Skip to content

Fix Utah relief household income and the Utah child tax credit phase-out income - #11182

Open
DTrim99 wants to merge 5 commits into
mainfrom
ut-relief-ctc-income
Open

DTrim99 wants to merge 5 commits into
mainfrom
ut-relief-ctc-income

Conversation

@DTrim99

@DTrim99 DTrim99 commented Nov 19, 2026

Copy link
Copy Markdown
Collaborator

Part of #10812 (state credit income audit).

Part of #10973 (its Utah alimony item and the page check; the issue covers more than Utah).

What was wrong

Homeowner's/renter's relief household income. Utah Code 59-2a-101(17) defines income as federal AGI plus nontaxable income, and (20)(a) defines nontaxable income as "amounts excluded from adjusted gross income under the Internal Revenue Code, including" a list (Utah Code Title 59 Chapter 2a, pages 3-4). The model's list:

  • held alimony_income in full, so alimony under a pre-2019 divorce instrument, which is already in AGI, counted twice (divorce_year defaults to 2010, so alimony with no divorce year did too);
  • left out SSI, which (20)(a)(vii) names as "payments received under the Social Security Act". (20)(b)(vi) leaves out only social security disability income, and (24) (page 5) limits "public assistance" to medical assistance, SNAP, Title 35A Chapter 3 benefits and foster care maintenance payments;
  • left out nontaxable disability insurance benefits: (20)(a)(xi) counts "the gross amount of 'loss of time' insurance";
  • left out capital loss carryovers: (20)(a)(ii) counts "loss carry forwards claimed during the taxable year", and Form TC-90CB line 5 (page 1) reports "Capital gain/loss (exclude carryforwards/carrybacks)". AGI already nets the carryover (long_term_capital_gains includes it as a loss).

Utah child tax credit phase-out. Utah Code 59-10-1047(1)(d) defines modified adjusted gross income as the claimant's (and a joint spouse's) AGI, plus interest income not in AGI, plus the additions of 59-10-114. Nothing is subtracted. The 2025 TC-40 instructions' Child Tax Credit Worksheet (page 24) starts at TC-40 line 6 (total income), takes out the code 57 municipal bond interest addition (line 4) and adds federal tax-exempt interest (line 6). ut_ctc_potential phased out on ut_taxable_income (line 9), which is net of ut_subtractions (line 8, including U.S. government interest) and the state tax refund (line 7), so the credit ran too high for anyone with either.

What changes

gov.states.ut.tax.property.homeowner_renter_relief.nontaxable_income.sources:

Entry Source
alimony_income replaced by the new ut_homeowner_renter_relief_nontaxable_alimony: alimony_income − taxable_alimony_income when the latter is a federal gross income source (the AZ, ND and RI pattern) 59-2a-101(20)(a), (a)(iv) "support money received" (page 4)
ssi added (20)(a)(vii), (20)(b)(vi), (24) (pages 4-5)
tax_exempt_disability_benefits added (the taxable part is already in AGI, so the gross amount counts once) (20)(a)(xi) (page 4)
New ut_homeowner_renter_relief_loss_carryforward added (20)(a)(ii) (page 4); TC-90CB line 5 (page 1)

ut_homeowner_renter_relief_loss_carryforward is the amount by which long_term_capital_loss_carryover lowered AGI: the return's capital gain or loss in AGI (net gain, or net loss up to the IRC 1211(b) limit) without the carryover, less that amount with it (the AZ/MT logic of adding the carryover back to the net gain). The head and spouse's return combines both spouses' gains, losses and carryovers under one limit and records the result on the head; a tax unit dependent's return is figured on its own, with the single or separate limit dependent_adjusted_gross_income uses. The unused part of a carryover adds nothing.

ut_ctc_potential: phase-out income is now ut_total_income + tax_exempt_interest_income, summed over the head and spouse (tax_unit_non_dep_add), per (1)(d)'s "claimant ... and the claimant's spouse". The model computes no code 57 addition (ut_additions is the ut_other_additions input alone), so the municipal bond interest is counted once, through tax_exempt_interest_income. The contrib reform ut_ctc_reform copies the baseline's income line, so it changes the same way (its tests enter only ut_total_income, so none changes). The reduction start and rate descriptions now say modified adjusted gross income.

The nontaxable income variable's statute link pointed at page 5; (20) is on pages 3-4 (#10973 asked to check it).

Who is affected

  • Utah homeowner's/renter's relief (property tax relief, not income tax): household income rises by the amount added, so relief falls or ends, for households with SSI, nontaxable disability insurance benefits or a capital loss carryover used in the year. It falls, so relief rises, for households with alimony under a pre-2019 divorce instrument (or no divorce year entered), which no longer counts twice. SSI reaches the most claimants: an aged claimant with little other income gets most of a full SSI benefit (11,604 in 2025) added.
  • Utah child tax credit (income tax): filers with a child under 6, modified AGI above the threshold, and Utah subtractions (U.S. government interest, other subtractions) or a state tax refund in AGI get a smaller credit, so their Utah income tax rises by up to 10% of those amounts. A child's own tax-exempt interest no longer reduces the claimant's credit (rare).

