving spouse," and Form 500, Schedule 3’s Line 10 instruction states the same rule in plain English, independently of Form 500’s own letter-coded Line 11. This is the OPPOSITE of Maryland, where a qualifying surviving spouse moves onto Filing Status 5 and the joint schedule (#353); each state was read on its own instructions, not by analogy.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household, which the same Filing Status D covers but which the engine does not model as a distinct status.
  • Whether a qualifying surviving spouse might also claim the qualifying-child dependent that produced federal QSS status — the engine models no dependants in any state.

  Georgia Department of Revenue — 2025 IT-511 Individual Income Tax Instructions Booklet, Form 500 Line 11 and filing-status key ("D - Head of household or Qualifying surviving spouse"), and Form 500, Schedule 3 Line 10 (independently, in plain English rather than letter codes) — settles the STATUS MAPPING and the 2025 dollar figure ($12,000/$24,000) · House Bill 463 (2025-2026 Regular Session), "AS PASSED HOUSE AND SENATE" — supplies the 2026 dollar figures ($15,000/$30,000) this claim also certifies; the Filing Status D mapping itself is untouched by HB 463

• The flat 3.07% rate, applied with no standard deduction, no personal exemption and no bracket: twice the taxable compensation is exactly twice the tax, and the first dollar is taxed.

  Tax years 2025 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Pennsylvania’s poverty-income forgiveness credit (PA Schedule SP), which can reduce or eliminate tax for a low-income household and is not modelled.
  • Local city, county, school-district and municipal earned-income taxes, which are out of scope in every jurisdiction by design.

  PA DOR — Personal Income Tax Guide, Gross Compensation

• That the FERS annuity and Social Security never enter the Pennsylvania base. THE ANNUITY RESTS ON THE PA-40’S OWN "INCOME NOT TAXABLE FOR PA PIT PURPOSES" LIST, which carries two consecutive bullets either of which suffices: "Commonly recognized pension, old age, or retirement benefits paid after becoming eligible to retire, and retiring" — exception 1 in list form, which a FERS retiree who retired satisfies directly — and, unconditionally, "Civil Service Annuity", which describes the PAYMENT rather than naming a legacy plan. The PIT Guide’s listing of the U.S. Civil Service Commission Retirement Disability Plan is corroboration and is the weakest of the three. A structural case additionally pins that `pensionIncome` is not read by the rule on ANY path, so no pension classification question can move a figure.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • NEITHER THE PIT GUIDE’S GROSS COMPENSATION CHAPTER NOR THE 2025 PA-40 INSTRUCTIONS WRITES "FERS" OR "FEDERAL EMPLOYEES RETIREMENT SYSTEM" — the scope checked, and the whole of the claim. AN EARLIER DRAFT OF THIS CLAIM SAID SOMETHING BROADER AND FALSE — that "the phrase ‘Federal Employees’ does not appear in the PIT Guide at all" — which was a failed search reported as an absence: the Guide writes "Federal Employee’s Thrift Savings Plan", singular possessive, which the plural pattern missed. Recorded because the correction is the point: the U.S. Civil Service Commission listing is NOT certified as identifying FERS, only as corroborating a conclusion the PA-40’s own list reaches without it.
  • THE LEGAL BASIS DIFFERS BY PENSION COMPONENT, and the structural case does not certify any of them: an ordinary or deferred FERS annuity travels by exception 1, and a SURVIVOR annuity by the PA-40’s separate exception for payments to a beneficiary. Only the ordinary-retiree route is certified here; the survivor route is asserted by neither a case nor this statement.
  • A FERS DISABILITY ANNUITANT would be the one population where the plan-name listing had to carry the weight, since they may not have "qualified for retirement and retired" in Pennsylvania’s sense. It is unreachable: `EligibilityType` is immediate_unreduced | mra_plus_10 | deferred | not_vested, and FERS disability retirement is not a modelled scenario. Recorded so a later pass that adds it knows this question comes with it.

  PA DOR — 2025 PA-40 instructions: "INCOME NOT TAXABLE FOR PA PIT PURPOSES", including "Commonly recognized pension, old age, or retirement benefits paid after becoming eligible to retire, and retiring" and "Civil Service Annuity" · PA DOR — Personal Income Tax Guide, Gross Compensation: income not taxable as compensation, naming the U.S. Civil Service Commission Retirement Disability Plan (corroboration only)

• That TSP withdrawals are exempt only after the taxpayer "qualified for retirement and retired", and that the test is PER PERSON — CERTIFIED GIVEN THE PREMISE that the TSP is an eligible employer-sponsored plan, which is disclosed below rather than certified. A household where one spouse is a qualified retiree and the other separated early is taxed on the early one’s withdrawals ALONE — a figure no household reading can produce, since a household rule would give either zero or the whole. Pennsylvania does not follow the federal early-distribution exceptions and 59½ is nowhere in its test; the gate is 61 Pa. Code § 101.6(m)(1), under which a person who separated before satisfying superannuation requirements is "deemed to be retired from service upon reaching retirement age".

  Tax years 2025 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE TSP’S ELIGIBILITY IS A PREMISE, NOT A CERTIFIED FACT. The PA-40 (page 11) DOES name it — "Eligible employer-sponsored retirement plans can, but do not necessarily, include ... 401(k) plans; thrift plans; thrift savings plans; and employee welfare plans" — but that names the CATEGORY, not this plan: "can, but do not necessarily" is expressly non-determinative and the next sentence directs the filer to ask their plan administrator. Pennsylvania publishes no determination that the FEDERAL Thrift Savings Plan in particular qualifies, so everything in this claim is conditional on the premise. The PIT Guide corroborates it: contributions to the "Federal Employee’s Thrift Savings Plan" are PA-taxable when made, the hallmark of an eligible plan and the same fact behind #379. TWO EARLIER DRAFTS OF THIS EXCLUSION WERE WRONG IN OPPOSITE DIRECTIONS — the first overclaimed by treating the naming as determinative, the second asserted the PA-40 does not name thrift savings plans at all, which was a failed PDF text search reported as an absence.
  • COST RECOVERY IS NOT MODELLED, and a case certifies that gap AS a gap so a later fix has something to fail against (#379). Pennsylvania exempts the portion of a distribution representing previously taxed contributions and taxes employee deferrals WHEN CONTRIBUTED, so a federal employee’s own TSP contributions are already PA-taxed basis returning tax-free. The calculator cannot know the composition of an entered opening balance, so a taxable distribution is taxed in FULL and Pennsylvania tax is OVERSTATED for an affected household. The previous behaviour understated it by the whole distribution instead.
  • Every non-federal account is treated as IRA-like at 59½ (#380).
  • The year containing the deeming date is treated as fully retired, where a real return would tax withdrawals taken earlier within it.

  PA DOR — 2025 PA-40 instructions: the eligible employer-sponsored retirement plan discussion, with its "can, but do not necessarily, include" and its referral to the plan administrator · 61 Pa. Code § 101.6(m)(1) — a person who separated before satisfying superannuation requirements is "deemed to be retired from service upon reaching retirement age" · PA DOR — Personal Income Tax Guide, Gross Compensation: the Federal Employee’s Thrift Savings Plan contributions rule

• Pennsylvania’s employee UNEMPLOYMENT COMPENSATION withholding, the sixth state payroll charge FERSCalc models and the only one besides California’s with NO CAP AT ALL (#429). Certified in four parts. FIRST, the rate and its absence of a ceiling: 0.07% of gross wages, "70 cents per $1,000", and "Employee contributions are not limited to the taxable wage base" — so $1,000,000 of wages owes $700 where every capped state would have flattened long before. Pennsylvania is uncapped AND flat, which no other modelled state is. SECOND, that the rate is the same in every modelled year, published as "0.07% for 2023 and thereafter". THIRD, that both earners are charged and the base is WAGES. FOURTH, and the reason this needed checking harder than anywhere else, that it is NOT a PA-40 liability AND does not reduce Pennsylvania taxable compensation: Pennsylvania taxes compensation directly rather than starting from a federal figure, so a deduction against compensation would have fed `applyTax`, and the PIT Guide allows exactly FOUR deductions — Medical Savings Account, Health Savings Account, § 529 and § 529A — of which this is not one. Federal employment is outside it: federal employees are covered by the separate federal UCFE programme under 5 U.S.C. ch. 85 rather than by a state’s unemployment law. Pennsylvania is one of only three states, with New Jersey and Alaska, where employees contribute to unemployment insurance at all — a category #429’s original survey never asked about, so this jurisdiction was CERTIFIED while carrying an undeclared payroll contribution.

  Tax years 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • NO EXEMPT BRANCH IS OFFERED, and that is a finding rather than an omission. Every other modelling state has an employment class the engine cannot identify; Pennsylvania has none on the employee side, because all employers withhold "regardless of whether an employer is contributory or reimbursable" and there is no employer-pickup provision. The rule therefore carries no `exemptBranch` and the branch module gates on that.
  • THE HOUSEHOLD COLLAPSE IS UNOBSERVABLE HERE, and the case says so rather than pretending otherwise. With no cap, 0.07% of two wages summed equals the sum of 0.07% of each, so a rule that collapsed the household would give the same answer. What the per-person case pins is that BOTH earners are charged, not the arithmetic of the split.
  • LOCAL EARNED-INCOME TAXES, which are out of scope in every jurisdiction by design and are separate from this contribution.
  • DEDUCTIBILITY ON A FEDERAL RETURN by an itemizing taxpayer; see the note on `StatePayrollWithholding`.

  PA Department of Labor and Industry — Employee Withholding: 0.07% "for 2023 and thereafter"; "Employee contributions are not limited to the taxable wage base"; "All employers are required to withhold employee contributions ... regardless of whether an employer is contributory or reimbursable" · PA Department of Revenue — PA Personal Income Tax Guide, Deductions and Credits: Pennsylvania allows four deductions (MSA, HSA, § 529, § 529A) and federal deductions are "not permitted in calculating the taxpayer’s Pennsylvania taxable income"

• The 2025 rate schedule (IT-201-I page 33), asserted at band ceilings that ROUND TO the NEXT line’s printed cumulative constant — $340, $484, $600 and $4,271 single; $686, $976 and $1,202 joint. TWO RATES WERE WRONG AND BOTH OVERSTATED TAX: the band from $13,900 to $80,650 single (and $27,900 to $161,550 joint) is 5.5% and was charged at 5.85%, and the band above it is 6% and was charged at 6.25%. The 5.5% band is where a New York federal retiree’s taxable income actually sits, so the error was not an edge case: it reached single or married-filing-separately returns with more than $13,900 of New York taxable income, and joint or surviving-spouse returns above $27,900 — not a household living on federal retirement income alone, whose New York taxable income is zero — about $161 a year on $60,000 of taxable income for a single filer, or $112.35 on the same joint taxable income. The schedule also continues past the old 9.65% ceiling into 10.3% above $5,000,000 and 10.9% above $25,000,000, neither of which was modelled. The 10.3% band is certified in the ny-recapture claim, where it is reachable.

  Tax years 2025 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE FILED-RETURN ARITHMETIC IS NOT CERTIFIED, on two counts. New York prints its cumulative constants TO THE DOLLAR and treats them as operands — "$600 plus 5.5% of the excess" — while this model integrates the bands exactly, computing $599.50 where the schedule says $600. AND BELOW $65,000 OF TAXABLE INCOME THE SCHEDULE IS NOT THE PRESCRIBED INSTRUMENT AT ALL: New York requires a $50-band TAX TABLE there, which this model does not implement, the same exclusion California and the District carry for their own rounded tables. Most cases in this pack sit under that threshold, so their absolute figures are the schedule’s answer rather than a filed return’s. What is certified is the TABLE — the rates and the breakpoints — which is what the cases discriminate.
  • THE BANDS ABOVE $107,650 OF NEW YORK AGI CANNOT BE ISOLATED, because the recapture applies to every return above that figure. They are certified in the ny-recapture claim instead, where the schedule tax is an input to the worksheet and the printed constants still verify. That is a fact about New York’s computation, not a gap in this pack.
  • Head of household, which the engine cannot express — it carries single, joint and qualifying surviving spouse only. New York gives a head of household its own schedule, so a HoH filer is treated as single here and OVERCHARGED.

  NYS DTF — 2025 Form IT-201-I page 33, New York State tax rate schedule (read as a page image)

• The standard deduction table (page 11): $8,000 for filing status 1 and $16,050 for status 2. Includes that filing status 5, QUALIFYING SURVIVING SPOUSE, takes the joint deduction AND the joint rate schedule — the schedule’s first table is headed "Married filing jointly and qualifying surviving spouse — filing status 2 and 5". The rule branched on `isSingle`, so a survivor was charged as a single filer, costing $610.25 on $50,000 of wages. A case pins that BOTH halves move together, since either alone would be a different defect.

  Tax years 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The reduced $3,100 deduction for a single filer who can be claimed as a dependent, which the calculator does not collect.
  • Itemized deductions (Form IT-196) and the $1,000 dependent exemption.

  NYS DTF — 2025 Form IT-201-I page 11, New York State standard deduction table (read as a page image)

• The line 29 $20,000 pension and annuity income exclusion, AS A PER-PERSON ALLOWANCE THAT IS NOT TRANSFERABLE, applied to NON-FEDERAL pension income. Page 10: "you and your spouse can each subtract up to $20,000 of your own pension and annuity income. However, neither of you can claim any unused part of your spouse’s exclusion." The certifying pair is the same $60,000 twice — all on one person ($994.625) against $30,000 each ($158.00) — identical income, ages and filing status, differing only in whose pension it is; the pooled rule returned $158.00 for both, which is what New York’s own test suite asserted before #416. Also certifies the gate AS A DATE: line 29 asks whether "you were 59½ before January 1, 2025", so the year must OPEN with the taxpayer past it. The discriminating pair is two people BOTH AGED 60 at the end of 2025 whose birth dates are two days apart — 30 June 1965 reaches 59½ on 30 December 2024 and takes the exclusion, 2 July 1965 reaches it on 2 January 2025 and does not — which an `age >= 59` proxy cannot reproduce. And that the allowance is capped at the person’s OWN income rather than lent to a spouse.

  Tax years 2025 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • THE ATTAINMENT YEAR IS NOT PRORATED. Line 29 says "If you became 59½ during 2025, enter only the amount you received after you became 59½, but not more than $20,000." The projection carries no within-year payment dates, so the model can only answer all-or-nothing and withholds the whole exclusion that year. That OVERSTATES tax by at most one year’s partial exclusion, the conservative direction. It is unfixable rather than unfixed: no input distinguishes a January distribution from a December one.
  • THE $20,000 CAP IS "REGARDLESS OF THE SOURCE OF THE INCOME", AND THE MODEL NOW APPLIES IT THAT WAY — ONE allowance per person, shared between a non-federal pension and a non-federal 401(k) or IRA, which is what § 612(c)(3-a) says. #416 shipped TWO, one on each category, and disclosed it; #420 dissolves that, and a case discriminates the readings ($1,595.00 against $499.75 on $30,000 of each). The same change un-pools the TSP side, which was still summing both spouses’ allowances against HOUSEHOLD withdrawals — #416’s defect 4, fixed for the pension there and for the account here. WHAT REMAINS UNMODELLED is any OTHER qualifying source: a household with § 457 or non-TSP IRA income the calculator does not carry separately would exhaust the allowance sooner than this model shows.
  • The beneficiary and decedent allocation rules on page 10, under which a decedent’s $20,000 is shared among beneficiaries.

  NYS DTF — 2025 Form IT-201-I pages 9-10, line 29 pension and annuity income exclusion and the "Married taxpayers" rule (read as page images)

• That FEDERAL retirement income leaves the New York base ENTIRELY — the FERS annuity under IT-201 line 26, which covers a pension from "The United States, its territories, possessions (or political subdivisions thereof), or any agency or instrumentality of the United States (including the military), or the District of Columbia"; a THRIFT SAVINGS PLAN under Tax Law § 612(c)(3)(ii), TO THE EXTENT ITS DISTRIBUTIONS ARE FEDERALLY SOURCED — which for an account with no nongovernmental roll-in is all of them, and which is what the model builds BY DEFAULT — a rolled-in share can now be supplied, and every certified figure here is the share-zero case (see the roll-in exclusion below); and Social Security under line 27. A household living on federal retirement income alone owes New York nothing, and a case pins exactly that on $200,000 of annuity plus $200,000 of TSP plus $40,000 of benefits. What does NOT leave the base is NON-FEDERAL retirement income, in either bucket: a pension from other employment, and a non-federal spouse’s private 401(k) or IRA. Two cases price the distinction on IDENTICAL DOLLARS — $60,000 of FERS annuity costing nothing against $60,000 of other-employer pension costing $1,595, and $60,000 of federal TSP costing nothing against $60,000 from a private account costing $994.625. THE TSP HOLDING TOOK TWO ATTEMPTS. #416’s first draft read the IT-201-I’s general sentence about "a supplemental annuity plan which was funded through a salary reduction program" onto the TSP and certified a $20,000 cap; #416 as shipped kept that cap deliberately and disclosed it; #420 removes it. The Department has held the opposite TWICE, naming the plan: TSB-A-20(6)I (29 Sep 2020) — "lump sum distributions from the TSPs may be subtracted ... Also, distributions from the TSPs to Petitioners other than by lump sum distribution may be subtracted" — and TSB-A-24(14)I (24 Apr 2024), for a "Federal Employees’ Retirement Thrift Savings Plan", resting on 20 NYCRR 112.3(c)(1)(i)(b): retirement benefits paid to an employee of the United States "all or a portion of which are actually contributed by the Federal government". The TSP receives agency automatic and matching contributions. A general instruction cannot carry a certification against specific, current, on-point agency guidance.

  Tax years 2025 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • BOTH ADVISORY OPINIONS BIND THE DEPARTMENT ONLY AS TO THEIR OWN PETITIONERS, so this is the Department’s reading rather than a rule of general application. It is certified as the best available reading — repeated four years apart, naming the plan, and grounded in the regulation — and the alternative was overcharging any New York federal retiree whose TSP distributions exceeded $20,000, which is an input the app collects on every scenario.
  • THE FEDERAL/NON-FEDERAL SPLIT IS PROVENANCE, NOT EMPLOYMENT, AND THAT IS WHAT #420 BUILT. The calculator supports a non-federal spouse whose private 401(k) arrives in the same `tspIncomePersonB` field — the UI only relabels the control, and person B DEFAULTS to non-federal — so exempting the bucket would hand that account an exemption reserved for federal service. A per-person employment flag was tried and REJECTED: when a person dies the engine merges their balance into the survivor, so classification by holder handed the survivor’s status to dollars that never earned it, taking a non-federal spouse’s $60,000 from $994.625 of tax to $0 — WORSE than the rationing it replaced. The projection now carries a non-federal SHARE of each balance, seeded from employment, blended on contribution, conversion and spousal merge, recorded on any annuity contract the premium funds, and split pro-rata at distribution. THE CASES IN THIS PACK CANNOT REACH ANY OF THAT: they build a breakdown directly and never run a projection. Two projection-level tests do — one for the merge, one for annuitization — and each was proved to fail against the defect it guards, the annuity one because a premium removes the dollars from the balance model entirely while the balances still reconcile.
  • THE SHARE ANSWERS A PROPORTIONAL QUESTION AND NOT AN ORDERING ONE. TSB-A-15(6)I splits a mixed account IN PROPORTION — its petitioner was 80% federally derived and 20% rolled in — which a fraction represents exactly. It cannot say WHICH dollars came out first, and must not be used for anything that turns on that; `hasInheritedTSPBalance` records the § 72(t) version of the question, which remains unanswerable from an aggregate balance model.
  • A ROLL-IN IS THE REMAINING GAP, AND IT IS NOW LIVE RATHER THAN LATENT. TSB-A-15(6)I splits an account holding a private-sector rollover: the rolled-in share "will not be eligible for an exclusion under Tax Law § 612(c)(3)(ii), because the contributions to that account were not related to services Petitioner performed as a Federal employee and none were contributed by her Federal employer", and "may be eligible for an exclusion, not in excess of $20,000, under Tax Law § 612(c)(3-a), provided that all the requirements of that section are met" — so $20,000, LESS, OR NOTHING. FERSCalc collects ONE TSP balance and the field guide tells the user to enter it undifferentiated, so an entered figure CAN contain rolled-in dollars, and the model DEFAULTS to treating the account as wholly federally sourced — which UNDERSTATES New York tax to the extent a rolled-in share would have remained taxable. THE ENGINE CAN NOW APPORTION IT: `Employee.tspRolledInShare` seeds the provenance share, and every certified figure here is the share-zero case. WHAT IS DELIBERATELY NOT DONE IS ASKING (#419). Under docs/INPUT_BURDEN_POLICY.md this is a step-3 disposition: the amount is a decades-old rollover most users cannot state precisely, a wrong answer moves the figure with false confidence, and the branch test — run against a correctly solved draw, after #423 — flips funding status or depletion year only at a LARGE rolled-in share on an already-marginal plan; at the 20% of the Department’s own published example neither funding status nor depletion year flips in the cells measured — though the ENDING SURPLUS does move there, so it is not true that the exclusion absorbs the whole effect. So it is surfaced instead: a note where the balance is entered, and a priced comparison copy. The uncertainty register in that policy carries the worked branch-impact table.
  • Interest on U.S. government bonds (line 28), which New York also subtracts. The calculator collects no ordinary interest at all, so there is nothing here to subtract — an earlier draft of this exclusion said the adapter “passes it through as ordinary income”, which was false. What the adapter DOES fold into the wages field is tax-exempt interest, sent through every state’s general-income path as a conservative baseline because the models do not yet distinguish an in-state exemption.
  • THE NON-FEDERAL PENSION SHARE IS IDENTIFIED FROM ONE ADDITIVE FIELD, `nonFederalPensionIncomePersonA/B`, carrying the ledger’s "fixed retirement income". Any other non-federal retirement income the projection folds into the pension category would still be excluded as if federal. No new user input was added; the amounts already existed per person upstream.
  • AND THAT FIELD CARRIES AN AMOUNT, NOT A SOURCE. `FixedRetirementIncome` is an annual figure and a COLA flag — nothing in it says who paid the pension — and the adapter classifies all of it as NON-GOVERNMENT. That is right for the private-employer pension the name describes and WRONG for three kinds IT-201 line 26 also exempts IN FULL: a New York State or local government pension, a MILITARY pension, and a District of Columbia one. A caller putting $60,000 of NYSLRS or military pension in that field is charged $1,595 where New York would charge nothing. THIS IS THE OPPOSITE DIRECTION FROM THE DEFECT CERTIFIED ABOVE, so the change is not strictly better for every meaning the field can bear: before #416 the category was excluded whole, which was right for those three and wrong for a private pension. What this claim certifies is therefore the TREATMENT OF A NON-GOVERNMENT PENSION, not a general rule for anything a caller may put in the field. #418 must collect the SOURCE alongside the amount.
  • AND TODAY THAT FIELD IS ALWAYS ZERO FOR A SCENARIO BUILT BY THE WEB APP, which bounds the PENSION half of this claim — though not the TSP half, which is reachable on every scenario. `RetirementScenario` supports `fixedRetirementIncome` and the engine threads it through, so the pension treatment is live for any caller that supplies it — but the web app’s only `Employee` builders are `buildEmployee`, which sets it to `undefined` unconditionally (src/lib/utils/build-projection-inputs.ts:199), and `buildDummyPersonB`, which never sets it at all — and `PersonDraft` carries no field it could be built from. Collecting it is #418.

  NYS DTF — Advisory Opinion TSB-A-15(6)I: a TSP holding a private-sector rollover is SPLIT — the rolled-in share is not eligible for § 612(c)(3)(ii) and may get at most $20,000 under § 612(c)(3-a), and only if that section’s requirements are met · NYS DTF — Advisory Opinion TSB-A-24(14)I: distributions from a Federal Employees’ Retirement Thrift Savings Plan qualify for the Tax Law § 612(c)(3)(ii) subtraction. EVIDENCE FOR THE TREATMENT DEFERRED TO #420, not for the rationing this rule still applies · NYS DTF — Advisory Opinion TSB-A-20(6)I: lump-sum and non-lump-sum TSP distributions qualify for the same subtraction, as does a TSP-funded IRA but only "to the extent that the distributions represent a return of the amount rolled over" · NYS DTF — 2025 Form IT-201-I page 9, line 26 pensions of New York State and local governments and the federal government (read as a page image)

• The tax table benefit recapture (pages 34-39), absent from the rule before #416. Above $107,650 of NEW YORK adjusted gross income the rate schedule is not the end of the computation: sixteen worksheets add back the benefit of the lower bands. Certified as starting strictly ABOVE $107,650 — and, in the same case, that the worksheet’s own four-decimal rounding of the phase fraction means the first two dollars past the floor add NOTHING, the first that bites being $107,653 where 3/50,000 rounds up to 0.0001. An earlier draft asserted that $107,651 started the phase-in; it does not, and that case would have passed against a rule with no recapture at all; as completing at $157,650, where the first-tier worksheets flatten the whole taxable income to one rate (6% single, 5.5% joint) — a clean flat-rate landing that verifies the phase arithmetic rather than snapshotting it; as adding the printed Recapture Base in full plus the printed Incremental Benefit phased over $50,000 in the later tiers; and as abandoning the schedule entirely above $25,000,000 for a flat 10.9%. ALSO CERTIFIED IS WHICH AGI IT READS: the NEW YORK figure, not the federal one. Federal AGI carries the FERS annuity and the taxable part of Social Security, both of which New York subtracts, so a household with $100,000 of wages, $300,000 of annuity and $40,000 of benefits has a federal AGI of $434,000 — the annuity in full plus at most 85% of the benefits under IRC § 86 — and a New York AGI of $100,000. A rule keying on the federal field would have charged $5,520 instead of $4,951.75.

