Add North Carolina's wagering losses itemized deduction from 2025 (G.S. 105-153.5(a)(2)e)

[MaxGhenis]

Summary

North Carolina allows itemizers to deduct wagering losses from tax year 2025. This PR adds nc_wagering_losses_deduction: the wagering losses allowed under 26 U.S.C. 165(d) (PE's wagering_losses_deduction), from 2025, counted in North Carolina itemized deductions outside the $20,000 mortgage and property tax cap.

Depends on two open PRs; its own change is six files on top of them:

• Count gambling winnings in federal gross income and fix the wagering losses deduction #10070 (the fix of Federal gross income omits gambling winnings but still deducts wagering losses; OBBBA 90% wagering loss limit (2026+) not modeled #9637): federal gross income counts gambling winnings, wagering_losses_deduction allows 90% of losses from 2026 (P.L. 119-21 § 70114), and it counts only the head's and spouse's gambling.
• Add North Carolina's claim of right repayment itemized deduction (G.S. 105-153.5(a)(2)d) #10028 (the claim of right deduction), which adds the other_sources list this extends.

Until both merge, the diff against main also shows their changes. Against the merge of their current heads (#10028 at 81528e0207, #10070 at 4bfb615ab7), this branch adds only nc_wagering_losses_deduction.py, its in_effect parameter, one line in other_sources.yaml, the YAML and property tests, and the changelog fragment.

The law

S.L. 2026-41 § 44.2.(a) (PDF pp. 607–608; ratified July 2, 2026, approved July 7, 2026) adds sub-subdivision e. to G.S. 105-153.5(a)(2):

e. The amount allowed as a deduction for wagering losses under section 165(d) of the Code, to the extent the losses are not deducted in arriving at adjusted gross income.

§ 44.2.(b): "This section is effective for taxable years beginning on or after January 1, 2025."

NCDOR's notice of July 23, 2026 (updated July 29) gives an example. Jane Smith had $10,000 of 2025 gambling winnings, all in AGI, and $12,000 of losses; she "can include $10,000 in gambling losses". The deduction goes on Form D-400 Schedule A line 9 for 2025. NCDOR's FAQs of September 11, 2026, Q36–Q40 (PDF pp. 11–12): the winnings must be reported as income, cannot be reduced by the losses, and the losses are claimed separately as an NC itemized deduction.

North Carolina's Code is the Internal Revenue Code as of July 5, 2025 (S.L. 2026-31 § 12.(a)). That includes P.L. 119-21 § 70114, which allows 90% of losses from 2026, so North Carolina follows the federal amount in 2026 too.

Households, main and this branch

NCDOR's example as a 2025 single return: wages $60,000, gambling winnings $10,000, losses $12,000, charity $5,000. Evidence: ~/reviews/us-hub/nc-itemized-other/evidence/households_main.json (main ff479aa74a) and ~/reviews/us-hub/gambling-9637/evidence/households_nc_wagering.json (this branch with the federal fix, head de24c038a8).

Case	main: AGI, NC income tax	Branch: AGI, deduction, NC taxable income, NC income tax	Law
2025, PE's own AGI	$60,000, $2,008.12 (winnings missing from AGI)	$70,000, $10,000, $55,000, $2,337.50	$2,337.50
2025, AGI given as $70,000	$70,000, $2,433.12 (no wagering deduction)	$70,000, $10,000, $55,000, $2,337.50	$2,337.50
2024, AGI $70,000	$70,000, $2,576.25	$70,000, $0 (none before 2025), $57,250, $2,576.25	$2,576.25
2025, no gambling	$60,000, $2,008.12	$60,000, $0, $47,250, $2,008.12	$2,008.12

$55,000 = $70,000 − ($10,000 wagering + $5,000 charity); 4.25% × $55,000 = $2,337.50.

Tests

• nc_wagering_losses_deduction.yaml: ten cases. None before 2025; NCDOR's example; losses below winnings; spouses combined; 2026; 2026 at 90% ($8,000 of losses → $7,200); a dependent's gambling ($0); a non-NC household; and nc_income_tax integration for 2025 and 2024.
• test_nc_wagering_losses_properties.py: a 196-point grid (winnings × losses × charity × property tax), three vectorized simulations per year for 2024–2026 (winnings in AGI, no gambling, PE's own AGI). Invariants: zero before 2025; equal to min(share × losses, winnings) with the share restated from the statute; bounded by winnings and losses and monotone in each; adds one for one to NC itemized deductions outside the cap; gambling never lowers NC taxable income and raises it by at most the winnings, by exactly the net gain for a filer who itemizes either way, with AGI given and with PE's own AGI.
• Five checks guarded this deduction on the federal fix: the YAML 2026 and dependent cases, the 2026 section 165(d) check, and the modeled-AGI checks for 2025 and 2026. They failed before Count gambling winnings in federal gross income and fix the wagering losses deduction #10070 and pass after it, with their inputs and expectations unchanged. Two other 2026 property checks compared sums near $30,000 at a tolerance of $0.000001; once the deduction is 90% of losses, float32 rounding reaches $0.0016, so their tolerance is now $0.01.
• At de24c038a8 (~/reviews/us-hub/gambling-9637/evidence/nc_r2.log): NC deduction and income tax YAML 63 passed; both NC property files 20 passed in 2.9 s of tests, 44 s wall, 1.04 GB peak RSS (local macOS, not a CI figure). NC_R3_RESULTS

Microsimulation

A real 2026 run on populace_us_2024.h5@populace-us-2024-spm-20260909 at de24c038a8, against the federal branch's run at f7f555818f (~/reviews/us-hub/gambling-9637/evidence/impact-ncw9637-2026.json): zero change on every record in state_income_tax, income_tax, taxable_income, adjusted_gross_income and household_net_income. gambling_winnings, gambling_losses and wagering_losses_deduction are zero on every record, because the dataset has no gambling, so the deduction is zero everywhere.

Axiom

axiom: N.C. Gen. Stat. 105-153.5(a)(2)e TheAxiomFoundation/rulespec-us#1656 queued

🤖 Generated with Claude Code