Tests

Moved from #10812:

  • ut_ctc_audit.yaml: its one case to ut_ctc.yaml ("Utah subtractions do not reduce modified adjusted gross income", 300).
  • ut_homeowner_renter_relief_audit.yaml: D1, D2, D4, D5 and D6 to ut_homeowner_renter_relief.yaml as Cases 13 (household income 10,000, renter's credit 1,140), 14 (10,000), 15 (6,000), 17 (16,000) and 19 (14,000). Each except D4 now enters ssi: 0, so it still isolates its own rule now that SSI counts. D1 and D2 also assert the new alimony variable and D6 the new carryforward variable. D3 (an adult dependent's wages) passes on main through Count strike benefits other than gifts in federal gross income #10727 and ut_homeowner_renter_relief_strike_benefits.yaml already has the same case, so it isn't moved.

Added, with the arithmetic in each comment:

  • ut_homeowner_renter_relief.yaml Case 16: SSI the model computes. 6,000 of Social Security gives SSI of (967 − (500 − 20)) × 12 = 5,844; household income 6,000 + 5,844 = 11,844; renter's credit 9.5% × 6,000 = 570. Case 18: partly taxable disability benefits count at their gross amount (AGI 14,000 + 2,000 = 16,000).
  • New ut_homeowner_renter_relief_loss_carryforward.yaml, 6 cases: a carryover used against the year's gain (4,000); a carryover above the gain, claimed up to the loss limit (1,000 + 3,000 = 4,000); the year's own loss exactly at the 3,000 limit (0); the year's own loss 1,000 under the limit (1,000); a joint return netting one spouse's carryover against the other's gain (4,000, on the head); a dependent's own return (3,000; household income 30,000 − 3,000 + 3,000 = 30,000).
  • ut_ctc.yaml: U.S. government interest end to end (AGI 45,000, subtractions 5,000, credit 800); a state tax refund end to end (AGI 45,000, credit 800); MAGI exactly at the 43,000 threshold with 3,000 of subtractions (1,000) and 10 above it (999); a dependent's own tax-exempt interest left out (1,000).

Existing tests changed, inputs only (no expected value changed). Eight cases give a person aged 66 or older little or no income of their own and no SSI input, so the model now computes SSI for them and counts it. Run w1-ut-a shows the failures:

  • ut_homeowner_renter_relief.yaml Case 7 (relief 1,550 became 1,450: SSI of 7,114 raised household income from 10,000 to 17,114 and the renter's rate from 9.5% to 8.5%), Case 9 (550 became 400) and Case 11 (850 became 400).
  • ut_homeowner_renter_relief_eligible.yaml Case 8 (household income 70,000 became 77,114).
  • ut_homeowner_renter_relief_strike_benefits.yaml: the three one-person nontaxable income cases (0 became 11,604, and 3,000 became 14,604 twice: full SSI of 967 × 12), and "a return's adjusted gross income adds to an adult dependent's own" (35,000 became 46,604).
  • gov/states/tax/property/state_property_tax_refunds.yaml, "A Utah owner who also rents" (the household of Case 7).

Each now enters ssi: 0 for every person, so it tests what it names and its expected values stand. No partner test changed or failed.

Not in this PR

  • Utah relief in households of more than one tax unit (the parent and grandparent flags describe a member's relation to their own filer; Strike benefits: follow-ups from the review of #10727 #10973).
  • Net operating loss carryforwards and short-term capital loss carryovers: (20)(a)(ii) counts them, but the model has no input for either.
  • Utah's state supplement to SSI (ut_ssp): whether it is a Social Security Act payment or Title 35A Chapter 3 public assistance under (24)(c) needs a reading.
  • The TC-40A code 57 addition has no input of its own. A filer who enters it in ut_other_additions gets it counted twice in the CTC's modified AGI; a dedicated input (as ca_out_of_state_municipal_bond_interest is for California) would let the worksheet's line 4 take it out.
  • railroad_benefits stays on the list: the model's AGI holds none of them, so they count once (Strike benefits: follow-ups from the review of #10727 #10973 asked to check).
  • The review lists no other Utah gaps; the adult-dependent case (D3) passes on main through Count strike benefits other than gifts in federal gross income #10727.

axiom: needed (rulespec-us has no us-ut module for Utah Code 59-2a-101 or 59-10-1047; no encoder run was authorized for this PR)

Tested on Modal: run w1-ut-b, 239/239 passed (233 YAML files: the Utah folder, every partner file, the Utah contrib reforms and the changed files; plus test_source_conflicts.py, test_aggregate_list_integrity.py, test_strike_benefits_properties.py, test_strike_benefits_source_reforms.py, test_household_spm_resource_parity.py and test_non_refundable_credit_downstream_consumers.py).