  Tax years 2025 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The head-of-household worksheets 12-16, unreachable because the status is not modelled.
  • THE RATE SCHEDULE’S OWN 10.9% BAND IS STRUCTURALLY UNREACHABLE, and is carried for completeness rather than certified. Taxable income is always BELOW New York AGI (it is New York AGI less the standard deduction), so taxable income above $25,000,000 implies New York AGI above $25,000,000 — at which point worksheets 6 and 11 discard the schedule entirely and tax the whole amount at 10.9% anyway. A mutation deleting that band SURVIVED the pack, correctly, and was narrowed to the 10.3% band, which IS reachable inside worksheet 10 and now has a case.
  • The worksheets’ own whole-dollar convention. Each says "Use whole dollars only" and this model compares raw Decimals; the phase fraction IS rounded to four decimal places as the worksheet directs, but the inputs to it are not rounded to dollars. Engine-wide, tracked as #411.

  NYS DTF — 2025 Form IT-201-I pages 34-39, tax computation worksheets 1-16 (read as page images)

• New York Paid Family Leave, the SECOND state employee payroll contribution FERSCalc models and the first whose cap is on the CONTRIBUTION rather than on the wage (#429). Certified in five parts. FIRST, that it is not an IT-201 liability: it is returned by `applyPayrollWithholding` and never by `applyTax`, so wages cost what wages cost while the contribution is a separate non-zero figure. It does not reduce New York taxable income either — the Department of Taxation and Finance says premiums are deducted "from your after-tax wages" and reported "on Form W-2 in Box 14 as state disability insurance taxes withheld" — which is what keeps the charge additive. SECOND, the rate and the maximum for BOTH published years: 0.432% capped at $411.91 for 2026, 0.388% capped at $354.53 for 2025, with the latest figures held beyond. THIRD, that the CAP IS ON THE CONTRIBUTION: $120,000 of wages owes $411.91 rather than $518.40, ten times that wage owes the same $411.91, and the two branches meet exactly at the $95,349.54 cap wage — a case that separates only if the rate and the maximum are both right. FOURTH, that the maximum is PER PERSON, since it applies "for each covered employer": two earners at $60,000 owe $259.20 each, so $518.40, which EXCEEDS the single maximum — a rule summing the wages and capping once would return $411.91 and be wrong by $106.49. FIFTH, that the base is WAGES and that federal employment is outside the programme, the latter DERIVED rather than quoted: New York publishes no federal-exemption sentence, and Workers’ Compensation Law § 201(4) defines a covered employer as a "person, partnership, association, corporation, legal representative of a deceased employer, or the receiver or trustee" of one, which the federal government is not.

  Tax years 2025, 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • NEW YORK’S DISABILITY-BENEFIT DEDUCTION, which is real and NOT modelled — an employer may deduct "one-half of one percent of their wages, but no more than sixty cents a week" (Workers’ Compensation Board). It is excluded on MATERIALITY rather than on principle: sixty cents a week is at most about $31 a year. That is why this state’s declaration reads `partial` rather than `full`, and why the public guide shows it as partly modeled.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. The maximum applies "for each covered employer", so someone with two jobs can pay up to two caps. The model carries one wage figure per person and applies ONE cap to it, which UNDERSTATES the contribution for that household. Nothing collected distinguishes one employer from two.
  • WHETHER A NEW YORK PUBLIC EMPLOYER OPTED IN. § 201(4) excludes "the state, a municipal corporation, local governmental agency, other political subdivisions or public authority" from the definition of employer, and such an employer may take on coverage voluntarily — many have, by collective bargaining. So a state or municipal employee may or may not be covered and nothing collected says which. The default CHARGES, and the exempt branch prices the alternative rather than asking.
  • The waiver available to an employee who will not work long enough to become eligible (fewer than 20 hours a week and under 175 days a year, or 20+ hours and under 26 consecutive weeks). It is optional and employee-specific, and nothing collected reaches it.
  • RATE YEARS BEYOND 2026. Both the rate and the maximum move most years — the cap tracks the statewide average weekly wage — so later rows hold 2026’s figures and a long projection asserts only "at today’s rate".

  NYS Paid Family Leave — 2026 updates: "employees will contribute 0.432 % of their gross wages per pay period" and "The maximum employee contribution for each covered employer in 2026 is $411.91" · NYS Paid Family Leave — Cost and Deductions: the 2025 rate of 0.388% and its $354.53 maximum, and that "Paid Family Leave is not optional for eligible employees" · NYS Department of Taxation and Finance — Paid Family Leave: "Your employer will deduct premiums for the Paid Family Leave program from your after-tax wages" · Workers’ Compensation Law § 201(4) — the definition of "employer", which reaches private forms and excludes New York’s own public employers

• Paid Leave Oregon’s employee contribution, the THIRD state payroll charge FERSCalc models and the third distinct cap shape (#429). Certified in five parts. FIRST, that it is not an Oregon income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the tax on a covered earner’s wages is not the contribution figure. SECOND, the employee share: "Employees pay 60% of the total 1% contribution rates", so 0.6% — stored as the PRODUCT of the published total rate and the published share rather than as a rate Oregon prints on its own, because a year that moves the total moves this and the arithmetic must be redone. THIRD, that the maximum is a WAGE and not a contribution: $184,500 for 2026 and $176,100 for 2025, so $500,000 of wages owes $1,107.00 and the branches meet exactly at the published figure. THAT MAXIMUM IS NOT OREGON’S OWN CONSTANT and is no longer certified as one — the state defines it by reference to the Social Security wage index, so the rule reads the engine’s published SSA table, which returns 2025 and 2026 exactly and PROJECTS beyond them. The first version of this claim certified the cap as held flat at $184,500 forever, which understated it, and the charge, for every year after 2026; the corrected case asserts the projected figure and labels it an estimate. THAT IS THE OPPOSITE OF NEW YORK, which caps the contribution at $411.91 — and Oregon’s correct behaviour is precisely the misreading New York’s pack carries a mutation against, which is why no rule may borrow another’s shape. FOURTH, that the maximum is applied SEPARATELY TO EACH MODELLED EARNER and never to a household total: two earners at $150,000 owe $900 each, where a rule summing the wages first would cap $300,000 at $184,500 and return $1,107, wrong by $693. THAT IS THE MODEL’S RULE, NOT OREGON’S — the statute applies the maximum "separately to each employer an individual works for", and the cases here cannot reach that, since nothing collected distinguishes one employer from two. What is certified is that the household is not collapsed; the per-employer reading is recorded as an exclusion below. FIFTH, that the base is WAGES, and that federal employment is outside the programme — which Oregon states OUTRIGHT where the other two did not: ORS 657B.010 provides that "‘Employer’ does not include the federal government or a tribal government."

  Tax years 2025, 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • OREGON’S INCOME TAX, which this pack does not touch at all. Its brackets, its federal-tax subtraction and its unmodelled pre-October-1991 service subtraction remain uncertified and on the `state-income-tax-baseline` provenance entry. This pack’s evidence is recorded against the cross-cutting `state-payroll-contributions` entry precisely so that nothing here reads as certifying the state.
  • WHO ACTUALLY PAYS THE EMPLOYEE SHARE. ORS 657B.150(5) provides that "an employer may elect to pay the required employee contributions, in whole or in part, as an employer-offered benefit", which is open to ANY employer rather than only one running an approved equivalent plan — and such a plan may separately "assume all or part of the costs". So a covered employee may pay all of this charge, some of it, or none. Not derivable from anything collected; the default charges the whole share and a priced branch shows the full-pickup endpoint, with partial pickup landing between. That endpoint is not free: per IRS Rev. Rul. 2025-4, an employer's voluntary payment of an employee's own required contribution is additional taxable W-2 compensation, which the branch's comparison does not net out.
  • TRIBAL-GOVERNMENT EMPLOYMENT, excluded by the same sentence of ORS 657B.010 as federal employment, and able to elect coverage. The engine cannot distinguish it from other non-federal work.
  • THE PER-EMPLOYER MAXIMUM FOR A TWO-JOB EARNER. The maximum wage applies separately to each employer, so someone with two jobs can be charged on more than $184,500 in total. The model carries one wage figure per person and applies one maximum, which UNDERSTATES the contribution for that household.
  • SELF-EMPLOYMENT, which participates only by election, and which the model does not represent as an income type in any state.
  • DEDUCTIBILITY ON A FEDERAL RETURN. Rev. Rul. 2025-4 holds that mandatory employee contributions of this kind "are employee payments of State income tax", deductible under § 164(a)(3) — but only by a taxpayer who ITEMIZES, and subject to the SALT limitation. FERSCalc models the standard deduction and no itemized deductions in any jurisdiction, so a real itemizing household would pay less federal tax than this model shows. The gap reaches California and New York identically.
  • RATE YEARS BEYOND 2026. Oregon sets the CONTRIBUTION RATE by 15 November each year and the total is statutorily capped at 1%, so later rows hold the 0.6% employee share and a long projection asserts only "at today’s rate". The MAXIMUM WAGE is different in kind: it follows the engine’s OASDI projection rather than being held, because Oregon does not set that figure — and a projected base is an estimate, disclosed as one on the FICA path already.

  Paid Leave Oregon — Contributions: the 1% total rate for 2025 and 2026, "Employees pay 60% of the total 1% contribution rates", and maximum wages of $176,100 (2025) and $184,500 (2026) · ORS 657B.010 — "‘Employer’ does not include the federal government or a tribal government" · Paid Leave Oregon — Equivalent Plan Guidebook: an approved-plan employer "may assume all or part of the costs" and "may not withhold more than the amount employees would have paid"

• Minnesota Paid Leave’s employee premium, the FOURTH state payroll charge FERSCalc models and the first with a START YEAR (#429). Certified in five parts. FIRST, that it is not a Minnesota income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that PREMIUMS BEGAN ON 1 JANUARY 2026 and nothing is charged before — an ABSENT programme rather than a zero rate, with the boundary pinned on both sides. THIRD, that what is modelled is a CEILING and not a rate: Minnesota sets a 0.88% total premium for 2026 and 2027 and provides that "employers can collect up to 0.44% from employees, or employers can choose to cover more", and may "not collect more than 50%" — so 0.44% is the most an employee can be charged, and charging it is the conservative assumption. FOURTH, that the WAGE is capped at "the Old-Age, Survivors, and Disability Insurance (OASDI) limit" in Minnesota’s own words, so $184,500 of 2026 wages owes $811.80 however large the wage, and the rule reads the engine’s published SSA table rather than keeping a copy of a figure Minnesota does not set. FIFTH, that the base is WAGES and that federal employment is outside the programme — the STRONGEST of the four states on this point, since Minnesota says both that "Federal entities are exempt from Paid Leave" and that federal agencies "are not covered by Paid Leave and cannot opt in".

  Tax years 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MINNESOTA’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • HOW MUCH THE EMPLOYER ACTUALLY COLLECTS. "Employers may choose to cover up to 100% of the total premium for employees, but may not collect more than 50% from employees." The model charges the 0.44% ceiling, so a household whose employer absorbs any of it is OVERSTATED. Not derivable from anything collected; a priced branch shows the full-pickup endpoint and partial pickup lands between. That endpoint is not free: Minnesota's own guidance says an employer "pick-up" beyond the required share is taxable income to the employee under IRS Rev. Rul. 2025-4, which the branch's comparison does not net out.
  • THE SMALL-EMPLOYER RATE. "The Paid Leave small employer rate is 0.66%", which halves to 0.33% on the employee side rather than 0.44%. Eligibility turns on the employer’s headcount (30 or fewer) and its average wage against the statewide average — neither of which this model represents — so such a household is OVERSTATED by up to a quarter of the charge.
  • TRIBAL NATIONS, which "are not automatically covered by Paid Leave, but they can opt in", and self-employed individuals and independent contractors on the same terms. The engine cannot distinguish any of them from other non-federal work.
  • THE PROJECTED OASDI BASE beyond 2026. The cap follows SSA’s published base exactly for 2025 and 2026 and the engine’s growth assumption after that, which is an estimate and is disclosed as one on the FICA path already.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. Paid Leave caps the premium at the OASDI limit for each covered employer, so someone with two employers in one year can have more than the annual maximum withheld in total. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for that household — the same gap New York’s and Oregon’s own claims disclose. New Jersey’s Form NJ-2450 reconciles the excess automatically on the tax return; Connecticut and Maine instead let a worker file a manual refund claim with the state authority. Neither route is something this model applies, and Minnesota’s own remedy, if any, is not modelled.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction. See the note on `StatePayrollWithholding` for why Rev. Rul. 2025-4 is not a single answer across these programmes.

  Minnesota Paid Leave — Premium rate and contributions: "For 2026 and 2027, the Paid Leave premium rate is 0.88%", "The Paid Leave small employer rate is 0.66%", and "employers can collect up to 0.44% from employees, or employers can choose to cover more" · Minnesota Paid Leave — Common questions: "Premiums are capped at the Old-Age, Survivors, and Disability Insurance (OASDI) limit"; "Federal entities are exempt from Paid Leave"; federal agencies "are not covered by Paid Leave and cannot opt in"

• New Jersey’s FOUR worker contributions, the fifth state payroll charge FERSCalc models and much the most complex (#429). Certified in five parts. FIRST, that none of them is a New Jersey income tax: they are returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, all four rates applying together below both bases — Unemployment Insurance 0.3825%, Workforce Development 0.0425%, Temporary Disability 0.19% and Family Leave Insurance 0.23% for 2026, so $40,000 of wages owes $338.00. THIRD, that THE TWO BASES ARE DIFFERENT and both are New Jersey’s own rather than the OASDI figure Oregon and Minnesota use: unemployment and workforce development stop at $44,800 for 2026 while disability and family leave run to $171,100, so $100,000 of wages owes $610.40 — a figure no single-base rule could produce from either base — and above both the charge holds at $909.02. FOURTH, that the 2025 rates and bases REPRODUCE NEW JERSEY’S OWN PUBLISHED CEILINGS: Form NJ-2450 names $184.02, $380.42 and $545.82 for excess withholding, each exactly a base times a rate, so the rates and the bases confirm each other rather than resting on one page. FIFTH, that both bases are applied to each earner’s own wages and never to a household total — two earners at $40,000 owe $676.00, where a collapse returns $526.40 and UNDERSTATES them — that the base is WAGES, and that federal employment is outside all four.

  Tax years 2025, 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • NEW JERSEY’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry.
  • THE ROUNDING REGIME, and the half-cent it produces. $43,300 x 0.425% is $184.025 exactly and New Jersey publishes the ceiling as $184.02. This model computes the product and never rounds, in any state, so the certified 2025 total is $1,110.265 against a filed $1,110.26. What is certified is the rates and the bases; the rounding is an exclusion rather than a disagreement, as it is for California’s Form 540.
  • A GOVERNMENTAL REIMBURSABLE EMPLOYER’S LOWER UNEMPLOYMENT RATE — 0.0825% rather than 0.3825% on New Jersey’s own table. Nothing collected says who the employer is, so the standard rate is charged, which OVERSTATES such a worker. It is disclosed rather than branched, because it is a lower rate and not an exemption: the priced branch’s all-or-nothing shape would misdescribe it.
  • PRIVATE DISABILITY PLANS. "A subject employer may opt for a private plan for the payment of temporary disability benefits instead of the mandated State Plan", and certain government entities are excluded from automatic disability coverage. Neither is derivable, so the default charges the state-plan amount and the priced branch offers the alternative.
  • THE PER-EMPLOYER CEILINGS FOR A TWO-JOB EARNER. Each employer applies each base afresh, and the model carries one wage figure per person, so such a worker is understated here. New Jersey lets them RECLAIM the excess on Form NJ-2450 with their return — an automatic reconciliation on the tax form itself, which California, New York, Oregon and Minnesota do not appear to offer. Connecticut and Maine offer a different remedy instead: a manual refund claim filed with the state authority, not a tax-form reconciliation — see their own claims.
  • DEDUCTIBILITY ON A FEDERAL RETURN by an itemizing taxpayer. FERSCalc models the standard deduction and no itemized deductions in any jurisdiction; see the note on `StatePayrollWithholding`.

  NJ Department of Labour, Division of Employer Accounts — Rate information: the 2026 and 2025 worker rates for U.I., D.I., W.F./S.W.F. and F.L.I, and both taxable wage bases ($44,800 / $171,100 for 2026; $43,300 / $165,400 for 2025) · NJ Division of Employer Accounts, Employer Handbook — "A New Jersey employer subject to the Unemployment Compensation Law is also subject to the Temporary Disability Benefits Law", and the private-plan option · NJ Division of Taxation, Form NJ-2450 — the 2025 excess-contribution limits of $184.02, $380.42 and $545.82, which cross-check the rates against the bases

• Alaska’s employee UNEMPLOYMENT INSURANCE contribution, the seventh state payroll charge FERSCalc models and the FIRST in a jurisdiction with no income tax at all (#429). Certified in four parts. FIRST, and the reason a no-income-tax state gets a pack: that the charge exists and is levied while `applyTax` returns zero and always will — "$0 state tax" is true and incomplete, which is the clearest possible demonstration of why a payroll contribution needed a category separate from the income-tax rules. SECOND, the rate: "The 2026 Employee Rate is 0.50%", unchanged in both published years. THIRD, that the WAGE is capped at Alaska’s OWN base — "The 2026 Taxable Wage Base is $54,200.00", against $51,700 for 2025 — a figure roughly a third of the OASDI base Oregon and Minnesota read, so this rule keeps its own table; the two branches meet exactly at the published figure, and because the rate held between the two years, the 2025 case tests the BASE alone. FOURTH, that the base is per modelled earner rather than per household, and is WAGES. Federal employment is outside it, DERIVED rather than quoted: Alaska publishes no exemption line and its handbook’s excluded-employment list never reaches the question, but Alaska "administers the Unemployment Compensation for Federal Employees (UCFE) Program for the federal government under statutory authority of Title 5 Chapter 85 United States code", so federal employees draw on that programme rather than Alaska’s own.

  Tax years 2025, 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • ALASKA HAS NO INCOME TAX, so this pack certifies no income-tax behaviour whatsoever — there is none. It is recorded against the cross-cutting `state-payroll-contributions` provenance entry rather than a per-jurisdiction one, which for Alaska also keeps its `no_income_tax` coverage classification untouched.
  • NO EXEMPT BRANCH IS OFFERED, as for Pennsylvania. The classes Alaska’s handbook excludes are ownership roles — sole proprietors, partners, LLC members — rather than employment an ordinary wage-earning spouse could be in without knowing, so there is no unidentifiable class to price.
  • EMPLOYERS WHO VOLUNTARILY ELECT COVERAGE for otherwise-excluded employment, which would make an excluded person covered. Nothing collected could reveal it.
  • ALASKA’S CAPITAL-GAINS AND OTHER NON-INCOME TAXES, which are outside both this pack and the model.
  • DEDUCTIBILITY ON A FEDERAL RETURN by an itemizing taxpayer; see the note on `StatePayrollWithholding`.

  Alaska DOLWD — 2026 Unemployment Insurance Tax Rates: "The 2026 Employee Rate is 0.50%" and "The 2026 Taxable Wage Base is $54,200.00" · Alaska DOLWD — 2025 Unemployment Insurance Tax Rates: the 2025 employee rate of 0.50% and taxable wage base of $51,700.00 · Alaska DOLWD — Federal workers and Unemployment Insurance: Alaska "administers the Unemployment Compensation for Federal Employees (UCFE) Program for the federal government under statutory authority of Title 5 Chapter 85 United States code"

• Washington’s TWO employee payroll contributions, the ninth state charge FERSCalc models, and the state that showed #429’s survey had missed a THIRD programme category (#429). Certified in five parts. FIRST, that both are charged although Washington levies no income tax at all — `applyTax` returns zero for a covered and an exempt earner alike, and the contributions are real. SECOND, the PAID LEAVE rate as the employee share of a total premium, where BOTH parts move by year: "The premium rate will be 1.13%" with "Employees will pay 71.43%" for 2026, against "The rate for 2025 is 0.92%" with a 71.52% share — so a case at the same wage in the two years separates them, and a rule that changed only the total, or only the split, lands elsewhere. THIRD, WA CARES at "0.58 percent of each employee’s gross wages— there is no Social Security cap", "paid by the employee" in full. FOURTH, and the case no single-rate rule can produce: ABOVE THE CAP THE TWO DIVERGE. At $500,000 of 2026 wages the paid-leave premium stops at $1,489.2084 while WA Cares reaches $2,900, so $4,389.2084 — where a wholly capped rule would stop below $2,000 and a wholly uncapped one would exceed $6,900. Doubling the wage moves only the WA Cares half. FIFTH, that both are per modelled earner, on WAGES, and outside federal employment.

  Tax years 2025, 2026 · 9 cases · re-check due 2026-12-01

  Not covered by this claim:

  • WASHINGTON HAS NO INCOME TAX, so this pack certifies no income-tax behaviour — there is none. It is recorded against the cross-cutting `state-payroll-contributions` provenance entry, which also leaves Washington’s `no_income_tax` coverage classification untouched.
  • WHO ACTUALLY PAYS. Employers may "pay some or all of the premium on their employees’ behalf" for Paid Leave, and "must collect WA Cares premiums or pay employee premiums on their behalf". The model charges the full employee share, which OVERSTATES a household whose employer picks any of it up; the priced branch shows the endpoint where none is withheld.
  • WA CARES EMPLOYEE EXEMPTIONS, which are real and individual: "Some employees may be exempt from contributing WA Cares premiums. It is the employee’s responsibility to formally notify you of their exemption." The exempt assertion clears BOTH programmes together, so it prices the full-relief endpoint rather than an exemption from WA Cares alone — which would leave the paid-leave premium standing. Stated because the branch’s all-or-nothing shape cannot express it.
  • VOLUNTARY EMPLOYER PLANS for Paid Leave, under which contributions may differ.
  • THE PROJECTED SOCIAL SECURITY CAP beyond the last published base, which is an estimate and disclosed as one.
  • DEDUCTIBILITY ON A FEDERAL RETURN by an itemizing taxpayer; see the note on `StatePayrollWithholding`.

  WA ESD — Paid Family & Medical Leave premium rate increases to 1.13% in 2026: "The premium rate will be 1.13%", "Employees will pay 71.43%", "The rate for 2025 is 0.92%" · Paid Leave Washington — Your role and responsibilities: the Social Security cap on paid-leave premiums; "The WA Cares premium rate is 0.58 percent of each employee’s gross wages— there is no Social Security cap"; "The WA Cares premium is paid by the employee" · Paid Leave Washington — Updates: "The Social Security cap for the 2026 calendar year will increase to $184,500", and the 2025 employee share of 71.52%

• Colorado FAMLI’s employee premium, the EIGHTH state payroll charge FERSCalc models, and the first with a rate that has already changed once (#429). Certified in seven parts. FIRST, that it is not a Colorado income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that PREMIUMS BEGAN 1 JANUARY 2023 and nothing is charged before — an ABSENT programme rather than a zero rate, pinned on both sides of the boundary. THIRD, that the RATE CHANGED for 2026: 0.45% of a 0.9% total premium for 2023-2025 dropped to 0.44% of 0.88%, both years splitting the total exactly in half — the statutory employee ceiling, and the ceiling this model charges as the conservative assumption. FOURTH, that federal employment is outside it BY STATUTE: the FAMLI Act’s own definitions say "‘Employer’ does not include the federal government" (C.R.S. § 8-13.3-503(8)(c)). FIFTH, that the WAGE is capped at the OASDI limit — $184,500 of 2026 wages owes $811.80 however large the wage — read from the engine’s published SSA table rather than a Colorado-specific copy, as Minnesota’s and Oregon’s do. SIXTH, that the cap and base apply per modelled earner, never to a household total. SEVENTH, that an employer may pay more of the premium than the modelled ceiling, which is priced as a branch rather than derived.