🤖 Generated with Claude Code

@codecov

codecov Bot commented Nov 19, 2026 •

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (8999568) to head (6d52046).
⚠️ Report is 296 commits behind head on main.

Additional details and impacted files
@@            Coverage Diff            @@
##              main    #11182   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            1         4    +3     
  Lines           24        97   +73     
  Branches         0         3    +3     
=========================================
+ Hits            24        97   +73     
Flag Coverage Δ
unittests 100.00% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@DTrim99
DTrim99 requested a review from hua7450 November 20, 2026 04:30

@hua7450 hua7450 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, David. The relief changes match Utah Code 59-2a-101(17), (20) and (24): SSI under (20)(a)(vii), the tax-free part of disability insurance, the capital loss carryover used in the year, and alimony only where AGI leaves it out. The renter's credit (claim year) and homeowner's credit (prior year) each read the right income year. The CTC phase-out income now matches 59-10-1047(1)(d), which is the same in 2024, 2025 and 2026, and drops main's wrong netting of ut_subtractions. The reform mirrors the baseline. No cycles and no new inputs. Two fixes before I approve.

Before merge

  1. Register the other-state bond interest reading. Read literally, 59-10-1047(1)(d) counts this interest twice: once under (ii) as interest not in AGI, and again under (iii) as the 59-10-114(1)(e) addition. The TC-40 worksheet takes the code 57 amount out on line 4 and counts it once, and the model follows the worksheet. That is a sensible reading, but CONTRIBUTING ("Source conflicts", step 1) asks the PR that takes it to add a source_conflicts/candidates.yaml entry. The entry should quote statute (ii) and (iii) against worksheet lines 3-6 and give the basis: no double counting, and agency practice.

  2. Nine existing cases had inputs changed. ssi: 0 was added so that main's expected values still pass:

    • Cases 7, 9 and 11 in ut_homeowner_renter_relief.yaml
    • Case 8 in ut_homeowner_renter_relief_eligible.yaml
    • four cases in ut_homeowner_renter_relief_strike_benefits.yaml
    • the Utah owner-renter case in state_property_tax_refunds.yaml

    Each is an older person with little income, who would get SSI under the model's own rules, so the changed result for those households is now asserted only in new Case 16. Please keep main's inputs and update the outputs with a one-line reconciliation, for example "10,000 + SSI 7,114 = 17,114". If you want the zero-SSI versions too, add them as new cases.

Optional

  • SSI comment in sources.yaml. Say that (20)(b)(vi) "social security disability income" means Title II SSDI, so SSI paid on account of disability still counts under (a)(vii).
  • Cases 14b, 19b and 19c each paste the full federal gross income source list with one entry removed. That copy will drift when main adds a source, so a comment noting when it was copied would help.

@DTrim99
DTrim99 force-pushed the ut-relief-ctc-income branch from 4e654a3 to 0a6dd14 Compare November 20, 2026 21:51
@DTrim99

DTrim99 commented Nov 20, 2026

Copy link
Copy Markdown
Collaborator Author

Thanks, Ziming. Fixed in 0a6dd14, rebased on main.

Fixed (your "Before merge" items)

  1. The other-state bond interest reading is registered. New source_conflicts/candidates.yaml entry us-ut:ctc-modified-agi-other-state-bond-interest, following the pattern of us-co:federal-deduction-addback-net-of-state-tax-addback.
    • The statute side: it quotes Utah Code 59-10-1047(1)(d)(i)-(iii), the same in the 2024, 2025 and 2026 editions, and 59-10-114(1)(e).
    • The form side: TC-40 worksheet lines 3-7 (2024 and 2025 instructions, p. 24), lines 5-6 (p. 10) and code 57 (p. 18).
    • The reading: the model follows the worksheet and counts the interest once. The basis is no double counting and agency practice.
    • The example: a 2025 joint return with one child gets a $1,000 credit under the worksheet and $800 under the literal reading. The model gives $1,000.
  2. Main's inputs are back in the nine cases. The ssi: 0 inputs are gone, so against main only outputs and comments differ. Each output now has a one-line reconciliation. In 2025, SSI is 11,604 − (10,000 − 240 − 780) / 2 = 7,114.
    • ut_homeowner_renter_relief.yaml:
      • Case 7: 10,000 + 7,114 = 17,114, renter's credit 850, homeowner's 600.
      • Case 9: 32,172.
      • Case 11: 2 × 17,114 = 34,228.
    • ut_homeowner_renter_relief_eligible.yaml Case 8: 77,114, still not eligible.
    • The four strike-benefit cases: 11,604, 14,604 (twice) and 46,604.
    • The Utah case in state_property_tax_refunds.yaml: 1,450 / 850 / 600, refund 850.