  Tax years 2023, 2024, 2025, 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • COLORADO’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • HOW MUCH THE EMPLOYER ACTUALLY COLLECTS. Colorado employers "may also choose to pay the full amount if they would like to offer this as an added perk for their employees." The model charges the 0.44%/0.45% ceiling, so a household whose employer absorbs any of it is OVERSTATED. Not derivable from anything collected; a priced branch shows the full-pickup endpoint and partial pickup lands between. That endpoint is not free: FAMLI's own IRS Tax Guidance says an employer "pick-up" of the employee's share becomes taxable compensation, which the branch's comparison does not net out.
  • LOCAL GOVERNMENT OPT-OUT. A local government employer may have opted out of FAMLI entirely, in which case its employees owe nothing unless they voluntarily elect in for a minimum three-year term. The engine cannot distinguish this employment from any other non-federal work.
  • THE RATE BEYOND 2026. The FAMLI Division recalculates the premium annually, capped by statute at 1.2% of wages; this table’s 2026 figure is the best available estimate for a later year, not a guarantee.
  • THE PROJECTED OASDI BASE beyond 2026. The cap follows SSA’s published base exactly for 2025 and 2026 and the engine’s growth assumption after that, which is an estimate and is disclosed as one on the FICA path already.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. FAMLI caps the premium at the OASDI limit for each covered employer, so someone with two employers in one year can have more than the annual maximum withheld in total. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for that household — the same gap New York’s and Oregon’s own claims disclose. New Jersey’s Form NJ-2450 reconciles the excess automatically on the tax return; Connecticut and Maine instead let a worker file a manual refund claim with the state authority. Neither route is something this model applies, and Colorado’s own remedy, if any, is not modelled.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  FAMLI Premium and Benefits Calculator — 2026 total premium 0.88%, split 0.44%/0.44% employer/employee, capped at the Social Security wage base · "Colorado FAMLI Rates Remain the Same for 2025" — the 2023-2025 rate of 0.9% total, split 0.45%/0.45% · FAMLI — Employers: "premium payments for FAMLI will decrease to 0.88 percent of a worker’s wages in 2026", and the statutory 1.2% ceiling · FAMLI news — "Employees of the federal government cannot access the state benefit" and see no FAMLI deduction · C.R.S. § 8-13.3-503(8)(c) — "‘Employer’ does not include the federal government"

• Connecticut Paid Leave’s employee contribution, the NINTH state payroll charge FERSCalc models, and the first funded by the employee ALONE (#429). Certified in six parts. FIRST, that it is not a Connecticut income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that federal employment is outside it BY STATUTE: "‘Employer’ does not mean the federal government" (Conn. Gen. Stat. § 31-49e(8)). THIRD, that the RATE IS A FLAT 0.5%, held since withholding began in 2021 and confirmed unchanged for 2026, so no year table is needed. FOURTH, that the WAGE is capped at the OASDI limit — employees "contribute until they reach the Social Security contribution limit" — applied per modelled earner, never to a household total. FIFTH, that there is NO employer-pays-more axis to price: "the funding to support CT Paid Leave comes from employee payroll deductions; there is no employer match" — unlike Colorado’s or Minnesota’s `exemptBranch`. SIXTH, that a DIFFERENT axis survives and earns its own `exemptBranch`: the state, a municipality, or a local or regional board of education is not an "employer" under the statute except as to its "covered public employees" (§ 31-49e(5)) — and state and municipal employment default OPPOSITE ways. A state employee is covered unless their bargaining unit has not negotiated inclusion, so most state employees ARE covered by default; a municipal or school employee is covered only once their employer has negotiated inclusion for at least one bargaining unit — Maryland’s and New Jersey’s shape, not Pennsylvania’s.

  Tax years 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • CONNECTICUT’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • PRIVATE PLAN SUBSTITUTION. An employer "may apply to CT Paid Leave to offer a private plan offering the same or better benefits as the CT Paid Leave public plan," and Conn. Gen. Stat. § 31-49o requires such a plan to "cost employees no more than the premium charged to employees under the state program" — permitting LESS, not just the same. The model always charges the full 0.5%, which is the conservative ceiling; a private-plan employee paying less than that is OVERSTATED.
  • STATE EMPLOYEES IN A NON-PARTICIPATING UNION, AND MUNICIPAL OR SCHOOL EMPLOYEES OF A NON-PARTICIPATING EMPLOYER. The statute’s own carve-out covers the state, a municipality, or a local or regional board of education "except as to its covered public employees", and state and municipal employment default OPPOSITE ways: most state employees ARE covered by default, while a municipal or school employee is covered only once their employer has negotiated inclusion — a fact about the NON-FEDERAL earner’s specific employer and union status this model does not collect, so it still charges the contribution by default. The `exemptBranch` prices what a non-federal spouse in either uncovered class would be worth instead, but FERSCalc does not ask which applies and does not adjust the default figure.
  • THE PROJECTED OASDI BASE beyond 2026. The cap follows SSA’s published base exactly for 2025 and 2026 and the engine’s growth assumption after that, which is an estimate and is disclosed as one on the FICA path already.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. CT Paid Leave’s own guidance says an employee "contribute[s] until they reach the Social Security contribution limit" WITH EACH EMPLOYER, so someone with two employers in one year can have more than the annual cap withheld in total. Connecticut lets a worker claim a refund from the CT Paid Leave Authority with proof of year-end earnings — manual, not automatic like New Jersey’s Form NJ-2450 reconciliation, and not something the model applies. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for that household. Nothing collected distinguishes one employer from two — the same gap New York’s and Oregon’s own claims disclose.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  CT Paid Leave — Contributions: "maintain the contribution rate at 0.5% for 2026"; "no employer match"; employees "may not opt-out"; contribute "until they reach the Social Security contribution limit" · Conn. Gen. Stat. § 31-49e(8) — "‘Employer’ does not mean the federal government, the state or a municipality, a local or regional board of education or a nonpublic elementary or secondary school, except that the state, a municipal employer or local or regional board of education is an employer with respect to each of its covered public employees" · CT Paid Leave — Coverage and Eligibility: municipalities not covered "unless their unionized employees collectively bargain to participate"; public school operators, as to certified employees, on the same terms

• Rhode Island TDI’s employee contribution, the TENTH state payroll charge FERSCalc models, and the SECOND (with Alaska) whose taxable wage base is its OWN rather than the OASDI figure (#429). Certified in seven parts. FIRST, that it is not a Rhode Island income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that the RATE FELL AND THE BASE ROSE for 2026: "the TDI contribution rate will be set at 1.1 percent for calendar year 2026, down from 1.3 percent in 2025", on a base that grew from $89,200 to $100,000 — Alaska’s shape, where a wage above the base tests the base and one below it tests the rate. THIRD, that federal employment is outside it BY STATUTE: TDI’s own definitions section adopts chapter 28-42’s definition of "employment", which exempts "service performed in the employ of... the United States government... or of an instrumentality of the United States" (R.I. Gen. Laws § 28-42-8(3)). FOURTH, that the base is Rhode Island’s OWN figure and not the $184,500 OASDI base — a wage between the two (proven at $150,000, above RI’s base and below the OASDI one) pins the distinction. FIFTH, that the cap is applied per modelled earner, never to a household total. SIXTH, that there is NO employer-pays-more axis to price: the programme’s own page says TDI/TCI "are financed entirely by payroll deductions," and unlike New Jersey, California or Connecticut, no section of R.I. Gen. Laws Title 28 authorizes a private plan or an employer contribution — matching Pennsylvania’s and Alaska’s precedent on THAT axis. SEVENTH, that a DIFFERENT axis survives and earns its own `exemptBranch`: "governmental entities... shall not be deemed to be employing units" subject to TDI by default (R.I. Gen. Laws § 28-39-3), unless the entity elects in under § 28-39-3.1 ("Employees of certain governmental entities eligible by election"), or a unionized state worker’s bargaining unit elects coverage — Maryland’s and New Jersey’s shape, not Pennsylvania’s.

  Tax years 2025, 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • RHODE ISLAND’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • STATE AND MUNICIPAL GOVERNMENT EMPLOYEES outside an election under § 28-39-3.1 or a collective-bargaining opt-in. FERSCalc’s payroll fields distinguish only federal and non-federal employment, so it still charges the contribution by default for a non-federal spouse employed by Rhode Island state or municipal government. The `exemptBranch` prices what that employment would be worth instead, but FERSCalc does not ask which applies and does not adjust the default figure.
  • THE RATE AND BASE BEYOND 2026. The DLT recalculates both annually; this table’s 2026 figures are the best available estimate for a later year, not a guarantee.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  RI DLT — 2026 Tax Rates for Unemployment Insurance and Temporary Disability Insurance: "the TDI contribution rate will be set at 1.1 percent for calendar year 2026, down from 1.3 percent in 2025", taxable wage base $100,000 · RI DLT quick-reference table — 2025 TDI/TCI taxable wage base $89,200 · RI DLT — TDI/TCI For Employers: TDI/TCI "are financed entirely by payroll deductions" · R.I. Gen. Laws § 28-42-8(3) — the federal-government exemption from "employment" · R.I. Gen. Laws § 28-39-3 — "Governmental entities as defined in § 28-42-3 shall not be deemed to be employing units subject to chapters 39 — 41 of this title" · R.I. Gen. Laws § 28-39-3.1 — "Employees of certain governmental entities eligible by election"; a political subdivision "may become subject to those chapters by election"

• Maine PFML’s employee premium, the ELEVENTH state payroll charge FERSCalc models, and the SECOND (with Minnesota) with a START YEAR inside this model’s reach (#429). Certified in six parts. FIRST, that it is not a Maine income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that WITHHOLDING BEGAN 1 JANUARY 2025 and nothing is charged before — an ABSENT programme rather than a zero rate, with the boundary pinned on both sides. THIRD, that federal employment is outside it BY STATUTE, stated more plainly than most: "wages do not include wages earned from federal employment." FOURTH, that the employee CEILING IS 0.5% REGARDLESS OF EMPLOYER SIZE — an employer with 15 or more employees may deduct up to half of its 1% joint rate (0.5%), and a smaller employer may deduct the entirety of its 0.5% joint rate (also 0.5%) — so, unlike Washington’s two-charge shape, no employer-size branch is needed to reach the ceiling. FIFTH, that the WAGE is capped at the OASDI limit — "the premium amount will be calculated to exclude amounts above the annual base limit set by the U.S. Social Security Administration" — applied per modelled earner, never to a household total. SIXTH, that an employer may deduct less than the ceiling, which is priced as a branch rather than derived.

  Tax years 2025, 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MAINE’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • HOW MUCH THE EMPLOYER ACTUALLY DEDUCTS. The law says an employer "may" deduct up to the ceiling, not "must" — an employer may cover some or all of the employee share instead. The model charges the 0.5% ceiling, so a household whose employer absorbs any of it is OVERSTATED. Not derivable from anything collected; a priced branch shows the full-pickup endpoint and partial pickup lands between. That endpoint is not free: per IRS Rev. Rul. 2025-4, an employer's voluntary payment of an employee's own required contribution is additional taxable W-2 compensation, which the branch's comparison does not net out.
  • PUBLIC-SECTOR COLLECTIVE BARGAINING CARVE-OUTS. Public employers and employees subject to a collective bargaining agreement in effect on 25 October 2023 are not required to participate until that agreement expires, on terms this model cannot represent (it distinguishes only federal and non-federal employment).
  • THE PROJECTED OASDI BASE beyond 2026. The cap follows SSA’s published base exactly for 2025 and 2026 and the engine’s growth assumption after that, which is an estimate and is disclosed as one on the FICA path already.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. Maine's quarterly wage reporting is employer-by-employer, with the state system calculating the premium due FROM WHAT THAT EMPLOYER REPORTS — a weaker inference than Connecticut's or New York's own explicit "for each covered employer" language, but the same structural conclusion: someone with two employers in one year can have more than the annual maximum withheld in total. Maine lets a covered employee seek a refund of the overpayment from the Department — manual, not automatic like New Jersey’s Form NJ-2450 reconciliation, and not something the model applies. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for that household — the same gap New York’s and Oregon’s own claims disclose.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  Maine PFML Employer FAQ (December 2024): withholding "will begin with pay dates on or after January 1, 2025"; "the joint contribution rate for employers and employees is set at either 0.5 or 1 percent of wages based on the size of the employer. No more than 0.5 percent can come from the employee"; wages exclude "amounts above the annual base limit set by the U.S. Social Security Administration"; "wages do not include wages earned from federal employment" · Maine PFML Employer FAQ (July 2026 update) — the same rate and wage-base answers, confirmed current for 2026

• Delaware Paid Leave’s employee premium, the TWELFTH state payroll charge FERSCalc models, and the FIRST whose EMPLOYER-SIZE AXIS changes the ceiling itself rather than only who pays a fixed one (#429). Certified in six parts. FIRST, that it is not a Delaware income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that CONTRIBUTIONS BEGAN 1 JANUARY 2025 and nothing is charged before — an ABSENT programme rather than a zero rate, with the boundary pinned on both sides (19 Del. C. § 3705(b)). THIRD, that federal employment is outside it BY STATUTE: "'Employer' does not include... the federal government" (§ 3701(7)b.2). FOURTH, that the modelled 0.4% CEILING is half of the 0.8% combined rate an employer with 25 or more employees owes across all three lines of coverage — medical (0.4%), family caregiving (0.08%) and parental (0.32%) (§ 3705(b)(1)-(3)(a)) — the largest defensible figure across employer sizes, not a universal one; a smaller employer’s own lower-or-zero ceiling is disclosed, not computed, since the model collects no employer-headcount input. FIFTH, that the WAGE is capped at the OASDI limit — wages are defined by reference to FICA old-age, survivors, and disability insurance remuneration (§ 3701(24)) — applied per modelled earner, never to a household total. SIXTH, that an employer may deduct less than the ceiling, which is priced as a branch rather than derived.

  Tax years 2025, 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • DELAWARE’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • HOW MUCH THE EMPLOYER ACTUALLY DEDUCTS. The law caps a deduction at "not greater than 50% of the contribution required for the employee. However, an employer may elect to pay all or any portion of the employee’s share" (§ 3705(d)-(f)). The model charges the 0.4% ceiling, so a household whose employer absorbs any of it is OVERSTATED. Not derivable from anything collected; a priced branch shows the full-pickup endpoint and partial pickup lands between. That endpoint is not free: per IRS Rev. Rul. 2025-4, an employer's voluntary payment of an employee's own required contribution is additional taxable W-2 compensation, which the branch's comparison does not net out.
  • AN EMPLOYER WITH FEWER THAN 25 EMPLOYEES OWES LESS, OR NOTHING. An employer with 10 to 24 employees is required to offer only parental leave (§ 3701(7)a.1), capping the employee share at 0.16% rather than the 0.4% modelled here — half of the 0.32% parental rate (§ 3705(b)(3)a.), per the employee's statutory 50% cap (§ 3705(f)). An employer with fewer than 10 employees falls outside the "Employer" definition entirely (§ 3701(7)b.1), so its employees owe nothing under the mandate — though such an employer may still voluntarily purchase a group policy the model cannot see. FERSCalc collects no employer-headcount input, so it charges the largest tier's ceiling for every household; a household at a smaller employer is OVERSTATED, potentially all the way to zero.
  • THE RATE AFTER 2026, AND THE TRUE CEILING IS HIGHER THAN WHAT IS MODELLED. From 2027 the Department sets each line's rate "based on sound actuarial principles" (§ 3705(b)(1)-(3)(b)); the combined rate may not exceed 1.00%, and if the math would push it past that the Department cuts the benefit percentage instead of the rate (§ 3705(c)(2)b.-(c)(3)). At the statutory 50% employee cap, that allows a share as high as 0.50%, not the 0.4% modelled here. 0.4% is this table's best available ESTIMATE for 2027+ — the observed 2025-2026 figure held flat, the same caveat Colorado's post-2026 rate carries — not itself a statutory maximum.
  • THE PROJECTED OASDI BASE beyond 2026. The cap follows SSA’s published base exactly for 2025 and 2026 and the engine’s growth assumption after that, which is an estimate and is disclosed as one on the FICA path already.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. Delaware’s quarterly Hours & Wage reporting is employer-by-employer, the same structural shape Maine’s weaker inference rests on, so someone with two employers in one year can plausibly have more than the annual maximum withheld in total. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for that household — the same gap New York’s and Oregon’s own claims disclose. Delaware’s own remedy, if any, is not modelled.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  19 Del. C. § 3701 — "Employer" defined; a.1-2 set the 10-24 and 25+ headcount thresholds and what each tier must offer; b.1-3 exclude an employer with fewer than 10 employees, the federal government, and a business closed 30+ consecutive days a year; (24) defines "Wages" by reference to FICA OASDI remuneration · 19 Del. C. § 3705 — Contributions: (b) sets the 1 January 2025 start and the 2025-2026 per-line rates (medical 0.4%, family caregiving 0.08%, parental 0.32%); (d)-(f) cap the employee deduction at 50% and permit the employer to pay more · Delaware Dept. of Labor, "Notice that Payroll Deductions for Delaware Paid Leave are Starting" — confirms the 1 January 2025 start date and names "federal government agencies, railroads, and seasonal businesses" as the only employer-mandate exceptions, in the agency's own plain-language voice; undated, no byline · Delaware Dept. of Labor, "Employers & TPAs Guide to Delaware Paid Leave" — confirms the 0.32%/0.40%/0.08%/0.8% rate table and the FICA-wage-base cap in worked-example form; undated, no byline

• Hawaii Temporary Disability Insurance’s employee premium, the THIRTEENTH state payroll charge FERSCalc models, and the FIRST whose statutory cap is expressed WEEKLY rather than annually (#429). Certified in six parts. FIRST, that it is not a Hawaii income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not. SECOND, that federal employment is outside it BY STATUTE: "employment" excludes "[s]ervice performed in the employ of the United States government... exempt under the Constitution... from the contributions imposed by this chapter" (HRS § 392-5(6)), checked on both modelled earners. THIRD, that the modelled ceiling is 0.5% of wages under the ORDINARY statutory plan (HRS § 392-43(a)) — the largest defensible figure this model can compute, not the premium-limited amount a real deduction would be, and not the higher amount an approved private plan may require by agreement (§ 392-41(a)(4)-(5), excluded below). FOURTH, that the wage base — a separate quantity, defined in § 392-43(b)(2) — is published WEEKLY and moves annually — $1,441.72 for 2025, $1,500.21 for 2026 — annualised here (× 52) to $74,969.44 and $78,010.92, each a genuinely different ceiling, proven on both sides of both boundaries, and proven to hold the 2026 figure flat for a later year rather than falling back to 2025’s. FIFTH, that the cap applies per modelled earner, never to a household total. SIXTH, that an employer may bear the whole cost instead of withholding any of it, which is priced as a branch rather than derived, proven independently on both modelled earners.

  Tax years 2025, 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • HAWAII’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • THE HALF-THE-PREMIUM MECHANISM. The statute caps the deduction at "one-half the cost but not more than .5 per cent of the weekly wages" (HRS § 392-43(a)) — a minimum of TWO separate limits, not one. This model has no input carrying the actual premium a carrier charges or a self-insured plan costs, so it charges only the 0.5%-of-wages half of that minimum. TDI premiums are typically inexpensive, so half of a small premium often lands well under 0.5% even at an employer that shifts the maximum allowable share onto the employee — meaning the modelled figure OVERSTATES a real deduction for most covered employees, not merely for those whose employer elects to absorb some of the cost.
  • THE WEEKLY-TO-ANNUAL ANNUALISATION. HRS § 392-43(b)(2) caps a WEEKLY wage, not an annual one; this model has no per-week wage granularity and instead caps the person’s annual wage at 52 times the published weekly figure, computed from the wage base itself rather than from DLIR’s separately-rounded "maximum weekly deduction" quick-reference figure ($7.21 for 2025, $7.50 for 2026). The two paths do not agree to the cent, and NOT in one direction only: 52 × $7.21 = $374.92 (a few cents ABOVE this model’s $374.8472 for 2025), while 52 × $7.50 = $390.00 (a few cents BELOW this model’s $390.0546 for 2026) — DLIR’s own cent-level rounding of the weekly deduction goes a different way each year. Immaterial in size, but not a one-directional overstatement the way the half-the-premium exclusion above is.
  • AN APPROVED PRIVATE PLAN CAN REQUIRE MORE THAN THE ORDINARY 0.5% CEILING, BY AGREEMENT. An employer may satisfy chapter 392 with its own director-approved plan instead of the statutory default (HRS § 392-41(a)(4)-(5)), and such a plan "does not require contributions of any employee... in excess of the amount authorized in section 392-43, EXCEPT BY AGREEMENT AND PROVIDED THE CONTRIBUTION IS REASONABLY RELATED TO THE VALUE OF THE BENEFITS as determined by the director" — richer, director-approved benefits can carry a richer, agreed employee contribution above the ordinary 0.5% this model charges. Not derivable from anything collected, so a household on such a plan is UNDERSTATED here, the opposite direction from every other exclusion on this claim.
  • THE 14-WEEK, $400 ELIGIBILITY THRESHOLD for a NEW hire. An employer "may not deduct premium cost from an employee who does not meet the eligibility requirements of Section 392-25" — 14 weeks of Hawaii employment with 20+ hours and $400+ wages in the preceding 52 weeks (HRS § 392-25). This model assumes an ongoing career already past that threshold and does not model a brief zero-withholding window for a household’s non-federal earner who has just started work.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. Hawaii lets an employee from whose wages more than the authorized amount has been withheld seek "a refund or credit of the excess as prescribed by regulation of the director" (HRS § 392-43(e)) — the same structural shape Delaware’s and Maine’s own remedies take. The model carries one wage figure per person and applies one cap, which UNDERSTATES the contribution for a household with two employers in one year — the same gap New York’s and Oregon’s own claims disclose. Hawaii’s own remedy, if any, is not modelled.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  HRS § 392-3 — "Employer" (includes the State and its political subdivisions), "Employment", and "Wages" defined · HRS § 392-5(6) — excludes "[s]ervice performed in the employ of the United States government... exempt under the Constitution... from the contributions imposed by this chapter" from "employment" · HRS § 392-25 — eligibility for benefits: 14 weeks of employment with 20+ hours and $400+ wages in the preceding 52 weeks · HRS § 392-43 — "Authority to withhold contributions, rate of contribution, maximum weekly wage base": (a) one-half-the-cost / 0.5%-of-weekly-wages ceiling; (b)(2) the weekly wage base formula; (e) the employee’s refund right for excess withholding · HRS § 392-41(a)(4)-(5) — an employer’s own director-approved plan "does not require contributions of any employee... in excess of the amount authorized in section 392-43, EXCEPT BY AGREEMENT AND PROVIDED THE CONTRIBUTION IS REASONABLY RELATED TO THE VALUE OF THE BENEFITS as determined by the director" · DLIR Disability Compensation Division, "2026 Maximum Weekly Wage Base and Maximum Weekly Benefit Amount" (10 Dec 2025) — $1,500.21 weekly wage base, $7.50 maximum weekly deduction; footnote 3 restates the one-half-the-premium / 0.5% minimum · DLIR Disability Compensation Division, "2025 Maximum Weekly Wage Base and Maximum Weekly Benefit Amount" (2 Dec 2024) — $1,441.72 weekly wage base, $7.21 maximum weekly deduction · DLIR, "About Temporary Disability Insurance" — "The employer may pay for the entire cost of providing TDI coverage, or the employer may share the cost equally with the employees eligible for coverage"; confirms coverage is ordinarily by insured carrier plan, DCD-approved self-insured plan, or an equivalent collective bargaining agreement

• Massachusetts Paid Family and Medical Leave’s employee premium, the FIFTEENTH state payroll charge FERSCalc models (#429). Certified in nine parts. FIRST, that it is not a Massachusetts income tax: it is returned by `applyPayrollWithholding` and never by `applyTax`, and a case pins that the same wages produce the same income tax whether the earner is covered or not — checked against an INDEPENDENTLY-derived reference figure ($2,060.00 on $50,000 of wages) rather than only against itself, so an unconditional fold-in bug is caught, not only one gated on federal status. SECOND, that federal employment is outside it: PFML’s own "employee"/"employer" definitions (M.G.L. c. 175M § 1) borrow chapter 151A’s, and that chapter excludes "service performed in the employ of the United States government or of an instrumentality of the United States" (M.G.L. c. 151A § 6(e)), checked on both modelled earners. THIRD, that the modelled ceiling is 0.46% of wages for 2025-2026 under DFML’s currently-published rate — 100% of the 0.18% family-leave rate plus 40% of the 0.70% medical-leave rate (M.G.L. c. 175M § 6(c)) — the largest figure this model can defend GIVEN THAT RATE, not an unqualified statutory maximum (see the first exclusion below). FOURTH, that the Social Security wage-base cap (§ 6(f)) applies, proven on both sides of the boundary, reading the engine’s own published SSA table the same way every other capped state here does. FIFTH, that the cap applies per modelled earner, never to a household total. SIXTH, that an employer may bear the whole cost instead of withholding any of it (both deduction caps are permissive, "shall not deduct MORE THAN"/"may deduct not more than"), priced as a branch rather than derived, proven independently on both modelled earners. SEVENTH, that the 0.46% ceiling holds flat for 2025, 2026, and past the 2027 boundary DFML’s own page uses — proven at 2025 explicitly, not merely assumed from the rate table’s single row. EIGHTH, that the branch actually names the deduction-cap mechanism it prices, not merely a number. NINTH, that the SAME branch also names the municipality/political-subdivision coverage carve-out it discloses but does not price, checked against the operative facts (outside PFML entirely, owes nothing, disclosed not priced) and not merely the keywords. NONE of this resolves a genuine, disclosed disagreement between two Massachusetts state agencies over whether the enacted 2027 split-flip is already effective for 2026 — see the exclusion below.

  Tax years 2025, 2026 · 12 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MASSACHUSETTS’S INCOME TAX, which this pack does not touch. It remains uncertified and on the `state-income-tax-baseline` provenance entry; the evidence here is recorded against the cross-cutting `state-payroll-contributions` entry so that nothing reads as certifying the state.
  • WHETHER THE 2027 SPLIT-FLIP IS ALREADY EFFECTIVE FOR 2026. St. 2026, c. 101, §§ 25-26 swap the family/medical deduction caps; its own § 45 says they "apply for taxable years beginning on or after January 1, 2026", and the Dept. of Revenue's own WORKING DRAFT Technical Information Release (06/23/2026 — a draft, not a final release) reads it exactly that way — "Effective for taxable years beginning on or after January 1, 2026, employers can withhold no more than 40% of the PFML family leave contribution... Employers can withhold the entirety of the PFML medical leave contribution." DFML's own contribution-rate page — updated July 10, 2026, describing the same act by name — instead groups "2025 & 2026" under the PRE-swap split and puts the swap under "2027". Two dated, post-enactment, official sources disagree, and this pack does not adjudicate between them: it certifies the DFML-aligned 0.46% figure the model actually charges. IF DOR'S READING GOVERNS ACTUAL 2026 WITHHOLDING, the true 2026 ceiling could be 40% x 0.18% + 100% x 0.70% = 0.772% of wages — 67.8% HIGHER than 0.46% (equivalently, 0.46% is 40.4% below 0.772%) — a gap this model cannot rule out.
  • THE 25-EMPLOYEE THRESHOLD’S EFFECT ON THE EMPLOYER’S OWN SHARE. M.G.L. c. 175M § 6(d) — "notwithstanding subsection (c)" — exempts an employer with fewer than 25 employees from the employer portion of premiums for BOTH family and medical leave, not medical leave alone, via its own separate remittance rule that happens to use the same two percentages as (c)'s deduction caps for 2025-2026. Under the current split, family leave's employer portion is already zero for every employer regardless of size, so in practice size only changes who pays medical leave's remaining 0.42% — never the 0.46% an employee can be charged, so the model correctly carries no size input at all. Not modelled because it is not part of the employee’s figure, not because it is unknown.
  • THE ACTUAL 2027 TOTAL RATE, WHICH DFML HAS NOT YET PUBLISHED. DFML’s own guidance states "the total contribution rate for 2027 has not yet been set" (rates are set annually, by October 1, for the following year). This model holds the 2026 employee-side total (0.46%) flat for 2027 and later, consistent with — but not established by — § 43(a)’s direction that DFML’s administrative adjustments keep "the overall balance of employee and employer contribution levels... unchanged" when implementing the swap; that is a legislative instruction to the rate-setter, not itself a rate. A policy-directed estimate, not a published figure, and liable to move once DFML sets the actual 2027 rate.
  • A MASSACHUSETTS MUNICIPALITY, DISTRICT OR POLITICAL SUBDIVISION IS OUTSIDE PFML BY DEFAULT (M.G.L. c. 175M § 1, "shall not be subject to this chapter unless it adopts this chapter under section 10"; confirmed by DFML’s own "PFML coverage for statutorily excluded employers" guidance) — the same shape Connecticut’s and Rhode Island’s municipal carve-outs take. Not derivable from anything this model collects, so a household whose non-federal earner works for a non-opted-in municipal employer OWES NOTHING, the same $0 the priced branch's own employer-pays-it-in-full endpoint already shows for a different reason. Disclosed in the `exemptBranch` body rather than priced as a second numeric branch, matching Delaware’s convention for a secondary uncertainty it also cannot price.
  • AN APPROVED PRIVATE OR SELF-INSURED PLAN’S COST CANNOT EXCEED WHATEVER THE STATE PROGRAM ITSELF CHARGES (M.G.L. c. 175M § 11(a)(1)(xii)), unlike Hawaii’s approved-plan exception — so, unlike the Hawaii pack’s identically-shaped exclusion, this is recorded as a REASSURANCE that a private plan cannot cost MORE than the state program. It does not independently guarantee 0.46%: if DOR’s reading of the 2026 split governs the state program itself, a private plan’s own ceiling would move with it, so this does not resolve the DOR/DFML 2026 disagreement above.
  • THE PER-EMPLOYER CAP FOR A TWO-JOB EARNER. The model carries one wage figure per person and applies one cap, which UNDERSTATES the raw CONTRIBUTION actually withheld from a household with two employers in one year — the same gap New York’s, Oregon’s, Maine’s and Hawaii’s own claims disclose. Unlike those, Massachusetts names a real remedy: DOR’s own FAQ (DOR administers PFML contributions; DFML only sets the rate) confirms a worker whose combined contributions "exceed that required for the calendar year... may qualify for a corresponding tax credit... on [Massachusetts] Form 1[,] Line 48", per that FAQ’s own current wording — a credit this model does not additionally compute.
  • DEDUCTIBILITY ON A FEDERAL RETURN, for an itemizing taxpayer under § 164(a)(3). FERSCalc models the standard deduction and no itemized deductions in any jurisdiction.

  M.G.L. c. 175M § 1 — "Employee" and "Employer" defined by reference to M.G.L. c. 151A § 1; municipalities/political subdivisions excluded absent a § 10 opt-in vote · M.G.L. c. 175M § 6 — "Contributions": (a) the combined rate; (c) the 40%/100% deduction caps; (d) the 25-employee threshold's own remittance rule; (f) the OASDI wage-base cap · M.G.L. c. 175M § 11(a)(1)(xii) — an approved private plan's employee cost "shall not be greater than the cost charged to employees under the state program" · M.G.L. c. 151A § 6(e) — excludes "service performed in the employ of the United States government or of an instrumentality of the United States" from "employment" · St. 2026, c. 101 ("An Act making appropriations for the fiscal year 2026 to provide for supplementing certain existing appropriations..."), §§ 25-26 (the 40%/100% swap), § 43(a) (guidance mandate; "impact... in calendar year 2027"; the balance-preserving directive), § 45 ("taxable years beginning on or after January 1, 2026"). Approved June 12, 2026 · Mass. DOR, "Working Draft TIR: Tax Provisions in the Fiscal Year 2026 Final Mid-Year Supplemental Appropriation Bill" (06/23/2026) — reads the §§25-26 swap as "Effective for taxable years beginning on or after January 1, 2026", conflicting with DFML's page below · Mass.gov DFML, "Paid Family and Medical Leave employer contribution rates and calculator" (updated July 10, 2026) — 2025-2026 0.88% total / 0.18% family / 0.70% medical split, the 0.46% small-employer effective rate, and the 2027 framing ("the total contribution rate for 2027 has not yet been set") · Mass.gov DFML, "Employers and employment excluded from Paid Family and Medical Leave" — confirms the c. 151A §§ 2/6/6A exclusions carry over to PFML · Mass.gov DOR (not DFML — this FAQ is offered by the Department of Revenue), "PFML Registration, Contributions, and Payments" — a worker with two-or-more MA employers whose combined contributions exceed the SSA annual limit "may qualify for a corresponding tax credit... on [Massachusetts] Form 1[,] Line 48", per that FAQ's own current wording · Mass.gov DFML, "Paid Family and Medical Leave (PFML) coverage for statutorily excluded employers" — "Municipalities, districts, political subdivisions or authorities are excluded from PFML law unless their governing body... votes to opt-in"

• The year-scheduled rate — 4.5% (2024), 4.25% (2025), 3.99% (2026), 3.49% (2027-2029), 3.24% (2030-2032), 2.99% (2033-2034, unconditionally, and held flat thereafter) — per G.S. § 105-153.7(a) AS REWRITTEN BY S.L. 2026-41 § 44.1 (signed 7 July 2026), and the standard deduction — $12,750 single / $25,500 joint OR qualifying surviving spouse — per G.S. § 105-153.5(a)(1), a flat statutory table with no year dimension at all (confirmed untouched by S.L. 2026-41, and confirmed identical across the 2024 and 2025 D-401 booklets rather than assumed from the coincidence of two adjacent years), read via `filingStatus` so a real surviving spouse — whose `isSingle` reads true by the engine's own cross-state convention — still receives the joint figure rather than the single one.

  Tax years 2024, 2025, 2026, 2027, 2028, 2029, 2030, 2031, 2032, 2033, 2034 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • G.S. § 105-153.7(a1)'s REWRITTEN revenue-reduction trigger for 2035 and later, which can cut the rate further — to the greater of (the prior year's rate minus 0.25 points, HALVED from the pre-S.L.-2026-41 0.50) or 2.49% — for any year whose preceding fiscal year cleared that year's own trigger amount, a chain running through FY 2038-2039 / tax year 2040. Whether any post-2034 year's cut actually lands cannot be derived from the statute alone; the engine holds 2.99% flat rather than assume a cut that would only ever lower the true tax further, so the MODEL can only OVERSTATE — never understate — a triggered year relative to what the retiree would actually owe.
  • Head of Household ($19,125) — one of § 105-153.5(a)(1)'s own four filing-status rows. FERSCalc resolves filing status to `isSingle`/`filingStatus` ('single' | 'mfj' | 'surviving_spouse') in every state and has no Head of Household status anywhere, so this row is unreachable engine-wide, not a North-Carolina-specific gap.
  • The $0 a Married-Filing-Separately taxpayer gets when their spouse itemizes. THIS IS NOT A FIFTH ROW IN THE STATUTE'S OWN TABLE — § 105-153.5(a)(1) lists four filing-status rows, in this order: MFJ/surviving spouse, Head of Household, Single, MFS (each a flat dollar figure), and, separately, a federal-eligibility sentence: "The standard deduction amount is zero for a person who is not eligible for a standard deduction under section 63 of the Code." An MFS filer whose spouse itemizes loses that federal eligibility, which is where the D-401 booklet's own "$0" outcome comes from. FERSCalc has no separate-filing status anywhere, so this outcome is unreachable engine-wide as well.
  • Itemized deductions, which the engine never elects in any state.

  N.C.G.S. § 105-153.7 — "Individual income tax imposed": (a) the rate table, (a1) the revenue-reduction trigger and its fiscal-year table — AS REWRITTEN by S.L. 2026-41 § 44.1. The consolidated mirror at this URL had not yet incorporated that rewrite as of this pack's first draft; see the S.L. 2026-41 source below, which is authoritative where the two disagree · Session Law 2026-41 (Senate Bill 257), "Current Operations Appropriations Act of 2026" — § 44.1 rewrites G.S. § 105-153.7(a)/(a1) exactly as this claim states ("In 2027, 2028, and 2029 3.49% ... In 2030, 2031, and 2032 3.24% ... After 2032 2.99%"; the trigger step drops to 0.25 points; the trigger table's first row becomes FY2033-2034/TY2035). Ratified 2 July 2026; signed by the Governor 10:30 a.m. 7 July 2026; § 44.1.(b) makes the rewrite effective on signature · N.C.G.S. § 105-153.5(a)(1) — "Standard deduction amount": the filing-status table (Single $12,750, MFJ/QSS $25,500, HoH $19,125, MFS $12,750), plus the separate federal-eligibility sentence that zeroes it for an ineligible filer. Confirmed untouched by S.L. 2026-41 § 44.2, which rewrites only the itemized-deduction subdivision (2) · NCDOR 2025 D-401 Individual Income Tax Instructions — "For tax year 2025, the individual income tax rate is 4.25%"; N.C. Standard Deduction Chart, whose "Married filing jointly/Qualifying widow(er)/Surviving spouse" row is one figure, not two · NCDOR 2024 D-401 Individual Income Tax Instructions — "For tax year 2024, the individual income tax rate is 4.5%"; same standard-deduction chart

• Social Security benefits are subtracted from adjusted gross income in full per G.S. § 105-153.5(b)(3) ("Benefits received under Title II of the Social Security Act"), regardless of the level of any other income — proven against a household whose other income is well above the standard deduction, not merely a zero-income smoke test.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Railroad Retirement Act benefits, the other half of the same subdivision (b)(3) subtraction. The engine has no railroad-retirement input in any state.

  N.C.G.S. § 105-153.5(b)(3) — "Benefits received under Title II of the Social Security Act and amounts received from retirement annuities or pensions paid under the provisions of the Railroad Retirement Act of 1937"

• What `NC_RULE.applyTax` does with two ALREADY-DETERMINED, per-person Bailey-excludable pension figures, per G.S. § 105-153.5(b)(5): the two figures are summed and subtracted from adjusted gross income once, against the household pension total (not merely deducted against a bracket), clamped so the total can never reach past the pension category into wages or TSP even when the excludable figure itself exceeds the pension, and TSP stays fully taxable regardless of the annuity's own exclusion. This claim does NOT certify the vesting DERIVATION `baileyVestingForEmployee` performs — the strict-count/generous-bound methodology, the exactly-five-years boundary at 12 August 1984, and the `uncertain` verdict's charge-anyway direction are certified in fers.test.ts and projection.test.ts, and this pack is deliberately blind to that adapter — including the invariant that each person's own attributed figure never exceeds that person's own actual pension, which the rule's household-level clamp does not independently re-verify.

  Tax years 2024, 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Bailey vesting DERIVATION itself — see the claim statement. Certified elsewhere, not here.
  • Bailey's treatment of the TSP, which the rule does not grant at all: Directive PD-99-2 resolves the Thrift Plan question with a component-balance fraction over the Form TSP-8 split (employee contribution, agency matching, agency automatic) this calculator does not hold and cannot derive from a traditional and Roth balance alone, so TSP withdrawals stay fully taxable for a qualifying retiree — a disclosed DATA gap, not unsettled law.
  • Whether military service, refunded service, or service requiring a deposit counts toward the five years — the NCDOR directives are silent, and the engine excludes all three, the reading that qualifies fewer people (#383 area).
  • The `uncertain` vesting outcome's own charge-anyway direction, priced as a future comparison at #383 rather than resolved here.
  • Whether one person's over-attributed exclusion figure could shelter the OTHER person's pension — the rule clamps the SUMMED exclusion against the HOUSEHOLD pension total, not each person's own figure against that person's own pension, so a figure that exceeded its own person's pension (an adapter-side invariant this pack does not re-verify) could in principle reach the other person's income. Out of scope by the same adapter-blindness this claim states throughout.

  N.C.G.S. § 105-153.5(b)(5) — the subtraction for amounts exempt under the Bailey/Emory/Patton court-order settlements · NCDOR — "Bailey Decision Concerning Federal, State and Local Retirement Benefits": the five-year vesting test for FERS/CSRS and the separate contribution-based test for §401(k)/§457 plans · NCDOR Directive PD-99-2 — the Form TSP-8 component-balance-fraction detail this claim's TSP exclusion relies on. Cited directly because the general Bailey Decision page above states only the five-year vesting test, not this level of TSP-specific mechanics

• The flat 2.5% rate, UNCONDITIONAL for the years this claim covers (no year-stepped schedule like North Carolina's or Georgia's), per AZ DOR Form 140 instructions (2024 and 2025 editions), Line 46: "Multiply line 45 by 2.5% (.025)." The standard deduction is year-sensitive per ARS §43-1041(H)'s federal-indexing-METHOD language (a methodology-parity clause, not a definitional pass-through — Session Laws 2026, Chapter 140 §16 had to manually re-strike and replace §43-1041.A's own dollar figures when the federal 2025 amount was restructured, which a routine CPI-only indexer would not have needed) and is read from each year's own DOR-published amount: 2024 single $14,600 / MFJ $29,200, 2025 single $15,750 / MFJ $31,500. A qualifying surviving spouse (`filingStatus: 'surviving_spouse'`) takes a THIRD figure, the HEAD OF HOUSEHOLD amount ($23,625 for 2025) — neither the single nor the joint figure — per Form 140's own Box 5/Box 7 filing-status instructions, read via `filingStatus` rather than `isSingle`, which reads TRUE for this household by the engine's own cross-state convention.

  Tax years 2024, 2025 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Two 2025-legislative-session bills, SB1318 (a "structural surplus" trigger toward roughly 2.42%) and HB2918 (2.47%), each of which would cut the rate further starting tax year 2026. Neither was confirmed enacted as of this certification — AZ DOR's own 2026 Form 140ES estimated-tax instructions were checked directly and state no rate change for 2026. Revisit if either bill is confirmed signed.
  • Tax year 2026 and later standard-deduction amounts. Arizona has not yet published a 2026 figure; the engine holds the 2025 amounts forward as a labelled projection rather than substituting the federal 2026 basic standard deduction, per the methodology-parity finding above. NO case in this claim certifies a 2026-or-later deduction figure as itself confirmed — only that the engine holds 2025 forward rather than assuming a new one.
  • Married Filing Separately, and Head of Household as a GENUINE, non-QSS filing status (an unmarried person with a dependent who is not a surviving spouse). FERSCalc resolves filing status to `isSingle`/`filingStatus` ('single' | 'mfj' | 'surviving_spouse') in every state and has no true Head of Household status anywhere; the HOH figure this claim certifies is reached ONLY through the `surviving_spouse` value.
  • Itemized deductions, which the engine never elects in any state.
  • Tax years before 2024. The flat 2.5% rate did not apply before TY2023 (TY2021 used the old graduated schedule plus the Prop 208 surcharge; TY2022 used an intermediate two-bracket 2.55%/2.98% schedule, confirmed against actual filed-year AZ DOR forms for both years), and this claim does not certify a TY2023 standard-deduction figure either, so the engine's own year clamp does not reach back before 2024.
  • Whether the projection adapter's 'immediate' filing-status-switch mode genuinely represents federal QSS in the DEATH YEAR specifically. It does not: that mode's own code comment calls the death-year mapping to `surviving_spouse` a "QSS-equivalent" approximation, not literal federal QSS, so in that mode this rule's HOH treatment reaches a death-year household federal law would still treat as jointly filing. An engine-wide approximation this rule inherits, not one specific to Arizona.

  Arizona Form 140 (Resident Personal Income Tax Return) Instructions, 2025 — Line 46 ("Multiply line 45 by 2.5% (.025)"); the standard-deduction table; and the Box 5/Box 7 filing-status instructions · Arizona Form 140 Instructions, 2024 — the same Line 46, standard-deduction table, and filing-status language for that year · Arizona Revised Statutes § 43-1011 — the flat-tax-rate statute (this URL is the corrected one — azleg.gov's statute-index entry, not the single-digit URL pattern most other cited ARS sections follow; paragraphs 1-9 of subsection A are present here, but subsection F, which those paragraphs reference, is still absent from this page — the codified mirror lags any law enacted DURING the current legislative session, per its own disclaimer, which is also why Chapter 140 does not appear there yet) · Arizona 2022 Tax Tables X and Y (Full-Year Residents) — "$0 ... $28,653 ... 2.55% ... $28,653 ... 2.98%" (single) and the doubled MFJ thresholds, confirming TY2022 was NOT flat 2.5%, for the rate-history correction in this claim's own exclusions

• Social Security benefits are subtracted from adjusted gross income in full per ARS §43-1022 ¶10 ("[t]he amount included in federal adjusted gross income pursuant to section 86 of the internal revenue code, relating to taxation of social security and railroad retirement benefits"), regardless of the level of any other income — proven against a household whose other income is well above the standard deduction, not merely a zero-income smoke test.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Railroad Retirement Act benefits, the other half of the same ¶10 subtraction. The engine has no railroad-retirement input in any state.

  Arizona Revised Statutes § 43-1022 — "Subtractions from Arizona gross income", paragraph 10

• The ARS §43-1022 ¶2 government-pension deduction is $2,500 PER PERSON, not per return, and each person's own cap is bounded by that same person's own pension income — a spouse's unused headroom cannot shelter the other's. Confirmed by Form 140i Line 29a, identical wording in the 2024 and 2025 editions: "If both you and your spouse receive such pension income, each spouse may subtract the amount received or $2,500, whichever is less." Proven against fixtures with income above the standard deduction, so a defect that folded the deduction into a zero-income coincidence would be caught.

  Tax years 2024, 2025 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • WHETHER TSP DISTRIBUTIONS QUALIFY FOR THIS DEDUCTION — DISCLOSED AS GENUINELY UNSETTLED. ARS §43-1022 ¶2(a) reaches "any other retirement system or plan established by federal law," and the Form 140i NOTE on that line narrows it to a plan "authorized and enacted into the U.S. Code" as distinct from one "only regulated by federal law." The Thrift Savings Plan is itself established by 5 U.S.C. ch. 84 — arguably "enacted into the U.S. Code" on that text — but no ARS §43-1022 ruling or DOR guidance naming the TSP was found either way (ITR 93-13 addresses OTHER-STATE public pensions, a different question, and general secondary retirement-planning sources are themselves split without citing authority). The rule taxes TSP in full; this is a guaranteed upper bound on tax only in the narrow sense of never being LESS than a rule that granted the deduction, not a guaranteed overstatement in every case. This claim certifies CURRENT BEHAVIOUR as an accurate description of what the code does, NOT as a resolved legal question.
  • Uniformed-services retired or retainer pay, excluded at 100% under Form 140's SEPARATE Line 29b, not the $2,500 deduction this claim covers. The calculator has no military-retired-pay input distinct from the FERS/CSRS annuity `pensionIncomePersonA/B` represents.
  • The deduction's application to a non-federal Arizona state or local government pension (ASRS, elected officials' plan, county/city/town plans, and others Form 140i's Line 29a list names), OR to a non-federal PRIVATE pension that could in principle reach `pensionIncomePersonA/B` through `fixedRetirementIncome`. That field's own documentation in state-taxes.ts records it as unreachable from the web app today (always zero), so this is a DORMANT rule gap — the deduction would be wrongly granted to a private pension if that field were ever populated — not a live one.

  Arizona Form 140i, 2025 — Line 29a, "Exclusion for U.S. Government, Arizona State or Local Government Pensions": the per-spouse $2,500 figure and the qualifying-plan list · Arizona Form 140i, 2024 — the same Line 29a language, confirming the $2,500 per-spouse figure is a flat statutory amount, not indexed · Arizona Revised Statutes § 43-1022 — "Subtractions from Arizona gross income", paragraph 2

• A $2,100 PER-PERSON age-65 exemption, subtracted from Arizona gross income BEFORE the standard deduction — a separate step, not folded into it. ARS §43-1023(E): "[a] taxpayer is allowed an exemption of $2,100" at 65+ by year end (E)(1), "[f]or the taxpayer's spouse if the spouse has attained sixty-five years of age" on a joint return (E)(2). Confirmed still operative, unchanged by Arizona's 2021 flat-tax overhaul, on Form 140's own Line 38 in BOTH the 2024 and 2025 editions ("Multiply the number in box 8 ... by $2,100"). Each person's own age independently controls their own leg of the exemption — proven with the qualifying age on EACH side of an MFJ return in turn — and a qualifying-surviving-spouse return, which carries no second filer, draws only the ONE exemption its own filer qualifies for even when the underlying household data happens to carry a qualifying age for the absent second person. A DECEDENT KEEPS THEIR OWN AGE ON THEIR FINAL RETURN (#360): the rule reads `ageAtEndOfTaxableYear(agePersonA, finalReturnDecedentPersonA)`, the same decedent-aware helper DC, Georgia, Ohio, Illinois and Missouri already call for their own age-gated provisions, rather than the raw, decedent-blind age field — proven both for a single decedent's own final return and for a joint return where only the decedent's own leg is affected.

  Tax years 2024, 2025 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The § 43-1023(E) condition that the person not be "claimed as a dependent by another taxpayer." The engine carries no dependency-status input for either person, so this claim assumes the condition is satisfied — an engine-wide limitation, not Arizona-specific.
  • Tax years before 2024, for the same reasons the rate-and-standard-deduction claim's own exclusions state.
  • THE AGE CONVENTION: this claim reads "attained sixty-five years of age before the close of the taxable year" using the engine's own plain year-end age (`agePersonA/B`, or a decedent's age at death), NOT the federal IRC § 63(f) day-before-the-birthday convention several other certified states use for their own age-gated provisions (e.g. California, Virginia, Missouri). Whether Arizona's own statute or DOR guidance would actually require § 63(f) here — which would move the answer only for a taxpayer born 1 January — was not independently researched for this claim; disclosed rather than resolved, matching how the engine-wide plain-age default is documented everywhere it applies.

  Arizona Revised Statutes § 43-1023(E) — the $2,100 per-person age-65 exemption · Arizona Form 140i, 2025 — Line 38, "Age 65 or Over": "Multiply the number in box 8 on the front of your return by $2,100 and enter the result" · Arizona Form 140i, 2024 — the same Line 38 language, confirming the exemption predates and survives Chapter 140

• From tax years beginning after 31 December 2024 (TY2025 onward), a further subtraction equal to the federal IRC §151(d)(5)(C) enhanced senior deduction ACTUALLY ALLOWED — the post-MAGI-phase-out figure, not the flat statutory ceiling. Session Laws 2026, Chapter 140 (House Bill 4168) §15 added ARS §43-1022 ¶35: "[f]or taxable years beginning from and after December 31, 2024, ... the amount deducted for a qualified individual under section 151(d)(5)(C) of the internal revenue code" — enacted retroactively to TY2025 by the bill's own applicability clause. Because the statute incorporates the federal provision by reference to the amount actually deducted, the rule reads the federal calculator's own already-phased-out figure (`federalEnhancedSeniorDeduction`) rather than recomputing IRC §151(d)(5)(C) inside the state rule — proven to reduce the base by exactly the subtraction amount, to floor at zero rather than go negative, and to stack correctly alongside the § 43-1022 ¶2 pension deduction and the § 43-1023(E) age exemption in the same computation.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether the federal calculator's own IRC §151(d)(5)(C) computation (amount, MAGI thresholds, phaseout rate, 2025-2028 window) is itself correct. That is certified separately in federal-tax-certification.test.ts; this claim is deliberately blind to that computation and certifies only what AZ_RULE does with the federal calculator's already-produced figure.
  • Tax years before 2025, and tax years after the federal provision's own current 2028 sunset — both produce a $0 federal figure by construction, which this rule then correctly subtracts as zero, but neither is separately exercised by a case naming that year.
  • Whether this subtraction would follow a FUTURE federal extension past 2028: ARS §43-105's "internal revenue code" definition is a static, date-locked reference Arizona's legislature re-enacts annually (Chapter 140 itself set it to "in effect on January 1, 2026 ... excluding any changes to the code enacted after January 1, 2026"), so that would need a later Arizona conformity act, not something this rule can determine now.
  • The MCTCP Worksheet's other three lines — a qualified tips deduction, a qualified overtime-compensation deduction, and a qualified passenger-vehicle-loan-interest deduction — each conformed by Arizona to its own new 2025 federal counterpart. None of the three federal deductions is itself modelled anywhere in this calculator, so there is no federal figure for an Arizona rule to conform to yet.

  Session Laws 2026, Chapter 140 (House Bill 4168), 57th Legislature, 2nd Regular Session, §15 — new ARS §43-1022 ¶35, the enhanced-senior subtraction, and its retroactive applicability clause · Arizona Department of Revenue, Middle Class Tax Cuts Package (MCTCP) Worksheet, calendar year 2025 (ADOR (25)) — DOR's own mechanic for this subtraction: "Enter the amount from I.R.S. Schedule 1-A Line 37 (Enhanced deduction for seniors)" into Other Adjustments

• A FERS Survivor Benefit Plan annuity is deducted from Ohio AGI, GROSS, under R.C. 5747.01(A)(4) and OAC 5703-7-08(A)(2)/(D) ("[s]urvivor benefits received pursuant to a survivorship plan if the benefit payment is conditioned on the death of the covered individual") — with NO age gate, unlike the District of Columbia's reading of the same underlying `survivorBenefitsPersonA/B` field (#352). It also does not count toward R.C. 5747.055(B)'s retirement income credit (defined by R.C. 5747.055(A)(1) as income received "on account of retirement", which a survivor annuity is not) or toward R.C. 5747.05(E)'s joint-filing-credit qualifying-income test (the IT-1040 booklet's own words: amounts deducted on the Schedule of Adjustments "are not included in Ohio AGI, and thus are not 'qualifying income'"). A REAL provision an adversarial review of this pack's first draft found entirely missing — `pensionIncomePersonA/B` folds a survivor annuity into the same figure as a person's own annuity, so it was being taxed in full.

  Tax years 2024, 2025, 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Survivor benefits from the Ohio Public Employees Retirement System or other non-federal plans, out of scope for the same reason those plans are out of scope elsewhere in this pack: the calculator has no non-federal-pension input reaching this rule.

  Ohio Revised Code § 5747.01(A)(4) — the disability and survivor's benefits deduction · Ohio Administrative Code 5703-7-08 — "Deduction of disability and survivorship benefits", narrowing "survivor benefits" to those conditioned on the covered individual's death

• Social Security Title II benefits (and tier-1 railroad retirement) are deducted from federal AGI to reach Ohio AGI, per R.C. 5747.01(A)(5)(a) and the IT-1040 booklet's own Line 16 instruction ("Deduct the amount on your federal 1040 or 1040-SR, line 6b"), regardless of the level of other income — proven against a household whose other income is well above the $26,050 threshold, not merely a zero-income smoke test.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Railroad Retirement Act benefits other than tier-1, the other half of the same R.C. 5747.01(A)(5)(b) deduction. The engine has no railroad-retirement input in any state.

  Ohio Revised Code § 5747.01(A)(5) — "Adjusted gross income" definition, the Social Security and tier-1 railroad retirement deduction

• The nonbusiness income tax schedule is YEAR-SENSITIVE per R.C. 5747.02(A)(3), which codifies three consecutive years' schedules directly: 2024 — nothing to $26,050, then $360.69 + 2.75%, then $2,394.32 + 3.5% above $100,000; 2025 — $342.00 + 2.75%, then $2,394.32 + 3.125% above $100,000; 2026 and thereafter — a SINGLE $332.00 + 2.75% bracket with NO upper bracket at all, the statute's own "and thereafter" text. Today falls inside tax year 2026, so this is a correction to the CURRENT year's modelled tax, not a future-year projection.

  Tax years 2024, 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Tax year 2027 and later DOLLAR FIGURES. R.C. 5747.02(A)(5) lets the tax commissioner move both the $26,050 threshold and the base dollar amounts each August by the GDP deflator, but HB96's own uncodified Section 757.120(A) (read from the enrolled bill) suspends that adjustment specifically for 2025 and 2026 — "[t]he Tax Commissioner shall not make adjustments in 2025 or 2026" — which CONFIRMS those two years' figures as exact rather than merely stable by coincidence. The suspension does not reach 2027, where the ordinary mechanism resumes absent further legislation, so 2026+ is held forward from the statute's own "and thereafter" STRUCTURE (permanent law) but the specific dollar figures beyond 2026 remain a disclosed projection, not a statutory certainty.
  • Tax years before 2024, which used a different, five-bracket schedule this pack does not model.
  • Estates and trusts (R.C. 5747.02(A)(1)/(2)), a different rate and income measure entirely.

  Ohio Revised Code § 5747.02(A)(3) — the three-year nonbusiness income tax rate schedule (2024/2025/2026-and-thereafter), and (A)(5), the annual GDP-deflator adjustment mechanism · Ohio IT-1040 Individual Income Tax Instructions, 2025 — the nonbusiness income tax schedule as published to filers, confirming the $342.00/3.125% figures · Ohio IT-1040 Individual Income Tax Instructions, 2024 — the same schedule as published for that year, confirming the $360.69/3.5% figures · Am. Sub. H.B. No. 96, 136th General Assembly, enrolled text, Section 757.120(A) — suspends the R.C. 5747.02(A)(5) GDP-deflator adjustment specifically for 2025 and 2026

• The personal exemption ($2,400/$2,150/$1,900 by MAGI band, per person, taxpayer and spouse only) is confirmed stable across the 2024 and 2025 IT-1040 booklets. The top band's own MAGI CEILING is separately year-sensitive per R.C. 5747.025(A): no ceiling clause exists for 2024 (the $1,900 band applies unconditionally to any MAGI above $80,000); "less than seven hundred fifty thousand dollars for taxable years beginning in 2025" (the exemption is zeroed entirely, not merely reduced, at $750,000 and above); "five hundred thousand dollars for taxable years beginning in 2026 or thereafter" — a full $250,000 lower than 2025's ceiling.

  Tax years 2024, 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The exact TY2026 dollar amounts ($2,400/$2,150/$1,900). These are held forward from 2025 as a disclosed projection — the one source found that might state an official DOR-confirmed TY2026 figure directly (an ohiocpa.com guidance article) returned a Cloudflare JS challenge to both a browser-User-Agent curl and a direct fetch.
  • R.C. 5747.025(C)'s own annual GDP-deflator adjustment mechanism for the exemption dollar amounts themselves, for the same reason given in the rate-schedule claim's exclusions.
  • Dependent exemptions. The engine has no dependant-count input in any state, so the exemption count here is always 1 (single) or 2 (joint), understating the true exemption for anyone claiming dependents.

  Ohio Revised Code § 5747.025(A) — the personal exemption dollar amounts and the year-sensitive MAGI ceiling clause · Ohio IT-1040 Individual Income Tax Instructions, 2025 — Table 1, the personal exemption by MAGI band · Ohio IT-1040 Individual Income Tax Instructions, 2024 — the same Table 1, confirming the dollar amounts are unchanged from 2025

• R.C. 5747.055(B)'s retirement income credit table (0 / $25 / $50 / $80 / $130 / $200 by eligible retirement income) is applied against pension and TSP income only — wages and Social Security (already deducted) never count toward "eligible retirement income" — and is barred once MAGI less exemptions reaches $100,000, tested exactly at that boundary. Confirmed unchanged since the section's October 17, 2019 amendment, unaffected by the 2025 rate/exemption legislation.

  Tax years 2024, 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The lump-sum retirement credit (R.C. 5747.055(C)/(E)) and its election history. Out of scope for a MISSING-INPUT reason, not a can't-arise one — a TSP can, in principle, be drawn as a genuine single-year lump-sum distribution (OPM's own FERS election options confirm the whole account may be taken as one payment) — but the calculator collects no lump-sum-distribution input or prior-election history in any state.

  Ohio Revised Code § 5747.055(B) — the retirement income credit table and its $100,000 MAGI-less-exemptions ceiling ("less than one hundred thousand dollars") · Ohio IT-1040 Individual Income Tax Instructions, 2025 — Table 2, the retirement income credit worksheet · Ohio IT-1040 Individual Income Tax Instructions, 2024 — the same Table 2, confirming the credit table is unchanged from 2025

• R.C. 5747.055(F)'s $50 senior citizen credit is PER RETURN, not per person — "fifty dollars for each return" — so a joint return with both spouses 65+ receives the identical single $50 credit as one with only one qualifying spouse, and the spousal leg does not apply at all on a return with no second filer (a single or QSS return) regardless of what a stale `agePersonB` value happens to hold. The 65-or-older test is "during the taxable year", the same any-moment-in-the-year reading Arizona, Georgia, DC, Illinois and Missouri's own age-gated provisions already certify (#360): a decedent still on their own final return keeps the age they reached before death. Also barred at the same $100,000 MAGI-less-exemptions ceiling as the retirement income credit.

  Tax years 2024, 2025, 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The lump-sum senior citizen credit election (R.C. 5747.055(G)), for the same reason the lump-sum retirement credit is excluded from the sibling claim.

  Ohio Revised Code § 5747.055(F) — the $50-per-return senior citizen credit and its "sixty-five years of age or older during the taxable year" test · Ohio IT-1040 Individual Income Tax Instructions, 2025 — the senior citizen credit worksheet and the "What if a Taxpayer Is Deceased?" filing instruction

• R.C. 5747.022 grants $20 per exemption claimed — the taxpayer, spouse, and each dependent under R.C. 5747.02 — when MAGI less exemptions is under $30,000, a SEPARATE and LOWER ceiling than the $100,000 the retirement and senior credits use. A REAL provision an adversarial review of this pack's first draft found entirely missing; every dollar figure below $30,000 of MAGI less exemptions anywhere else in this pack, and in the pre-existing #366-era ordinary tests, changed once it was added.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Dependent exemptions, for the same engine-wide reason the personal exemption claim above excludes them: the engine has no dependant-count input in any state.

  Ohio Revised Code § 5747.022 — the $20-per-exemption credit and its $30,000 MAGI-less-exemptions ceiling · Ohio IT-1040 Individual Income Tax Instructions, 2025 — Line 9, "Exemption Credit"

• R.C. 5747.05(E) grants a joint filing credit on an actual MFJ return where EACH spouse independently has at least $500 of qualifying income (AGI exclusive of interest, dividends, capital gains, rent, and royalties — categories this engine does not track as Ohio income in the first place, so every dollar of wages, pension, and TSP already qualifies). The credit is the lesser of $650 or a percentage (20%/15%/10%/5%, by MAGI-less-exemptions band) of the tax remaining after the credits that precede it. The credit's own EXISTENCE — distinct from its percentage — is barred by Ohio AGI itself (not MAGI less exemptions) reaching a ceiling that is YEAR-SENSITIVE in the identical shape to the personal exemption's own: no ceiling in 2024, under $750,000 in 2025, under $500,000 in 2026 and thereafter, confirmed both by R.C. 5747.05(E)(2)'s own text and by the 2024 IT-1040 booklet's Line 12 table having no such ceiling where the 2025 booklet's does. The SECOND provision an adversarial review found entirely missing — the pack's own first-draft MFJ fixtures put all household income on one spouse, which cannot expose this credit's existence any more than it could expose its absence.

  Tax years 2024, 2025, 2026 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The table's own 20% band ("$25,000 or less" of MAGI less exemptions) — STRUCTURALLY UNREACHABLE for a nonzero credit under this rule, in any of the three certified years: Ohio's own $26,050 zero-tax threshold (R.C. 5747.02(A)(3)) exceeds $25,000, so MAGI less exemptions can never be both "$25,000 or less" and attached to a nonzero tax bill. Not a defect — a genuine structural fact about how these two provisions interact — and not certified for that reason.
  • Married Filing Separately, out of scope for the same reason the rate-and-standard-deduction claims above exclude it: the engine has no such filing status anywhere.

  Ohio Revised Code § 5747.05(E) — the joint filing credit, its per-spouse $500 qualifying-income floor, its percentage table, and its year-sensitive MAGI ceiling · Ohio Revised Code § 5747.98 — the statutory order for claiming credits, placing the $20 exemption credit and then the joint filing credit after the retirement and senior citizen credits · Ohio IT-1040 Individual Income Tax Instructions, 2025 — Line 12, "Joint Filing Credit", including the definition of "qualifying income" and the Kevin/Krysten worked example · Ohio IT-1040 Individual Income Tax Instructions, 2024 — the same Line 12, confirming the credit's table had NO MAGI ceiling that year (the "$75,001 or more: 5%" band ran unbounded)

• The graduated rate schedule under § 40-18-5 — 2% to $500, 4% $500-$3,000, 5% above (Single/MFS); double those breakpoints for MFJ/Head of Family — read from the 2025 Form 40 booklet's own Tax Table (its two-column structure) and corroborated by the booklet's own "Mr. and Mrs. Brown" worked example ($23,360 MFJ taxable income, $1,088 tax) and its "Over $100,000.00" worksheet, which applies the same flat 5% above that threshold.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Alabama's own rounded Tax TABLE — mostly $100-wide bands, though its first two rows are $50-wide — which the booklet requires below $100,000 of taxable income, and the "Over $100,000" worksheet's own carried-forward constant above it — confirmed $2 below this model's exact continuous arithmetic ($4,958 vs. $4,960 Single/MFS, $4,918 vs. $4,920 MFJ at the $100,000 boundary). This model computes the bracket rates exactly instead, the same disclosed gap as this engine's DC, Maryland, California and New York rules; a real filed return can differ from these figures by a few dollars.
  • Head of Family's own rate breakpoints, which are identical to MFJ's in the Tax Table but unreachable here: FERSCalc has no Head-of-Household-equivalent filing status anywhere. The "Over $100,000" worksheet places Head of Family under the Single/MFS constant ($4,958) rather than MFJ's ($4,918) — ADOR's own materials disagree with the tax table on which column Head of Family belongs to; unreachable either way, so not certified in either direction.

  2025 Form 40 Individual Income Tax Booklet — the Tax Table (p.24), its own "Mr. and Mrs. Brown" worked example, and the "Over $100,000.00" worksheet

• The standard deduction chart on Form 40 booklet p.9, both the Single column ("$0-$25,999 $3,000", stepping down $25 per $500 of AL AGI from $26,000, floored at $2,500 from $35,500) and the MFJ column ("$0-$25,999 $8,500", stepping down $175 per $500, floored at $5,000 from $35,500) — confirmed identical across the 2024 and 2025 booklets, so it is a STABLE table with no year dimension. The pre-fix Single-filer formula matched neither column: it borrowed MFS's own income thresholds ($12,999/$17,750) with Single's own base and floor amounts, under-deducting every Single filer below $26,000 of AL AGI.

  Tax years 2024, 2025 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Married Filing Separately's own column ("$0-$12,999 $4,250", stepping down $88 per $250, floored at $2,500 from $17,750) and Head of Family's own column ("$0-$25,999 $5,200", stepping down $135 per $500, floored at $2,500 from $35,500) — both unreachable, since FERSCalc has neither filing status anywhere.
  • The printed MFJ column's own typo at its third row ("$25,500 – $26,999" where the sequence demands "$26,500 – $26,999", reproduced identically in both the 2024 and 2025 booklets — an ADOR error, not a scan artifact). This is a genuine chart ambiguity, not a harmless one: the typo's literal range conflicts with BOTH the row above it ("$0-$25,999 → $8,500") AND the row it was meant to replace ("$26,000-$26,499 → $8,325") — a filer with AL AGI in $26,000-$26,499 sees two different printed deductions, $8,325 or $8,150, an $8.75 difference at the 5% margin. Not certified as a literal range: the formula reproduces the REGULAR step sequence every other column follows at that row ($8,150 for that band), which is a defensible reading but not the ONLY one a real filer could take from ADOR's own inconsistent chart.

  2025 Form 40 Individual Income Tax Booklet — Standard Deduction chart, p.9 · 2024 Form 40 Individual Income Tax Booklet — the same Standard Deduction chart, byte-for-byte identical, confirming the table is stable rather than a coincidence of two adjacent years

• The flat, non-income-tested personal exemption from the blank Form 40 return's own filing-status boxes: $1,500 for Single (Box 1) and Married Filing Separate (Box 3), $3,000 for Married Filing Joint (Box 2) and Head of Family (Box 4) — a single figure per status with no income column at all, unlike the standard deduction.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Married Filing Separate's own $1,500 exemption and Head of Family's own $3,000 exemption — both unreachable, since FERSCalc has neither filing status anywhere.
  • The dependent exemption (a separate line on Form 40), which the engine does not model in any state.

  Blank 2025 Form 40 return — the four filing-status exemption boxes directly

• Social Security and a defined-benefit (FERS annuity) pension are fully exempt from Alabama AGI under Ala. Admin. Code r. 810-3-19-.01(b)-(c), while TSP and other defined-contribution distributions remain fully taxable under r. 810-3-19-.04's defined-benefit test, which excludes a defined-contribution plan by definition.

  Tax years 2024, 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Non-federal defined-benefit plans and their own vesting or service-length tests. The engine's `nonFederalPensionIncomePersonA/B` field CAN reach `pensionIncome` and therefore this rule, but the web app's only `Employee` builders never populate it (always zero from the UI today, matching this field's own documentation), so the gap is dormant rather than structurally absent. Alabama exempts a defined-BENEFIT plan specifically (r. 810-3-19-.04), not every pension or annuity, and this engine cannot distinguish a non-federal defined-benefit plan from a non-federal defined-contribution one even once that field is populated.

  Ala. Admin. Code r. 810-3-19-.01, "Exempt Retirement Allowances" — (b) federal civil-service retirement benefits, (c) Social Security · Ala. Admin. Code r. 810-3-19-.04, "Defined Benefit Plans" — the test excluding defined-contribution plans (TSP) from the exemption

• Every full-year-resident Form 40 filer may deduct their federal income tax liability from Alabama AGI in full, uncapped and independent of the standard/itemized election, under §§ 40-18-15 and 40-2A-7(a)(5) and Ala. Admin. Code r. 810-3-15-.20. `AL_RULE.applyTax` never read the engine's own `federalOrdinaryIncomeTax` field before this pack's companion fix, so this deduction was entirely missing and Alabama tax was overstated for every filer with federal tax liability.

  Tax years 2024, 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • r. 810-3-15-.20(2)(b)'s nonresident apportionment of the deduction by Alabama-source AGI, and the part-year proration the booklet separately requires. This engine models a full-year Alabama resident only, in every state, so "every filer" above means every filer of THAT baseline, not literally every filer Alabama recognizes.
  • The Form 40 worksheet's own instruction to ADD the Net Investment Income Tax to the deduction base, and r. 810-3-15-.20(1)(d)2.'s own instruction to SUBTRACT federal credits — neither NIIT nor any federal credit is computed by this calculator at all, so for every household it can represent, the engine's federal tax figure already equals what Alabama's own worksheet would compute; see `AL_RULE.notes` for the full disclosure.
  • The optional cash-basis election r. 810-3-15-.20(2)(a) permits; this model always uses the engine's own current-year federal tax figure, matching the Department's own DEFAULT of the accrual basis under r. 810-3-15-.20(3)(c)2.(ii).
  • The AMT, which the worksheet's own federal-tax figure would include but this calculator does not compute anywhere.

  Ala. Admin. Code r. 810-3-15-.20, "Federal Income Tax Deduction - Individuals" — (1)(d)2. the federal-credits netting instruction, (2)(a) the optional cash-basis election, (2)(b) nonresident apportionment, (3)(c)2.(ii) the Department's own accrual-basis default

• Code of Alabama § 40-18-19(a)(13), added by Act 2022-294 (HB162) and effective for tax years beginning on or after 1 January 2023, exempts the first $6,000 of a taxpayer's taxable retirement income once that taxpayer turns 65. Alabama Schedule RS Parts II/III test and cap this PER PERSON — each of "Primary" and "Spouse" gets an independent age-65 test and an independent $6,000 cap against that person's own taxable retirement income. `AL_RULE.applyTax` modelled no such exclusion at all before an independent adversarial review of this pack's first draft found the gap; every TSP-vehicle case in that draft used an age of 67, silently qualifying for an unmodelled exclusion.

  Tax years 2023, 2024, 2025, 2026 · 10 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The exclusion's reach into non-federal pension or annuity income and IRA/401(k)/403(b) distributions generally — Schedule RS Part II/III covers all of these, but AL_RULE.applyTax excludes pensionIncome from Alabama AGI outright (correct for the federal civil-service benefits r. 810-3-19-.01(b) exempts, but overbroad for a non-federal defined-contribution pension, which Schedule RS would NOT exempt and this $6,000 exclusion alone would not fully cover either). The engine's `nonFederalPensionIncomePersonA/B` field could in principle carry that distinction but is always zero from the web app's UI today (dormant, not structurally absent — see that field's own documentation), so the gap cannot currently bite.
  • A decedent's exclusion is tested on the age reached at death via `ageAtEndOfTaxableYear`, matching this engine's existing precedent for Missouri, Georgia, Ohio and Arizona's own bare (no-date-specified) age gates, since Schedule RS line 10 asks only "is the...taxpayer 65 or older" with no year-end qualifier of its own — this pack does not independently re-derive that precedent's own correctness, only that AL_RULE calls it.

  HB162 (2022 Regular Session), enrolled text — adds Code of Alabama § 40-18-19(a)(13), "the first six thousand dollars ($6,000) of taxable retirement income," available "only [to] individual taxpayers who are 65 years of age or older," effective 1 January 2023 · 2025 Alabama Schedule RS (Form 40 return package) — Parts II/III line 10, testing and capping the exclusion PER PERSON against that person's own "Retirement Income Taxable to Alabama" (line 9)

• The MO-1040 Tax Rate Chart is a DIFFERENT table each year — bracket width inflation-indexed under RSMo § 143.011.5 ($1,273 for 2024, $1,313 for 2025, $1,348 for 2026) and the top rate itself stepped down between years (4.80% for 2024, 4.70% for 2025-2026, no further trigger-step for 2026) — read from the 2024 and 2025 MO-1040 instructions' own published charts and, for 2026 (whose MO-1040 form does not exist yet), the DOR's own 2026 Withholding Tax Formula, whose worked example independently confirms the derived $263 cumulative base at $9,436.

  Tax years 2024, 2025, 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Any year beyond 2026 or before 2024 — held forward/backward from the nearest published year rather than projected, matching this rule's existing convention for the pension cap and standard deduction.
  • A further trigger-step reduction below 4.70% for a future year Missouri has not yet published; this pack certifies only that no such step has fired for 2026 as of this pack's certification date.

  2025 Form MO-1040 Instructions — the 2025 Tax Rate Chart and Tax Calculation Worksheet, p.21 · 2024 Tax Chart — the 2024 Tax Rate Chart, confirming the 4.80% top rate and $1,273 bracket width · 2026 Missouri Withholding Tax Formula — the only DOR-published source for 2026 bracket figures; its own worked example confirms the derived chart

• FERS and TSP together are ONE "public pension" under MO-A Part 3 Section A ("Pensions received from any federal, state, or local government"), sharing a single cap equal to that year's maximum Social Security benefit ($46,381 for 2024, $47,633 for 2025, $48,967 for 2026), read from the DOR's own Pension FAQ.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The Section B private pension deduction, which the model cannot apply because it cannot distinguish privately funded retirement income from public.
  • Any year beyond 2026 or before 2024 — held forward/backward from the nearest published figure rather than projected.

  Missouri DOR Pension FAQ — maximum public pension exemption by year · 2025 Form MO-A — Part 3, Section A (public pension exemption)

• Section A line 4 floors the capped-pension-minus-Social-Security difference at zero, and Section D adds Section C's taxable Social Security BACK — so the combined Part 3 exemption is max(capped public pension, that person's own taxable Social Security), never their sum, computed PER PERSON on a joint return rather than pooled.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Social Security DISABILITY, which Missouri's own statute carries no age limit for, but which this engine cannot distinguish from ordinary Social Security — a recipient under 62 with disability benefits is overcharged, already disclosed in MO_RULE.notes rather than fixed here.

  2025 Form MO-A — Part 3, Sections A, C and D, including line 4's floor-at-zero instruction and line D's "Add Line 5 (Section A), Line 9 (Section B), and Line 4 (Section C)"

• Section C's Social Security deduction requires the taxpayer be 62 by 31 December, RSMo § 143.125.1's OWN plain age test with no federal reference — distinct from line 14's federal § 63(f) age-65 test. MO-1040 withdraws nothing from a decedent (a per-person "Deceased in [year]" box sits beside the same age boxes a living filer uses), so a decedent's age at death is tested exactly as a living person's would be, on WHICHEVER of the two conventions the line in question uses.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The fuller boundary sweep (both gates at their own exact thresholds, for both a decedent and a living person born 1 January) — already certified by `state-taxes.test.ts`'s own dedicated, passing "#360" suite; this claim does not re-derive that precedent, only that MO_RULE calls it for the age-62 gate itself and the one day the two conventions disagree.

  2025 Form MO-1040 Instructions — the decedent instruction ("select the appropriate box... attach a copy of Form MO-1310D..."), page 6, and the per-person age/deceased boxes on MO-1040 page 1 · 2025 Form MO-A — Part 3 Section C, "you must be 62 years of age by December 31"

• Line 12 grades the federal-tax deduction by Missouri AGI ROUNDED to the whole dollar (35% at $25,000 or less, down to 0% above $125,000), and Line 13 caps the resulting deduction at $5,000 individual / $10,000 combined — read directly from the 2025 MO-1040 instructions' own Line 12/13 text.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MO-1040 lines 9-11's own Schedule 2/3 additions to the federal-tax base, which this engine's federal tax figure may not fully replicate — the existing disclosed bound in `MO_RULE.notes`.

  2025 Form MO-1040 Instructions — Line 12 (federal tax percentage table) and Line 13 (federal income tax deduction, capped $5,000/$10,000), and the "Dollars and Cents" rounding instruction

• Line 14's standard deduction is "the standard deduction amount for your filing status" from the federal return under RSMo § 143.131 — which EXCLUDES the separate federal 65+ senior deduction Congress added for 2025-2028 — and a qualifying widow(er) takes the JOINT figure plus the $1,400 line-15 exemption despite filing alone.

  Tax years 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household's own standard deduction and $1,400 exemption — unreachable, since the engine has no head-of-household-equivalent filing status.

  2025 Form MO-1040 Instructions — Line 14 ("enter the standard deduction amount for your filing status") and Line 15 (head of household and qualifying widow(er) $1,400 exemption)

• A joint return apportions COMBINED taxable income by each spouse's share of Missouri AGI (Line 7, rounded to the nearest whole percent), then computes and rounds a SEPARATE tax for each share against the chart — not one tax on the pooled total. A single filer has no second person to apportion against.

  Tax years 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The narrow model choice for who absorbs the rounding remainder at an exact half-percent split (person A rounds, person B takes the remainder) — a design choice this claim does not independently exercise.

  2025 Form MO-1040 Instructions — Line 7 ("The total entered on Line 7 must equal 100 percent - round to the nearest percentage") and "A separate tax must be computed for you and your spouse"

• Colorado taxable income starts from federal TAXABLE income (DR 0104 line 1, i.e. federal AGI less the federal standard deduction and any enhanced senior deduction) — NOT federal AGI and NOT a sum of the engine's own income categories, which is what the pre-#301 `CO_RULE` used and which overstated Colorado tax by the federal standard deduction times 4.4% for every household.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The engine's own simplified federal model that produces this figure — no itemized deductions, no QBI, no above-the-line adjustments beyond what the projection computes; the same disclosed bound every other field sourced from the federal calculator already carries.
  • Colorado's own DR 0104 addition lines 2-9 (e.g. a state tax refund addback, a SALT Parity Election addback) — narrow items not modelled, uncommon for a FERS retiree.
  • Exactness below $50,000 of taxable income. DR 0104 line 13 directs a full-year resident to Colorado's own published tax TABLE for taxable income under $50,000, whose $100-wide bands round each band's MIDPOINT to the nearest dollar; only the "Worksheet for taxable incomes over $50,000" multiplies the rate directly. This rule always multiplies directly, the same simplification this program's own California certification discloses for a larger rounded-table range — a filed Colorado return under $50,000 can differ from this model's figure by a small, bounded amount.

  2025 Book 104 (DR 0104 booklet) — DR 0104 line 1 ("Enter your federal taxable income from your federal income tax form 1040")

• The rate is YEAR-SENSITIVE: 4.25% for tax year 2024 alone (SB24-228, a one-year-only TABOR refund), and 4.4% for 2025 and 2026 (the Proposition 121 baseline, confirmed unmoved by the 2025-2035 TABOR reduction mechanism for either of those two years as actually filed) — confirmed directly against the actual 2024 and 2025 DR 0104 booklets and the 2026 DR 1098 withholding worksheet as filed, not assumed from a single year's check generalized backward or forward.

  Tax years 2024, 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A TABOR-triggered rate reduction for a future year Colorado has not yet published; this claim certifies only that no such reduction appears on the actual 2024-2026 forms as filed.

  2024 Book 104 (DR 0104 booklet) — DR 0104 line 13 ("The income tax rate is currently 4.25%") and the 2024 Colorado Income Tax Table/worksheet, both headed "4.25%" · 2025 Book 104 (DR 0104 booklet) — DR 0104 line 13 ("The income tax rate is currently 4.4%") · DR 1098 (2026 Colorado Withholding Worksheet for Employees) — "4.40% PIT" · SB24-228 (2024) — the enacted one-year TABOR reduction of the 2024 rate from 4.40% to 4.25%, signed 14 May 2024

• The Social Security subtraction is computed PER PERSON: age 65+ subtracts the full federally taxable benefit with no dollar cap (DR 0104AD line 3, and the cap-raise mechanic of § 39-22-104(4)(f)(III)(B)); age 55-64 subtracts the full benefit if household federal AGI does not exceed $75,000 (single) / $95,000 (joint), otherwise capped at $20,000 — and did NOT EXIST AT ALL for anyone under 65 before tax year 2025 (HB24-1142).

  Tax years 2024, 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The Social Security survivor-benefit subtraction available to a person under 55 "due to the death of the person originally entitled to the benefits" — a status this engine's breakdown does not carry.
  • Whether the $75,000/$95,000/$20,000 figures are inflation-indexed for a future year; no indexing provision was found and none is assumed.

  2025 Book 104 (DR 0104 booklet) — DR 0104AD line 3/5 instructions (age and AGI-gated Social Security subtraction) · HB24-1142 (2024) — expanded the Social Security subtraction to ages 55-64, AGI-gated, for tax years commencing on or after 1 January 2025

• A `surviving_spouse` household reads the $95,000 (joint) Social Security AGI threshold, the same as `mfj` — a DISCLOSED DEFAULT rather than a form citation. DR 0104AD line 3 names only "single" and "married filing jointly"; no Colorado instruction naming surviving-spouse treatment for this specific threshold was found. Colorado requires filing under the same status as the federal return, and a federal qualifying surviving spouse uses MFJ brackets and the MFJ standard deduction, which this engine's own federal calculator already treats as MFJ-equivalent.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • This is NOT sourced to an explicit Colorado instruction naming surviving-spouse treatment for the $75,000/$95,000 split — unlike Arizona, whose Form 140 explicitly assigns a qualifying surviving spouse to a DIFFERENT bracket (head of household) for its own standard deduction. Revisit if Colorado publishes clearer guidance.

  2025 Book 104 (DR 0104 booklet) — p.6, "You must file using the same filing status on both your federal and Colorado income tax returns", and DR 0104AD line 3 (naming only "single" and "married filing jointly")

• The pension/annuity/TSP subtraction (DR 0104AD lines 4/6) is computed PER PERSON, sharing ONE ceiling with that SAME person's own Social Security subtraction: $24,000 (age 65+) or $20,000 (age 55-64) MINUS Social Security already claimed, floored at zero, then capped at that person's own taxable pension/annuity/TSP income — NOT a household-pooled figure, which is what the pre-#301 rule computed and is the central defect this rebuild fixes.

  Tax years 2024, 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The pension/annuity survivor subtraction available to a person under 55 who received the income as a secondary beneficiary due to the death of the person who earned it — a status this engine's breakdown does not carry.

  2025 Book 104 (DR 0104 booklet) — DR 0104AD line 4/6 instructions ("Pension/annuity income should not be intermingled between spouses. Each spouse must meet the requirements for the subtraction separately") · DR 0104AD (rev. 09/15/25) — the Subtractions from Income Schedule itself, showing separate Taxpayer (line 4) and Spouse (line 6) pension/annuity lines

• A decedent's final Colorado return reads their PLAIN age at death for both the Social Security and pension/annuity age gates — Colorado's own instruction is a plain "as of December 31" test with no reference to any federal age convention, and a decedent's taxable year ends at death. The same convention Maryland, Missouri, Georgia and Ohio use for their own age-65 gates.

  Tax years 2024, 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A decedent under 55 at death — this pack does not separately certify that boundary, since no Colorado subtraction is available at that age for a non-survivor either way.

  2025 Book 104 (DR 0104 booklet) — DR 0104AD line 3/4 ("If you were 65 years of age or older on December 31, 2025 ...") and the Deceased Taxpayer instructions

• The NJ-1040 Line 28a pension/annuity/IRA-withdrawal exclusion is available ONLY to a person who was age 62 or older, or blind/disabled as defined by Social Security guidelines, on the last day of the tax year. A household of any other age gets NO exclusion at all — the pre-#301 rule granted the exclusion regardless of age, so long as household total income cleared the dollar threshold.

  Tax years 2024, 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Blind/disabled as an alternative qualifying condition — this engine's breakdown carries no disability status anywhere, matching the disclosed pattern for every other state's age-based retirement exclusion.

  2025 Form NJ-1040 instructions, p.20 — Line 28a: "You (and/or your spouse if filing jointly) were age 62 or older or blind/disabled as defined by Social Security guidelines on the last day of the tax year"

• On a joint return where only one spouse is 62+/disabled, ONLY that spouse's own pension/annuity/IRA income qualifies for the exclusion — the other spouse's retirement income, however large, does not qualify at all and is never reduced by it. The pre-#301 rule pooled both persons' pension and TSP into one household figure unconditionally.

  Tax years 2024, 2025 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A household where BOTH spouses independently qualify (both 62+) — a simple sum of two qualifying amounts, exercised by this pack's other claims (e.g. the tier-2/tier-3 percentage cases) rather than by a dedicated case here, since the interesting question this claim isolates is what happens when they DIFFER.

  2025 Form NJ-1040 instructions, p.20 — Line 28a, Line A: "Joint Filers: If only one spouse is 62 or older or disabled, enter only the pension income of that spouse. You cannot exclude the pension income of the spouse who is younger than 62" · GIT-1 & 2 (January 2026), p.21 — worked example: A (64, $10,000 pension) and B (61, $8,000 pension), combined total income $127,000; "Income Qualified for Pension Exclusion $10,000"

• Single, Head of Household and Qualifying widow(er) filers share the SAME three total-income bands as MFJ ($0-$100,000 / $100,001-$125,000 / $125,001-$150,000) — NOT the pre-#301 rule's scaled-down $0-$50,000/$50,000-$62,500/$62,500-$75,000 bands. Within the first band their cap is $75,000, not MFJ's $100,000 and not the pre-#301 rule's $50,000 (which is actually Married-Filing-Separately's own figure).

  Tax years 2024, 2025 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Married Filing Separately's own $50,000/$25,000/$12,500 row — unreachable, since this engine has no MFS status.

  2025 Form NJ-1040 instructions, p.20 — Line 28a chart: "$0-$100,000 / Single, Head of household, Qualifying widow(er) ... $75,000" · 2024 Form NJ-1040 instructions, p.20 — the identical chart, confirmed independently rather than assumed from the 2025 booklet alone

• In the $100,001-$125,000 and $125,001-$150,000 total-income bands, the exclusion is a PERCENTAGE (50%/25% for MFJ, 37.5%/18.75% for Single/HoH/QW) of the qualifying person's ACTUAL pension/TSP income — not, as the pre-#301 rule computed, the lesser of a flat dollar figure and actual income. Because the percentage is always at most 100%, this is mathematically the same as the form's own "lesser of Line A or Line B" instruction once Line B is read correctly as a percentage of Line A, so no separate cap is needed once the base is right.

  Tax years 2024, 2025 · 7 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether the percentage-of-actual figure can ever exceed the filing status's own dollar cap for an unusually large pension — not separately tested, since qualifying pension income cannot exceed total household income, which is itself bounded by each band's own $125,000/$150,000 ceiling, making this a mathematical certainty rather than a case-worthy boundary.

  GIT-1 & 2 (January 2026), p.20 — worked example: combined total income $114,000, qualifying pension $6,200 (both spouses 65+), "Maximum Pension Exclusion $3,100" (exactly 50% of $6,200) · 2025 Form NJ-1040 instructions, p.20 — Line 28a chart: "$100,001-$125,000 / Married ... 50% of line 20a" and "$125,001-$150,000 / Married ... 25% of line 20a" (Line A, already age-filtered, per the chart's own cross-reference)

• Once household total income (NJ-1040 Line 27) exceeds $150,000, the pension exclusion is $0 regardless of age, filing status or how much qualifying pension income exists — a hard cliff on TOTAL INCOME, not on age-qualification.

  Tax years 2024, 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A surviving-spouse or single household above $150,000 — this claim's own cases use MFJ; the cliff's existence does not depend on filing status, and the bracket/cap split those statuses raise is certified separately below.

  2025 Form NJ-1040 instructions, p.20 — Line 28a: "Your income on line 27 is $150,000 or less" as a qualifying condition, and the chart's own $125,001-$150,000 ceiling

• A `surviving_spouse` household uses the JOINT bracket table (NJ Tax Rate Schedules Table B, which groups "Qualifying widow(er)/surviving CU partner" with MFJ and Head of Household) for the tax RATE, but the SINGLE/HoH/QW row's $75,000 cap (NOT MFJ's $100,000) for the Line 28a exclusion — two DIFFERENT filing-status groupings on the same return, which a single `isSingle` boolean cannot express at once.

  Tax years 2024, 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The $100,001-$150,000 percentage tiers for a surviving-spouse household specifically — this claim's own cases exercise only the tier-1 cap and the bracket-table split; the percentage-tier claim above certifies the 37.5%/18.75% mechanic against MFJ and Single, and a surviving spouse shares Single's percentage row by the same grouping this claim already establishes, but that combination is not independently exercised by its own case.

  2025 Form NJ-1040 instructions, p.63 — Tax Rate Schedules, Table B: "Married/CU couple, filing joint return; Head of household; Qualifying widow(er)/surviving CU partner" · 2025 Form NJ-1040 instructions, p.20 — Line 28a chart: "Single / Head of household / Qualifying widow(er) ... $75,000" — the same row, for the exclusion cap alone

• Social Security benefits are entirely outside New Jersey gross income — never reported on Line 27 "Total Income" at all, not merely excluded from the taxable base. The graduated bracket schedule (1.4% to 10.75%, MFJ's breakpoints wider than Single's at several steps) was checked directly against BOTH the 2024 and 2025 Tax Rate Schedules and found identical.

  Tax years 2024, 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • 2026. No NJ-1040 booklet has been published yet for tax year 2026; this rule holds the 2025 bracket and exclusion figures forward as a disclosed projection.

  NJ Division of Taxation, "NJ Income Tax – Retirement Income" (njit6.shtml) — "Social Security and Railroad Retirement benefits ... are not taxable and should not be reported as pension income" · 2024 and 2025 Form NJ-1040 instructions, p.63 in each — Tax Rate Schedules Table A and Table B, confirmed byte-identical across both years

• H.4216 (Act 110 of 2026) added S.C. Code §12-6-50(21), which specifically does NOT adopt IRC §63(b)-(g) (the federal standard/itemized deduction) — so, starting tax year 2026, the SC1040 no longer starts from federal taxable income the way the pre-2026 form (and the prior rule) did. SCDOR's own H.4216 announcement states federal AGI is now the starting point.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The FULL composition of federal AGI beyond wages, pension, TSP and taxable Social Security — this engine's federal AGI figure carries the same simplified-income-model gap as every other certified state (no interest, dividends, capital gains, business income or Roth-conversion income), so this claim proves the RULE reads the right binding, not that the binding itself is complete.

  H.4216 (Act 110 of 2026), Section 2 — amending S.C. Code §12-6-50 to add "(21) Section 63(b) through (g) ... specifically not allowed or adopted" · SCDOR, "Information about H. 4216" — "Federal Adjusted Gross Income (AGI) is now the starting point for the South Carolina Individual Income Tax return"

• S.C. Code §12-6-1120 computes South Carolina gross income "without application of [IRC] Sections 78, 86, and 87" — IRC §86 is the federal rule taxing a portion of Social Security, so South Carolina exempts it in full regardless of federal taxability, unaffected by H.4216. Social Security counts toward federal AGI (and so toward the SCIAD phase-out) but never enters SC taxable income.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Railroad retirement benefits, which the same S.C. Code §12-6-1120 exemption also covers — this calculator has no railroad-retirement income category to exercise it with.

  S.C. Code Ann. §12-6-1120 — "South Carolina gross income is determined without application of Internal Revenue Code Sections 78, 86, and 87"

• S.C. Code §12-6-1170(A), untouched by H.4216: an individual taxpayer who is the original owner of a qualified retirement account may deduct up to $3,000/year of their OWN qualifying retirement income under 65, or up to $10,000/year at 65 or older — NOT the $15,000 the prior rule used at 65+ (which conflated this deduction with the separate age-65 deduction below), and NOT a household-pooled figure — each person's deduction is capped by THEIR OWN qualifying income only.

  Tax years 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The §12-6-1170(A) surviving-spouse retirement deduction for a deceased spouse's own separate retirement account (SC1040 lines p-3/p-6) — distinct from the ordinary per-person deduction modeled here.

  S.C. Code Ann. §12-6-1170(A) (via Justia) — "$3,000 of retirement income received" under 65; "not more than ten thousand dollars of retirement income" beginning the year the taxpayer reaches 65 · 2025 SC1040 Instructions, p.8-9 — "Worksheet for taxpayer (line p-1)" and "Worksheet for spouse (line p-2)", each testing only that person's own qualified retirement income

• S.C. Code §12-6-1170(B), untouched by H.4216: beginning the year a resident taxpayer turns 65, they get a SEPARATE $15,000 deduction against ANY South Carolina income (not just retirement income), reduced by whatever that same person claimed under (A). SC Revenue Ruling 22-11 confirms this applies against any income on the return with no per-person income ceiling of its own.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The exact $10,000 figure in claim `sc-retirement-deduction-3000-under-65-10000-65plus-per-person` cannot be independently distinguished from any OTHER cap value up to $15,000 by testing this claim's own combined total alone, since (A)+(B) is a mathematical constant at $15,000 for any such cap — see the pack's own header note and `check-certification-mutations.ts` for the resulting one-directional mutation coverage.

  S.C. Code Ann. §12-6-1170(B) (via Justia) — "an amount not to exceed fifteen thousand dollars reduced by any amount the taxpayer deducts pursuant to subsection (A)" · SC Revenue Ruling 22-11 (2022-11-17), p.61-62 — surviving-spouse worked example: "$25,000 in his own qualified retirement income, his general retirement income deduction is the maximum $10,000 ... the husband's general age 65 and older deduction is $5,000 ($15,000 less husband's $10,000 general retirement income deduction). The $5,000 deduction can offset any remaining taxable income on the return"

• H.4216 §3 (S.C. Code §12-6-1140(15)) creates the South Carolina Income Adjusted Deduction: $15,000 for single/MFS, $30,000 for MFJ or surviving spouse, phased out over a $55,000 (single) or $110,000 (MFJ) range starting at $40,000 or $80,000 of federal AGI respectively, fully gone at $95,000/$190,000. The reduction amount itself rounds DOWN to the next $10 before being subtracted.

  Tax years 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of Household's own $22,500 tier, phased out over $60,000-$142,500 — unreachable, since this engine has no HoH filing status.

  H.4216 (Act 110 of 2026), Section 3 — S.C. Code §12-6-1140(15)(a)-(c): base amounts by filing status, the phase-out fractions, and "(c) Any reduction amount which is not a multiplier of ten dollars must be rounded to the next lowest ten dollars" · SCDOR, "Information about H. 4216" — confirms the $15,000/$22,500/$30,000 base amounts by filing status

• H.4216 §1 (S.C. Code §12-6-510(C)(1)), for tax years beginning after 2025: 1.99% on the first $30,000 of taxable income, 5.21% of the full amount minus $966 above it — algebraically the same two-bracket marginal schedule, not a cliff. §12-6-510(C)(1) supplies ONE table with no separate MFJ column, so the $30,000 break and $966 constant are per return, not doubled for MFJ — the pre-2026 law's own tax table said as much explicitly ("Use this Tax Rate Schedule for any filing status claimed on the SC1040"), and nothing in H.4216 creates a second, doubled table for 2026. Replaces the prior rule's six-bracket "0/3/4/5/6/6.2%" ladder, which matched no actual year of South Carolina law.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • 2027-and-later rates, which §12-6-510(C)(2)-(3) makes conditional on an annual revenue-growth trigger the Board of Economic Advisors determines, not a fixed schedule — this rule holds 2026's rate flat forward as a disclosed projection.
  • The exact $30,000 boundary itself is unobservable by output alone: the two formulas agree there by construction (both give exactly $597), so a `<=`/`<` mutation at the boundary cannot be caught by any certified case.

  H.4216 (Act 110 of 2026), Section 1 — S.C. Code §12-6-510(C)(1): "$0 ... $30,000 ... 1.99% times the amount; $30,000 ... or more ... 5.21% times the amount minus $966" · SCDOR, "Information about H. 4216" — "The tax rate for income less than $30,000 is 1.99%. The tax rate for income from $30,000 and above is 5.21%, minus $966"

• For the §12-6-1170(A)/(B) deductions, a decedent "continues to age" — the deduction amount is measured by the age the decedent WOULD HAVE reached at 31 December of the tax year had they lived, NOT their age at the actual date of death. This is the OPPOSITE of the default final-return convention (`ageAtEndOfTaxableYear`) most other jurisdictions this program has certified use; South Carolina deliberately does not call it.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A surviving spouse's OWN separate §12-6-1170(A) deduction for a deceased spouse's retirement account (SC1040 lines p-3/p-6), which uses the same "age had they lived" convention but is itself unmodelled — see the retirement-deduction claim above.

  SC Revenue Ruling 22-11 (2022-11-17), "Important Points to Remember and Assumptions Used in Examples" — "The age of a deceased individual had he or she been living on December 31st of the current tax year is used to calculate the deduction amounts (i.e., the deceased continues to age in the current tax year and thereafter, without regard to the actual date and year of his or her death)"; repeated in the ruling's own worked examples, which reject an age-at-death figure in favor of the year-end one

• Social Security benefits are subtracted from AGI on Schedule 1, not reported on Form 4884, and this rule omits them from every route's base regardless of which retirement-deduction election applies — unaffected by PA 4 of 2023 or 2025 PA 24.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The full composition of federal AGI beyond wages, pension and TSP — this rule's income model carries the same simplified-income-model gap as every other certified state.

  2025 Form 4884 Instructions — "Social Security benefits ... are subtracted from AGI and should be reported on the Schedule 1, line 11 or line 14, not on Form 4884"

• Michigan Treasury Letter Ruling 2025-1 holds that a Thrift Savings Plan distribution is not a "retirement or pension benefit" under MCL 206.30(8)(d) — the TSP "does not prescribe retirement age or years of service" and "allows the employee to set the amount of compensation to be deferred," the same exclusion that applies to 457/401(k)/403(b) deferred-compensation plans, with no carve-out for the FERS automatic 1%/matching agency contributions the way a 401(k) gets one under MCL 206.30(1)(a)(iv). TSP is fully taxable at 4.25% regardless of birth year, age, or which retirement-deduction route otherwise applies.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A non-federal spouse's 401(k)/IRA, which arrives in this engine's TSP-shaped bucket but is not a federal TSP — Letter Ruling 2025-1 addresses the TSP specifically, and a 401(k)'s own MCL 206.30(1)(a)(iv) carve-out for employer/mandated-employee contributions is not modelled here.

  Michigan Treasury Letter Ruling 2025-1 — "The distributions from your federal TSP in 2023 and future years are not a 'retirement or pension benefit' under the MITA. ... they are not eligible for subtraction under MCL 206.30(1)(f)"

• MCL 206.30(9)(a): a filer born before 1946 has no additional restriction on public retirement benefits, and a private-source pension is capped at the inflation-adjusted maximum under MCL 206.30(1)(f)(iv) — $67,610 single/$135,220 MFJ for 2026, not 2025's $65,897/$131,794.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The public/private split itself: MCL 206.30(9)(a) leaves PUBLIC (federal) pension unlimited and caps only PRIVATE pension. This rule still caps the WHOLE pension at the private maximum, understating a Tier 1 filer's deduction — fixing it needs `nonFederalPensionIncomePersonA/B`, which no branch of this rule reads.

  MCL 206.30(9)(a) — "For a person born before 1946, this subsection provides no additional restrictions or limitations under subsection (1)(f)" · Form 446 (Rev. 2026), 2026 Michigan Income Tax Withholding Guide — "recipients born before 1946 ... may subtract qualifying private pension and retirement benefits up to the remaining balance of $67,610 if single or married filing separately, or $135,220 if married and filing a joint return"

• MCL 206.30(10)(d): "For the 2026 tax year and each tax year after 2026, a taxpayer may deduct retirement or pension benefits as provided under subsection (1)(f), except that the amounts deductible under subsection (1)(f)(i) and (ii) combined are subject to the same maximum amounts allowed under subsection (1)(f)(iv)" — the SAME $67,610/$135,220 cap Tier 1's private pension uses, now available to combined public+private pension for EVERY filer born 1946 or later, with no age gate and no upper birth-year bound. This retires the pre-2026 4-tier structure's Tier 3 (MCL 206.30(10)(c)'s 75%-of-cap figure, itself only 2025's phase-in percentage — 25% for 2023, 50% for 2024, per MCL 206.30(10)(a)-(b)) and Tier 4 (no deduction at all, MCL 206.30(9)(e)) for the deduction's SIZE, though Tier 2/3's own standard-deduction alternative survives as a separate election — see the standard-deduction claim below.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MCL 206.30(9)(c)/(9)(d)/(11): the SSA-exempt-employment bonus tracks and the public-safety (fire/police/corrections) unlimited deduction, which this engine cannot model without an SSA-exempt-service or qualifying-public-safety-service input.
  • Reciprocal public pensions from another state (MCL 206.30(1)(f)(ii)).

  MCL 206.30(10)(d) — "For the 2026 tax year and each tax year after 2026, a taxpayer may deduct retirement or pension benefits as provided under subsection (1)(f), except that the amounts deductible under subsection (1)(f)(i) and (ii) combined are subject to the same maximum amounts allowed under subsection (1)(f)(iv) ... for that same tax year" · Revenue Administrative Bulletin 2026-1, Issue 3 — "for the final phase-in year of 2026, and each tax year thereafter, subtractions of retirement income from these public sources are limited to the private retirement maximum under subsection 30(1)(f)(iv) of the MITA (except for taxpayers born before 1946 ...)"

• Once the key person reaches 67, MCL 206.30(9)(b) (born 1946-1952) or MCL 206.30(9)(e) (born 1953+) makes a $20,000/$40,000 standard deduction against ALL income available as an ALTERNATIVE to the subsection (10) pension-cap route — Michigan's own "Which Section of Form 4884 Should I Complete?" questionnaire repeatedly instructs a filer to compute both and "claim the most beneficial subtraction." This rule takes the minimum taxable-income result of whichever routes apply, matching Form 4884's own election. Under 67, no standard-deduction alternative exists at all.

  Tax years 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MCL 206.30(9)(e)'s post-2028/pre-2026 coordination, which forbids the Social Security subtraction alongside the standard deduction — see the PA 24 window claim below.

  Revenue Administrative Bulletin 2026-1, Example J — reduces $35,000 AGI ($18,000 wages + $5,000 pension + $12,000 taxable Social Security) to exactly $3,000 taxable income via the standard-deduction route for tax year 2026 · 2025 Form 4884 Instructions, "Which Section of Form 4884 Should I Complete?" — repeatedly instructs completing a worksheet and Form 4884 section together and to "claim the most beneficial subtraction"

• MCL 206.30(9)(b) (born 1946-1952) claims the $20,000/$40,000 standard deduction IN FULL ALONGSIDE the personal exemption, permanently, with no sunset. MCL 206.30(9)(e) (born 1953+), for tax years 2026-2028 only, instead forgoes the personal exemption when taking the same standard deduction — a full personal-exemption's difference in taxable income between two households one birth year apart on either side of the boundary.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MCL 206.30(9)(c)'s own, narrower $35,000/$55,000/$70,000 SSA-exempt-employment enhancement to the 1946-1952 cohort's standard deduction — needs an SSA-exempt-service input this engine does not collect.

  MCL 206.30(9)(b) — "for a person born in 1946 through 1952 ... that person is eligible for a deduction of $20,000.00 for a single return and $40,000.00 for a joint return, which deduction is available against all types of income" (no personal-exemption coordination stated) · MCL 206.30(9)(e) — "For tax years that begin on and after January 1, 2026 and before January 1, 2029, if a person takes the deduction of $20,000.00 for a single return or $40,000.00 for a joint return, that person shall not take the personal exemption under subsection (2)"

• 2025 PA 24's enhancement to MCL 206.30(9)(e) — claiming the standard deduction WITHOUT forgoing the Social Security subtraction — applies only to "tax years that begin on and after January 1, 2026 and before January 1, 2029." Outside that window (2029 onward, in this rule's scope), MCL 206.30(9)(e) instead forbids the Social Security subtraction alongside the standard deduction, a trade-off this rule cannot evaluate without a taxable-Social-Security input no other part of it reads, so projections outside 2026-2028 fall back to the subsection (10) pension-cap route only.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The actual post-2028 mechanic itself (SS-subtraction-vs-standard-deduction coordination) — disclosed as unmodelled, not approximated.
  • Tax years before 2026, where the pre-PA-24 MCL 206.30(9)(e) coordination (reducing the std-deduction pool by personal exemption, Social Security, and military/railroad/Guard deductions) also applies and is equally unmodelled.

  MCL 206.30(9)(e) — "For tax years that begin on and after January 1, 2026 and before January 1, 2029, if a person takes the deduction ... that person shall not take the personal exemption ..." (the enhancement's own date bounds) · Revenue Administrative Bulletin 2026-1, Issue 11 — "For tax years 2026 through 2028, these taxpayers who have Social Security income included in AGI may subtract both the Social Security income and a full standard deduction. ... For tax years prior to 2026 and after 2028, their standard deduction must be reduced by the personal exemption and any deductions claimed for taxable Social Security ..."

• MCL 206.30(9)(f) and (10)(e): "the limitations and restrictions in this subsection shall be applied based on the date of birth of the older spouse filing the joint return" — confirmed independently by Form 446's Withholding Guide ("age is determined using the age of the older spouse"). Every birth-year and age test in this rule reads whichever of Person A/B has the earlier full DATE of birth, not Person A unconditionally and not merely the earlier calendar YEAR — a same-birth-year tie broken by year alone can pick the wrong person once one of them is a final-return decedent, whose age is their age at death rather than a year-end age (found by an independent adversarial review; see the certification pack's own case).

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A later-year surviving spouse electing to use a DECEASED spouse's older birth year (MCL 206.30(9)(f)/(10)(e)) instead of their own — conditioned on inputs (a joint retirement/SS subtraction claimed in the year of death; no remarriage) this engine does not track.

  MCL 206.30(9)(f) — "For a joint return, the limitations and restrictions in this subsection shall be applied based on the date of birth of the older spouse filing the joint return" · Form 446 (Rev. 2026), 2026 Michigan Income Tax Withholding Guide — "For married couples that file a joint Michigan income tax return, age is determined using the age of the older spouse"

• For the tax year of a spouse's death, that decedent's birth year is entered on Form 4884 like any other filer's — Form 4884's "surviving spouse" birth-year carve-out (Part 2) explicitly applies only to a spouse who died PRIOR to the current tax year. This rule reads `finalReturnDecedentPersonA/B.birthDate` for a person who died this year, rather than the `-1`/`null` sentinels that would otherwise reconstruct a birth year of `projectionYear + 1` and deny the household any deduction tied to the decedent's actual birth year.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • A decedent's age (as opposed to birth year) for the 67-test uses `finalReturnDecedent.ageAtDeath` — this codebase's default final-return convention. Unlike South Carolina's own SC Revenue Ruling 22-11, nothing in MCL 206.30 or Form 4884's instructions was found specifying whether a decedent instead "continues to age" through 31 December for this purpose, so the default convention is used rather than independently confirmed for Michigan.

  2025 Form 4884 Instructions — "When completing Form 4884, surviving spouse means the deceased spouse died prior to the current tax year (e.g., when filing a 2025 return the spouse died in 2024 or prior). Deceased spouse benefits do not include benefits from a spouse who died in 2025"

• The 2026 personal exemption is $5,900 single / $11,800 MFJ, not 2025's $5,800/$11,600 — Michigan indexes this figure to inflation annually.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Years beyond 2026: this rule holds the $5,900/$11,800 figure flat forward as a disclosed projection, the same "held forward, not projected" convention this program uses for every other inflation-indexed state figure.

  Form 446 (Rev. 2026), 2026 Michigan Income Tax Withholding Guide — "Withholding Rate: 4.25% Personal Exemption Amount: $5,900"

• Schedule 511-A line A2: "Social Security benefits that are included in the Federal AGI shall be subtracted." This rule takes no Social Security parameter at all, so SS has no code path into the taxable base.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The full composition of federal AGI beyond wages, pension and TSP — this rule's income model carries the same simplified-income-model gap as every other certified state.

  2025 Form 511 Packet, Schedule 511-A Instructions, line A2 — "Social Security benefits that are included in the Federal AGI shall be subtracted. Provide a copy of your federal return."

• HB 2764 (68 O.S. § 2355), effective for TY2026 and after, replaces the prior six-bracket 0.25%-4.75% schedule with a four-tier 0%/2.5%/3.5%/4.5% schedule, and gives MFJ/surviving-spouse filers thresholds exactly DOUBLE the single/MFS ones ($7,500/$9,800/$14,400 vs $3,750/$4,900/$7,200) — a structural change from the old law's same-thresholds-for-both-statuses design this rule used to encode unconditionally.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household as a distinct filing status, which shares HB 2764's doubled bracket set with MFJ/surviving spouse but which this engine does not model.

  Oklahoma Tax Commission, "Summary of 2025 Tax Legislation," HB 2764 — TY2026 bracket tables for both "Single & Married Separate Filers" ($0-$3,750 0.0%, $3,751-$4,900 2.5% plus $0.00, $4,901-$7,200 3.5% plus $28.75, $7,201+ 4.5% plus $109.25) and "Married Joint, Head of Household & Surviving Spouse Filers" ($0-$7,500 0.0%, $7,501-$9,800 2.5% plus $0.00, $9,801-$14,400 3.5% plus $57.50, $14,401+ 4.5% plus $218.50)

• TY2025 and earlier keep the old six-bracket schedule (HB 2764 is effective "for tax year 2026 and subsequent tax years," not retroactively), with MFJ/surviving-spouse thresholds EXACTLY DOUBLE the single ones — confirmed directly against the 2025 Form 511 packet's own printed tax table, which prints separate "Single or married filing separate" and "Married filing joint or head of household" columns charging different tax at the same income. 2027 and later years hold the 2026 schedule forward as a disclosed projection, since 62 O.S. § 34.103 lets the State Board of Equalization trigger a further 0.25% cut on a revenue test evaluated each December, and no year past 2026 has a certified rate yet.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Any further rate cut 62 O.S. § 34.103 may actually trigger for 2027 or later — this rule cannot know a future December's revenue-growth certification in advance, so it holds 2026's rates forward rather than guessing.
  • Years before 2024: an independent adversarial review found the OTC's own historical rate material shows a different top-bracket breakpoint for 2022-2023 than for 2024-2025, which this rule does not distinguish — a projection for those older years is out of scope.

  Oklahoma Tax Commission, "Summary of 2025 Tax Legislation" — "HB 2764 ... Effective November 1, 2025 ... Individual income tax rate reduction, effective for tax year 2026 and subsequent tax years. (68 O.S. § 2355) ... individual income tax rates can be reduced by 0.25% across all brackets when certain revenue conditions are met. The State Board of Equalization evaluates this each December and makes final certification in February. (62 O.S. § 34.103)"

• Oklahoma's standard deduction is $6,350 single/MFS and $12,700 MFJ/surviving spouse, frozen at 2017 pre-TCJA federal levels by statute (HB 2348) and NOT tied to the current federal figure. The prior rule's $15,000/$30,000 figures, labeled "mirrors federal 2025," were actually federal 2025's own standard deduction — an unverified assumption never checked against Form 511, overstating the deduction by more than double.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household's own $9,350 standard deduction, since this engine does not model HoH as a distinct filing status.
  • Oklahoma itemized deductions (Schedule 511-D) as an alternative to the standard deduction.

  2025 Form 511 Packet, "Standard Deduction" instructions — "If your filing status is 'single' or 'married filing separate,' your Oklahoma standard deduction is $6,350. ... If your filing status is 'married filing joint' or 'qualifying surviving spouse,' your Oklahoma standard deduction is $12,700."

• 68 O.S. § 2358 / Form 511 line 11: "Oklahoma allows $1,000 for each exemption claimed."

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The "Special" additional $1,000/person exemption for a taxpayer or spouse 65+ — see the retirement-exclusion claim below.

  2025 Form 511 Packet, Select Line Instructions, line 11 — "Oklahoma allows $1,000 for each exemption claimed on the top of the return."

• The $10,000 retirement-income exclusion (Schedule 511-A lines A5/A6) caps against each INDIVIDUAL's own qualifying retirement income — line A5's own note requires the income be "in your name." The prior rule pooled the household's entire pension + TSP into one figure and capped it at $10,000 times a head count, so a household where all the retirement income belonged to one spouse wrongly received two $10,000 exclusions instead of one.

  Tax years 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The FERS Annuity Supplement's separate eligibility for the CSRS 100% in-lieu-of-Social-Security exclusion — see the FERS-vs-CSRS claim below.

  2025 Form 511 Packet, Schedule 511-A Instructions, line A5 — "Each individual may exclude their retirement benefits up to $10,000 ... (To be eligible, you must have retirement income in your name.)"; line A6 — "the amount of the exclusion on this line cannot exceed $10,000 minus the amounts already claimed on Schedule 511-A, line 5"

• Pension and TSP distributions share ONE $10,000 combined cap per person, not an independent $10,000 for each income stream, per Schedule 511-A line A6's own coordination against line A5.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Which of the two coordinated lines (A5 government/civil-service retirement, A6 other retirement income) a given dollar of pension or TSP is deemed to occupy — this rule applies the combined $10,000 cap directly against the summed figure rather than allocating a specific dollar to a specific line, which the combined-cap arithmetic itself makes immaterial to the final tax.

  2025 Form 511 Packet, Schedule 511-A Instructions, line A6, "Other Retirement Income" — qualifying IRC § 401/408/403(a)/(b) benefits share the same $10,000 combined cap as line A5's government/civil-service retirement. Form 511 itself does not name the TSP; that a TSP distribution falls under IRC § 401 comes from 5 U.S.C. § 8440, which treats the Thrift Savings Fund as a trust described in IRC § 401(a) · 5 U.S.C. § 8440 — treats the Thrift Savings Fund as a trust described by, and exempt under, IRC § 401(a), the basis for TSP distributions qualifying under Schedule 511-A line A6's own IRC § 401 category (an inference from federal law, not a Form 511 statement)

• A FERS annuity is NOT eligible for 68 O.S. § 2358 paragraph 18's 100% "in lieu of Social Security" exclusion (Schedule 511-A line A3), which is specific to CSRS: "retirement benefits paid under [FERS] do not qualify for this exclusion," because FERS employees, unlike CSRS employees, are covered by Social Security. A FERS pension goes through the general $10,000/person cap (line A5, "Civil Service of the United States") like any other qualifying pension.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The FERS Annuity Supplement, which the SAME line A3 note carves back IN as an exception ("paid to certain FERS retirees until eligible for Social Security at age 62, will qualify for the exclusion") — not modelled, since the engine cannot isolate the supplement from the rest of `pensionIncome`; it stays inside the general $10,000/person cap, which OVERSTATES tax for an early retiree drawing one.

  2025 Form 511 Packet, Schedule 511-A Instructions, line A3, "Federal Civil Service Retirement in Lieu of Social Security" — "Note: Retirement benefits paid under the Federal Employees Retirement System (FERS) do not qualify for this exclusion, except: ... The FERS Annuity Supplement ... paid to certain FERS retirees until eligible for Social Security at age 62, will qualify for the exclusion."

• An additional $1,000 exemption applies for each taxpayer or spouse 65+ whose Federal AGI (Roth-conversion income excluded from the test) is at or below $15,000 single/MFS or $25,000 joint/surviving spouse. An independent adversarial review found this rule originally dismissed the provision as "essentially never" applicable to this engine's population without checking a concrete counterexample; a real, low-income federal-retiree scenario (a modest pension/wages combination at or under the ceiling) shows a genuine, nonzero tax difference, so it is modelled rather than merely disclosed.

  Tax years 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The $12,500 MFS and $19,000 head-of-household AGI ceilings — this engine does not model either as a distinct filing status; an MFS-shaped household defaults to the single figure, the same convention this engine already applies elsewhere for MFS.

  2025 Form 511 Packet, "Special" exemption instructions (p.8) — "An additional exemption may be claimed for each taxpayer or spouse who meets the qualifications based on filing status and Federal AGI limits ... and who is 65 years of age or older at the close of the tax year. (1) Single return with line 1 equal to $15,000 or less. (2) Joint return with line 1 equal to $25,000 or less. ... *Note: If your Federal AGI includes income from the conversion of a traditional individual retirement account to a Roth individual retirement account, this income shall be excluded in determining the Federal AGI limits."

• A qualifying surviving-spouse return uses the SAME doubled bracket thresholds and $12,700 standard deduction as MFJ, per Form 511's own bracket table and standard-deduction instructions, which group "married filing joint" and "qualifying surviving spouse" together for both — not the single-filer figures, even though this engine's `isSingle` flag reads true for a surviving-spouse filing status upstream. BUT it gets only the filer's OWN $1,000 personal exemption, not a second $1,000 spouse exemption: Form 511's own "Exemption Terms — Spouse" instructions grant the second exemption only for an actual joint return, for MFS/HoH with a non-filing spouse, or on a death-year return where the filer could have claimed the spouse on the date of death — and never lists "qualifying surviving spouse," which by definition has no living or death-year spouse in the year modelled. An independent adversarial review found this rule originally granted the second exemption to `surviving_spouse` by analogy with other certified states' own conventions, without checking Oklahoma's own exemption rules; independently re-verified directly against Form 511's own text before fixing.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of household, which Form 511 groups into the SAME doubled bracket table as MFJ/surviving spouse but a DIFFERENT ($9,350) standard deduction — moot here since this engine does not model HoH as a distinct filing status at all.
  • A dependent-based qualifying surviving spouse's own additional dependent exemption(s) — this engine collects no dependent count.

  Oklahoma Tax Commission, "Summary of 2025 Tax Legislation" — HB 2764's doubled bracket table is captioned "Married Joint, Head of Household & Surviving Spouse Filers"; 2025 Form 511 Packet, "Standard Deduction" instructions group "married filing joint" and "qualifying surviving spouse" under the same $12,700 figure; "Exemption Terms — Spouse" instructions list only married filing joint, MFS/HoH with a non-filing spouse, and a death-year return, never a later-year qualifying surviving spouse

• 2025 Form 1 instructions: "Massachusetts does not tax Social Security income; therefore, you should not report such income on Massachusetts Form 1." This rule takes no Social Security parameter at all, so SS has no code path into the taxable base.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The full composition of federal AGI beyond wages, pension and TSP — this rule's income model carries the same simplified-income-model gap as every other certified state.

  2025 Form 1 Instructions, Line 4 area — "Note: Massachusetts does not tax Social Security income; therefore, you should not report such income on Massachusetts Form 1."

• "Distributions made to you from a federal employee contributory plan are excluded from Massachusetts gross income... These pensions which are paid to surviving spouses are also tax exempt." (MGL ch. 62 §2(a)(2)(E).) `pensionIncome` — the one household figure that already folds in a person's own FERS annuity, the FERS supplement, AND any survivor annuity they receive — is discarded entirely, so this needed no fix: the exemption's own text confirms it reaches a survivor annuity, not merely the retiree's own.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • MA state/local government pensions (ch. 32) — exempt too, but out of scope: those employees are not FERS federal employees.

  mass.gov, "Tax Treatment of Government Pensions in Massachusetts," "Federal employee contributory pension" section — "Distributions made to you from a federal employee contributory plan are excluded from Massachusetts gross income... These pensions which are paid to surviving spouses are also tax exempt."

• "Distributions made to the retiree from the [Thrift Savings Plan] are fully taxable in the year paid." The TSP is a defined-contribution plan distinct from the FERS annuity and gets no exclusion of its own.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Any Massachusetts-previously-taxed-contributions adjustment of the kind §403(b)/§404 plans get on this same form — no analogous TSP adjustment was found in either source, and the engine has no field to carry one regardless.

  mass.gov, "Tax Treatment of Government Pensions in Massachusetts," "Federal employee Thrift Savings Plan (TSP)" section — "Distributions made to the retiree from the plan are fully taxable in the year paid."

• Personal exemption of $4,400 for single/married-filing-separately, $8,800 for married filing jointly (MGL ch. 62 §3(B)(b)(1)(A)), confirmed unchanged between the 2025 Form 1 instructions and the 2026 Form 1-ES. FOUND BY REVIEW, CORRECTING THIS PACK'S OWN FIRST DRAFT: these are NOT unindexed flat dollar amounts. The statute grows each figure by a fixed increment ($275 single/MFS, $550 MFJ) whenever the Commissioner certifies sufficient inflation-adjusted revenue growth, capped at exactly these two dollar amounts — the current figures are correct because they sit AT that statutory ceiling, not because the mechanism never adjusts them.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The $6,800 head-of-household exemption — this engine has no Head of Household filing status anywhere.
  • Any future year in which the statutory cap itself changes by further legislation — this claim certifies the CURRENT cap, not that it is permanent.

  2025 Form 1 Instructions, Line 2a — "If you are single or married filing a separate return, enter $4,400 in line 2a... If married filing a joint return, enter $8,800 in line 2a." · Massachusetts General Laws c. 62, §3(B)(b)(1)(A) — "For taxable years beginning on or after January 1, 2004, the personal exemption shall be: (i) the exemption in the previous year plus $275 [single/MFS] / $550 [MFJ] if the inflation-adjusted growth in baseline taxes... exceeds 2.5 per cent... The personal exemption shall not exceed $4,400 [single/MFS] / $8,800 [MFJ]."

• THE DEFECT THIS PACK FIXES. An additional $700 exemption applies for a filer age 65 or over before the close of the tax year, with no filing-status gate on the filer's OWN exemption. A SPOUSE's own $700 additionally requires "filing a joint return" — read from `filingStatus === 'mfj'` (self-documenting; provably equivalent to `!isSingle` given this engine's own `isSingle` convention, which groups `surviving_spouse` WITH `single`, unlike Oklahoma's `isJoint`), so a `surviving_spouse` household (which by definition has no living spouse to claim on this return) gets no second $700 regardless of person B's age. A decedent keeps the age reached before death via `ageAtEndOfTaxableYear`, matching this engine's Maryland/Missouri/Georgia/Ohio/Arizona precedent for a bare age gate. The prior rule read no age at all.

  Tax years 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The $2,200 blindness exemption — the engine collects no blindness input anywhere.

  2025 Form 1 Instructions, Line 2c — "Age 65 or Over Before 2026. You are allowed an additional $700 exemption if you were age 65 or over before January 1, 2026. If your spouse was age 65 or over and you are filing a joint return, you may also claim a $700 exemption for your spouse."

• THE DEFECT FOUND BY AN INDEPENDENT ADVERSARIAL REVIEW, AND THE MORE SERIOUS OF THE TWO THIS PACK FIXES. This pack's own first draft modelled neither No Tax Status nor the Limited Income Credit at all, overstating tax for every low-income household — including some of that SAME draft's own test cases, which asserted a nonzero tax for an income that was actually always $0. A household with Massachusetts AGI at or below $8,000 (single, including `surviving_spouse`) or $16,400 (MFJ) owes $0 outright; one just above either threshold has tax capped at 10% of the excess (the Limited Income Credit), which is provably equivalent to the worksheet's own $14,000/$28,700 eligibility ceiling without a separate branch for it — see `MA_LIMITED_INCOME_CREDIT_RATE`'s own comment for the algebraic proof.

  Tax years 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Head of Household's own $14,400/$25,200+ thresholds — this engine has no Head of Household filing status anywhere.
  • Any dependent count — both thresholds are modelled at 0 dependents; this engine collects no dependent count anywhere.
  • "Massachusetts AGI" as the form actually defines it (5% income plus interest/dividends/capital gains, less Schedule Y deductions) — this rule uses `gross` (wages + TSP) as a faithful stand-in, since the engine models none of the other components in any state.

  2025 Form 1 Instructions, Line 27 — "If your Massachusetts AGI was $8,000 or less if single, $14,400 or less plus $1,000 per dependent if head of household, or $16,400 or less plus $1,000 per dependent if married filing a joint return, you qualify for No Tax Status and are not required to pay any Massachusetts income taxes." · 2025 Form 1 Instructions, Line 29 Worksheet (Limited Income Credit) — caps tax at 10% multiplied by (Massachusetts AGI minus the No Tax Status threshold), for a single filer with AGI up to $14,000 or an MFJ filer up to $28,700 (0 dependents)

• A `surviving_spouse` household reads the $4,400 single exemption, the same as `isSingle` reading true for that status already produces — UNLIKE Maryland (joint schedule), Arizona (a third, head-of-household figure) or Missouri, which each needed a rule-level fix for this same federal status. Massachusetts's own "Single" filing-status definition explicitly covers "a widow or widower whose spouse died before [the current tax year]" — exactly the genuine federal QSS window this engine's `surviving_spouse` value represents in the projection's default mode — and Massachusetts has no qualifying-widow(er) status distinct from Single. FOUND BY REVIEW: $4,400 IS A VALID BUT POSSIBLY CONSERVATIVE READING, NOT THE ONLY CORRECT ONE. Massachusetts's own Head of Household status is gated on federal HoH ELIGIBILITY ("if you qualify to file this status federally"), a facts test, not on which status was actually elected — and this engine sets `surviving_spouse` only when a dependent-child attestation exists, so such a household typically ALSO meets HoH's own underlying facts (an unmarried person paying over half the cost of a home for a qualifying child) and could elect the more favourable $6,800 HoH exemption instead. This rule cannot reach that election.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The projection adapter's own disclosed 'immediate'-mode approximation, which assigns `surviving_spouse` starting in the death year itself rather than only the two years that follow. Massachusetts actually permits a joint return for the death year itself ("If your spouse died during [the year], you may still choose to file a joint return"), which that mode would understate as a Single return — an engine-wide caveat this rule inherits, not a defect of its own.
  • Whether the household ALSO meets Head of Household's "paid over half the cost of keeping up a home" test specifically — the federal QSS dependent-child attestation this engine collects does not itself confirm that spending test, only the dependent relationship.

  2025 Form 1 Instructions, Line 1 — "Single. Fill in the Single oval if you were single as of December 31, 2025... You were a widow or widower whose spouse died before 2025... A joint return may be filed by a surviving spouse. ... If your spouse died during 2025, you may still choose to file a joint return."

• The flat 5.0% rate is unconditional and confirmed current for tax year 2026 directly from the 2026 Form 1-ES worksheet's own "Taxable 5% income" line, not merely held forward from 2025 as a disclosed projection — unlike Maryland's or California's own indexed figures, Massachusetts's rate has no year-stepped schedule at all.

  Tax years 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The 4% surtax on taxable income over the annually-indexed threshold ($1,107,750 for 2026) — out of scope for essentially every FERS household this calculator models.
  • The 8.5%/12% short-term-capital-gains and collectibles rate schedules — the engine models no capital-gains income category in any state.

  2026 Form 1-ES, Estimated Tax Worksheet — "1 Taxable 5% income* (after deductions and exemptions)... *5% income includes: wages, salaries, tips, business income, partnership and S corporation income, trust income, rental income, unemployment... interest and dividend income and other taxable income not taxed at the 8.5%..." · mass.gov, "Massachusetts 4% Surtax on Taxable Income" — confirming the 2026 surtax threshold ($1,107,750) as an out-of-scope high-income provision rather than an unaddressed one

• 2025 IT-40 booklet, Schedule 1 Lines 5-6, "Taxable Social Security and/or Railroad Retirement Benefits Deduction": Indiana subtracts the federally-taxable portion of Social Security back out of the state base, so it is never taxed. Social Security only ever appears in this rule as an offset against the civil service annuity deduction.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The full composition of federal AGI beyond wages, pension and TSP — this rule's income model carries the same simplified-income-model gap as every other certified state.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet, Schedule 1, Lines 5-6 — "Taxable Social Security and/or Railroad Retirement Benefits Deduction"

• DOR's "2026 Tax Chapter for the 2025 Filing Year" (Oct. 2025): "Individual Income tax rate is lowered to 3% for tax year 2025 and 2.95% for tax year 2026." Confirmed independently by Departmental Notice #1 effective Jan. 1, 2026: "the state adjusted gross income tax rate for individuals is 2.95%." The pre-existing rule applied 3.0% to every projection year.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • IC 6-3-2-1's own scheduled further step to 2.90% in 2027 — contingent on a State Budget Agency revenue-trigger determination this pack found no confirmed primary source for. Years 2027 and later hold 2.95% forward as a disclosed projection, not a verified figure; re-verify once DOR publishes the 2027 rate.

  Indiana Department of Revenue, "2026 Tax Chapter for the 2025 Filing Year" (last revised October 2025) — "Individual Income tax rate is lowered to 3% for tax year 2025 and 2.95% for tax year 2026." · Indiana Department of Revenue, Departmental Notice #1, effective Jan. 1, 2026 — "For 2026, the state adjusted gross income tax rate for individuals is 2.95%."

• IC 6-3-2-3.7 and the 2025 IT-40 booklet's own worked example ("Matthew and Claire, both age 68 ... Matthew's taxable civil service annuity is $13,700; he also received $17,500 in Social Security ... he is not eligible for a deduction. Claire's taxable civil service annuity is $21,900; she also received $6,300 ... Allowable deduction ... $9,700.") establish the $16,000 cap and Social Security offset are computed PER ANNUITANT, not pooled at the household level. The prior rule capped household pension at $16,000 times eligible headcount and subtracted household Social Security in one combined step, which for this same example would yield $8,200, not $9,700. Indiana's own materials never use the word "FERS" — they say "federal civil service annuity" and require retaining "Form CSA 1099-R". IRS Publication 721 confirms Form CSA 1099-R is issued for BOTH CSRS and FERS annuity income (Form CSF 1099-R for survivor annuitants), which is the bridge this pack uses to confirm the deduction reaches the FERS population this app models.

  Tax years 2025, 2026 · 3 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Railroad retirement (tier 1/tier 2) benefits, which also offset this deduction per IC 6-3-2-3.7 — this app has no railroad-retiree income category to read.
  • A single filer receiving both their own annuity and a deceased spouse's survivor annuity, whose "combined deduction cannot exceed $16,000" per the booklet — not separately modeled, since this engine folds all of one person's civil-service income into a single `pensionIncomePersonA/B` figure already subject to the same per-person $16,000 cap.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet, Schedule 2, "Civil Service Annuity Deduction 601" — the Matthew-and-Claire worked example · Income Tax Information Bulletin #6 (June 2025, effective Jan. 1, 2025) — "The allowable adjustment is equal to the federal civil service annuity received up to a maximum of ... $16,000 for 2016 and thereafter, minus the total amount of Social Security ... received." · IRS Publication 721, "Tax Guide to U.S. Civil Service Retirement Benefits" — confirms Form CSA 1099-R covers both CSRS and FERS annuity income, and Form CSF 1099-R covers survivor annuitants

• IC 6-3-2-3.7 reaches only "a federal civil service annuity". `pensionIncomePersonA/B` is a mixed bucket that also carries non-federal `fixedRetirementIncome`. FOUND BY AN INDEPENDENT ADVERSARIAL REVIEW: this pack's own first draft applied the $16,000 cap to the whole mixed bucket, reaching non-federal pension income the statute never authorizes. Fixed by subtracting `nonFederalPensionIncomePersonA/B` before applying the cap.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether Indiana law would treat a MIXED federal/non-federal single payment stream differently from this engine's clean per-category split — not addressed in any source read for this pack, and not expressible by the engine's own field model regardless.

  law.justia.com's reproduction of Indiana Code 6-3-2-3.7 — the deduction reaches only income "received ... from a federal civil service annuity" (direct fetch returned HTTP 403; retrieved via search, corroborated by IB #6's restatement of the same language)

• IB #6: "To qualify for the civil service annuity adjustment, the taxpayer must be at least 62 years old at the close of the tax year." This applies to a person's OWN annuity share (see the separate surviving-spouse-waiver claim below for the survivor-sourced share, which has no age requirement). A decedent still on the return keeps the age they reached before death, per this engine's `ageAtEndOfTaxableYear` convention.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether a person exactly 62 at some point mid-year but not "at the close of the tax year" could partially qualify — the statute's own test is a single year-end snapshot and this engine's age fields carry no finer time resolution than that snapshot already provides.

  Income Tax Information Bulletin #6 — "To qualify for the civil service annuity adjustment, the taxpayer must be at least 62 years old at the close of the tax year and have received a civil service annuity... The individual's surviving spouse also qualifies for the deduction but does not have to be age 62 or older to qualify."

• IC 6-3-2-3.7 and IB #6: "The individual's surviving spouse also qualifies for the deduction but does not have to be age 62 or older to qualify." The IT-40 booklet's own Marie example ("Marie is a surviving spouse. The taxable amount of her civil service annuity is $14,500, and she received $1,200 in Social Security income... Allowable deduction... $13,300") has no age stated. FOUND BY AN INDEPENDENT ADVERSARIAL REVIEW: this pack's own first draft wrongly claimed the engine had no field to identify survivor annuity income (`survivorBenefitsPersonA/B` already carries it, added for D.C. Code § 47-1803.02(a)(2)(N)(ii), #352) and disclosed this as unmodelable rather than modeling it. The waiver reaches only the survivor-sourced SHARE of a person's pension — their own annuity, if any, still requires age 62.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether a survivor herself later becomes eligible under her OWN age-62 test once any of her income stops being survivor-sourced and starts being her own separate annuity — not a scenario any source read for this pack addressed, and the engine has no field to distinguish it from continuing survivor income regardless.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet, Schedule 2, "Civil Service Annuity Deduction 601" — the Marie worked example · Income Tax Information Bulletin #6 — "The individual's surviving spouse also qualifies for the deduction but does not have to be age 62 or older to qualify."

• IC 6-3-2-3.7 reaches only "a federal civil service annuity" — a defined-benefit annuity, not the TSP, a defined-contribution account. TSP distributions get no deduction of their own and do not enlarge the $16,000 annuity cap.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Any Indiana adjustment specific to TSP contributions or basis — no such provision was found in any source read for this pack, and this rule's income model carries TSP as a single gross-distribution figure regardless.

  law.justia.com's reproduction of Indiana Code 6-3-2-3.7 — the deduction reaches only income "received ... from a federal civil service annuity" (direct fetch returned HTTP 403; retrieved via search, corroborated by IB #6's restatement of the same language)

• 2025 IT-40 booklet, Schedule 3 line 1: "If you are married filing jointly, enter $2,000 on this line. All other filers should enter $1,000 on this line." The prior rule modeled no exemption of any kind.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The dependent exemptions on Schedule 3 lines 2-3 (and their own additional $1,500 dependent-child exemption) — this engine models no dependents in any state.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet, Schedule 3, Line 1

• 2025 IT-40 booklet, Schedule 3 lines 4-5: an unconditional $1,000 exemption per person age 65 or older, plus a further $500 per person if household federal AGI (Form IT-40 line 1) is under $40,000.

  Tax years 2025, 2026 · 6 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The $1,000 blindness exemption sharing Schedule 3 line 4's own box-and-multiply mechanic — the engine collects no blindness input anywhere.
  • The married-filing-separately variant of the Line 5 AGI threshold ($20,000 rather than $40,000) — this engine has no MFS filing status.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet, Schedule 3, Lines 4-5

• A household whose exemptions and deductions exceed its Indiana AGI owes $0, never a negative figure.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether Indiana permits any carryforward of unused exemptions to a future tax year — no such provision was found in any source read for this pack, and this rule computes each year independently regardless.

  2025 IT-40 Full-Year Resident Individual Income Tax Booklet — the tax computation schedule has no provision for a negative liability

• 2025 Publication OR-17: "Oregon doesn't tax Social Security benefits." This rule takes no Social Security parameter into the taxable base at all.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The full composition of federal AGI beyond wages, pension and TSP — this rule's income model carries the same simplified-income-model gap as every other certified state.

  2025 Publication OR-17, Oregon Individual Income Tax Guide — "Oregon doesn't tax Social Security benefits."

• The 2025 second-bracket boundary is $11,100 (single) / $22,200 (MFJ). Single is independently re-derived from the 2025 Form OR-40 Instructions' own Chart S worked figure ("$4,065 plus 8.75% of excess over $50,000"). MFJ is confirmed directly from Publication OR-17's own multi-year "Marginal tax rates" table, whose 2025 column prints the 6.75% MFJ band as "$8,801–$22,200" and the 8.75% band starting at "$22,201" — both fail with the prior rule's $11,050/$22,100 figures (2024's own boundary column in the same table, never updated for 2025).

  Tax years 2025 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Years before 2025, which this rule does not certify and which held no prior claim.
  • A $1 discrepancy between Form OR-40's own ROUNDED Chart J prose ("$3,756... over $50,000") and the dollar-exact bracket-ladder computation ($3,755) — this rule computes the exact bracket ladder rather than the rounded chart prose, the same choice this engine's other states make when a summary chart and a computation formula disagree by a rounding amount.

  2025 Form OR-40 Instructions, "2025 Tax rate charts" — Chart S worked tax-at-$50,000 figure · 2025 Publication OR-17, "Marginal tax rates" table — 2025 column MFJ 6.75% band "$8,801–$22,200", 8.75% band "Over $22,201"

• DOR's 2026 Oregon Withholding Tax Formulas confirms the 2026 standard deduction ($2,910 single / $5,820 MFJ) and bracket boundaries ($4,550/$11,400 single, $9,100/$22,800 MFJ) directly, distinct from 2025's figures. The prior rule applied 2025's $2,835/$5,670 standard deduction and $11,050/$22,100 boundaries to every projection year, including 2026.

  Tax years 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • 2027 and later years, which hold 2026's own confirmed figures forward as a disclosed projection — no 2027 Oregon tax table has been published as of this pack's own research date.

  2026 Oregon Withholding Tax Formulas (150-206-436, Rev. 12-31-25) — the 2026 standard deduction, exemption credit and bracket boundaries

• 2025 Publication OR-17's own multi-year reference table gives the personal exemption credit as $256 for 2025 (the prior rule held the stale 2023 figure, $236) and DOR's 2026 withholding formulas confirm $263 for 2026. ORS 316.085 denies the credit outright — not on a phase-out — once federal AGI exceeds $100,000 (single/MFS) or $200,000 (MFJ/HoH/QSS).

  Tax years 2025, 2026 · 4 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The severe-disability additional exemption credit and the child-with-a-disability additional exemption credit (ORS 316.752, 316.758, 316.099) — this engine models no disability status for any state.
  • ORS 316.085's own per-dependent multiplier — the credit is granted "for each" qualifying exemption, including each qualifying child and each other qualifying relative, not merely the taxpayer(s) themselves. This engine models no dependents in any state, the same convention already disclosed identically for Virginia, Maryland and Georgia; a household with dependents is UNDERCREDITED here.

  2025 Publication OR-17 — multi-year reference table ("Exemption credit $219 or $0x% / $236 or $0x% / $249 or $0x% / $256 or $0x%" for 2022-2025) and ORS 316.085's AGI-cliff text: "A personal exemption credit isn't allowed if your federal AGI ... exceeds ... $200,000 for married filing jointly ... or $100,000 if your filing status is single or married filing separately." · 2026 Oregon Withholding Tax Formulas — "$263" personal exemption credit and "greater than $100,000"/"greater than $200,000" allowance cliffs

• 2025 Publication OR-17, "Standard deduction—Age 65 or older, blind": an additional $1,200 (single/HoH) or $1,000-per-qualifying-spouse (married/QSS) once a taxpayer "turned 65 on or before January 1" of the following year. The prior rule modeled no age-based standard-deduction addition at all. "Turned 65 on or before January 1 [of the following year]" is the IRC § 63(f) day-before-your-birthday convention, not a plain 31-December age — an earlier version of this fix used a plain age and wrongly denied the addition to a person born 1 January, corrected by this pack's own second Codex review round using `federalAgeAtEndOfTaxableYear`, the same helper Missouri's and Virginia's own rules already use.

  Tax years 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • The identically-structured $1,200/$1,000 blindness addition sharing the same box-and-multiply mechanic — the engine collects no blindness input anywhere.

  2025 Publication OR-17, "Standard deduction—Age 65 or older, blind" — "$1,000 if married or qualifying surviving spouse. $1,200 if single or head of household."

• ORS 316.695/Table 9: Oregon allows a subtraction for the household's federal income tax liability after credits, capped at $8,500 for 2025 and $8,750 for 2026 (below the AGI phase-out threshold). The prior rule disclosed this as unmodelable, claiming the engine had no federal-tax-paid field — false; `federalOrdinaryIncomeTax` already exists and is read by this engine's Missouri and Alabama rules.

  Tax years 2025, 2026 · 5 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Table 9's own AGI phase-out above $125,000 (single) / $250,000 (MFJ), which steps the maximum subtraction down in increments until it reaches $0 — no 2026 edition of Table 9 was found, so this pack applies only the flat, year-gated cap. This OVERSTATES the subtraction (and UNDERSTATES Oregon tax) for a household whose federal AGI sits in the phase-out band.
  • Part B (federal tax paid for a PRIOR year) and Part C (foreign income tax) of the same worksheet — this engine has no prior-year or foreign-tax field for any state to read.
  • `federalOrdinaryIncomeTax` is itself NARROWER than the worksheet's own Line 1 ("Federal tax liability, Form 1040, line 22") and Line 4 ("Other taxes," e.g. AMT, excess-APTC repayment, and additional tax on early distributions — this engine tracks the 10% early-distribution additional tax separately in the projection audit trail and deliberately does not fold it into this field, the same scope gap already disclosed for Missouri's analogous federal-tax deduction). This UNDERSTATES the subtraction (and OVERSTATES Oregon tax) for a household with any such items on the federal return.

  2025 Publication OR-17, "Federal income tax liability" (ORS 316.680, 316.685, 316.695) and Table 9 — "$8,500" maximum subtraction, single, federal AGI under $125,000 · 2026 Oregon Withholding Tax Formulas — "federal tax withheld (not to exceed $8,750)"

• TSP distributions are ordinary retirement income, Oregon-taxable the same as a FERS pension distribution. Publication OR-17's own "Federal Thrift Savings Plan (TSP)" section states that TSP withdrawals ARE eligible for a subtraction — the SAME ORS 316.680(1)(e) pre-October-1991-service subtraction already disclosed as unmodeled for the pension share (below): "Once a taxpayer is a retiree, withdrawals from federal TSP accounts are eligible for the subtraction based on dates of service." An earlier draft of this claim said no TSP-specific exclusion existed at all — that was false, found by this pack's own independent Codex review round; corrected here to disclose the SAME gap the pension share already has, rather than assert none exists.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • ORS 316.680(1)(e)'s pre-October-1991-service subtraction, extended to TSP by OR-17's own text — unreachable for the same reason as the pension share: `StateTaxIncomeBreakdown` carries no federal-service-date field for any state to read. OR-17 also notes that once TSP money is moved to another type of account it "loses its character" and is no longer eligible, so this gap narrows over time as fewer retirees hold un-rolled-over TSP balances with pre-1991 service credit.

  2025 Publication OR-17, "Federal Thrift Savings Plan (TSP)" — "withdrawals from federal TSP accounts are eligible for the subtraction based on dates of service"

• THE SIXTH DEFECT THIS PACK FOUND: the prior rule keyed brackets, standard deduction and exemption-credit AGI limit off `isSingle`, which this engine's own convention groups `surviving_spouse` WITH `single` — so a qualifying surviving spouse was taxed as SINGLE, contrary to Chart J's own heading ("For persons filing jointly, head of household, or qualifying surviving spouse") and OR-17's own standard-deduction table, which lists "Married filing jointly or qualifying surviving spouse" as a single combined row. Fixed by keying those three branches off `filingStatus !== 'single'` instead. BUT OR-17's own "Exemption credit" section grants the spouse's own second exemption credit only to "you're filing a joint return," so a `surviving_spouse` return (which has no living co-filer) gets exactly ONE credit, not MFJ's two — kept keyed on `filingStatus === 'mfj'` specifically.

  Tax years 2025, 2026 · 2 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether a qualifying surviving spouse could instead elect Head of Household treatment where more favorable — Oregon's own HoH eligibility rests on federal facts (an unmarried person maintaining a home for a qualifying person) this engine does not independently track for a `surviving_spouse` household, so this rule reads only the QSS branch Chart J itself names.

  2025 Form OR-40 Instructions, Chart J heading — "For persons filing jointly, head of household, or qualifying surviving spouse" · 2025 Publication OR-17, "Exemption credit" section — "You are allowed one personal exemption credit for your spouse if ... You're filing a joint return."

• THE NINTH FIX, found and modeled during this pack's own second Codex review round: ORS 316.157's retirement income credit was entirely unmodeled. Age 62+ with taxable pension/TSP income under a narrow combined income/Social-Security test earns a nonrefundable, non-carryforward credit of 9% of the lesser of (a) the eligible person(s)' own Oregon-taxable pension/TSP income or (b) a household-income-and-Social-Security-headroom figure computed by OR-17's own 12-line worksheet — capped in practice at $675 (single) / $1,350 (MFJ). "Married filing jointly" ONLY, per the worksheet's own bullets — unlike the brackets/standard deduction/exemption credit, it does not extend to a qualifying surviving spouse. THE CREDIT SUNSETS FOR TAX YEARS BEGINNING ON OR AFTER 1 JANUARY 2032 (ORS 316.157, as amended by 2025 c.562 §5, which extended what had been a 2026 sunset) — a defect this pack's own THIRD Codex review round caught: the initial implementation of this fix computed the credit for every projection year, with no gate at all.

  Tax years 2025, 2026 · 8 cases · re-check due 2026-12-01

  Not covered by this claim:

  • "Household income" is approximated as federal AGI less each spouse's federally taxable Social Security (`taxableSsIncomePersonA/B`), rather than OR-17's full definition. Most of the items OR-17's own instructions add back are never in federal AGI at all (veterans/military benefits, gifts and scholarships over $500, disability pay, nontaxable interest, lottery winnings, RRB Tier 2) and this engine's income model carries no field for any of them. TAX-EXEMPT INTEREST IS A NARROWER GAP THAN THAT, caught by this pack's own THIRD Codex review round after the second round's "no field for any of them" wording overstated it: the app DOES collect tax-exempt interest (`taxExemptInterestAnnual`) and it already reaches Oregon's own taxable base (folded into the household `wages` category, a project-wide convention), but it never reaches `federalAdjustedGrossIncome` — which correctly excludes it, as real federal law does — and no separate field threads it to this worksheet apart from `wages`, so it is genuinely unreachable HERE specifically. "IRA conversions" is NOT a gap at all — OR-17's own bullet text is "IRA conversions included in AGI," meaning they are already inside `federalAdjustedGrossIncome`, corrected from an earlier version of this claim that grouped it with the genuinely-excluded items. Missing any of the genuine gaps UNDERSTATES household income and so slightly OVERSTATES the credit for a household with any such item.
  • The federal-pension subtraction (ORS 316.680(1)(e), disclosed as unreachable elsewhere in this pack) reduces the worksheet's own Line 3 for anyone who receives it — since this engine never applies that subtraction, Line 3 here is always the FULL pension/TSP amount, which OVERSTATES Line 3 (and so can overstate the credit) for a CSRS-to-FERS transferee with substantial pre-October-1991 service.
  • Tier 1 Railroad Retirement Board benefits, which the worksheet folds into the same Social Security line — this engine has no Railroad Retirement field for any state.

  2025 Publication OR-17, "Retirement income (ORS 316.157)" [Credit code 811] — full eligibility bullets and the 12-line worksheet: "Enter the retirement income of the eligible individual(s)... Multiply line 11 by 9% (0.09). This is your credit." · ORS 316.157 (2025 Oregon Revised Statutes, as amended by 2025 c.562 §5) — "A credit may not be claimed under ORS 316.157 for tax years beginning on or after January 1, 2032."

• A household whose standard deduction and federal tax subtraction exceed its Oregon gross income owes $0, never a negative figure.

  Tax years 2025, 2026 · 1 cases · re-check due 2026-12-01

  Not covered by this claim:

  • Whether Oregon permits any carryforward of an unused deduction to a future tax year — no such provision was found in any source read for this pack, and this rule computes each year independently regardless.

  2025 Publication OR-17 — the tax computation schedule has no provision for a negative liability