Skipped (optional, not in this round):

  • the SSDI note in sources.yaml;
  • a "copied on" comment on the pasted source lists in Cases 14b, 19b and 19c.

Verification:

  • After the rebase, the four changed test files (60) and the source-conflict tests (251) pass locally. Before it, the Utah relief and CTC directories passed (87).
  • An independent check matched the quotes against the live statute and TC-40 pages and confirmed every reconciliation.

Already on main, noted only: the model's SSI does not count union strike benefits. Under 20 CFR 416.1120 they are unearned income ("all income that is not earned income"), and no exclusion in 416.1124 applies. So the strike-benefit cases get full SSI. With strike pay counted, the first case would get about 1,844. The outputs assert what the model does today.

@DTrim99
DTrim99 requested a review from hua7450 November 20, 2026 21:51

@hua7450 hua7450 left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, David. Both items are fixed. The us-ut:ctc-modified-agi-other-state-bond-interest entry is complete, and its quotes, page locators and statute editions match the sources. Against main, the nine cases now change only outputs and comments, and every new output checks by hand (SSI 7,114 in 2025; 1,450/400/400; 77,114; 11,604, 14,604 twice and 46,604). One gap before I approve.

Before merge

  1. A test case for the registered reading. CONTRIBUTING ("Source conflicts", step 1) says the ruling PR "also adds ... the YAML test cases for its reading". No case in ut_ctc.yaml exercises it. Every case with tax_exempt_interest_income sets ut_total_income as an input, which bypasses ut_additions. Please add a 2025 case built from the entry's example: a joint return, $50,000 of employment_income, a child aged 3, $3,000 of tax_exempt_interest_income and no ut_other_additions. It should assert ut_total_income 50,000 and ut_ctc_potential 1,000, with a comment citing the candidate id.

    While you're there, please fix the example's wording too. $1,000 and $800 are the amounts before the liability limit. That household's Utah tax after the taxpayer credit is about $330, and the credit is nonrefundable, so both readings allow about $330. Either say "before the liability limit" or have the example assert ut_ctc_potential.

Optional

  • PR body. It still says the existing tests changed "inputs only" with ssi: 0 added. It leaves out the register entry and reports the earlier test run.
  • Strike benefits and SSI (follow-up). The model's SSI doesn't count union strike benefits, though 20 CFR 416.1120 makes them unearned income, and three changed cases rest on that. This is already on main, and issue 10973 doesn't cover it, so it needs its own issue. The 46,604 case also enters AGI on the return, which SSI doesn't read; a one-line comment would explain the full SSI there.
  • The entry's sources list the 2025 TC-40 instructions three times. Other register entries do the same, so this is optional.
  • From the first round, still open: the SSDI note in sources.yaml, and the copied source lists in Cases 14b, 19b and 19c.

DTrim99 and others added 5 commits November 20, 2026 22:26
Homeowner's/renter's relief nontaxable income (Utah Code 59-2a-101(20)(a))
now counts only alimony that federal AGI leaves out, and adds SSI,
nontaxable disability insurance benefits ("loss of time" insurance) and
the capital loss carryover claimed in the year. The Utah child tax credit
phases out on modified adjusted gross income under 59-10-1047(1)(d):
total income plus the claimant's and spouse's tax-exempt interest, with
no Utah subtractions or state tax refund taken out. The contrib reform
uses the same income.

Moves the Utah audit cases from #10812 into the real test files, adds
cases for each rule, and enters no SSI in the existing cases that assumed
none.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…income

Covers the alimony fallback when gross income doesn't list
taxable_alimony_income, and the carryover's zero when gross income doesn't
list capital_gains, which codecov flagged as uncovered patch lines.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ncome

Covers the false side of the non_sch_d_capital_gains check in
ut_homeowner_renter_relief_loss_carryforward, the remaining partial line.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…eep main's inputs in the SSI cases

Record the Utah relief income bond-interest interpretation in
source_conflicts/candidates.yaml and restore main's household inputs in
the SSI test cases for the Utah homeowner/renter relief and the
state property tax refund tests.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…fy the liability limit

Add a Utah CTC test case for the TC-40A code 57 bond interest reading,
state in the source-conflict candidate that both readings give the same
credit after the nonrefundable liability limit, and note in the
homeowner/renter relief test that SSI reads the head's own income.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@DTrim99
DTrim99 force-pushed the ut-relief-ctc-income branch from 0a6dd14 to 6d52046 Compare November 21, 2026 04:31

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants