Repository navigation
Treat income exactly at 400% FPL as premium tax credit eligible - #10312
Conversation
26 U.S.C. 36B(c)(1)(A) covers household income that "does not exceed 400 percent" of the poverty line, and Form 8962 Worksheet 2 compares income with four times the poverty line in dollars before truncating the percentage. The income scale's 400% bracket starts at 4.00, so evaluating it on the truncated aca_magi_fraction excluded a household at exactly 400%. aca_ptc_income_eligible keeps such income on the eligible side and is read by is_aca_ptc_eligible and the Basic Health Program payment. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Read the exact ratio from aca_fpg, the poverty line the credit's fraction has divided by since #10041, and point is_aca_ptc_eligible_before_premium at aca_ptc_income_eligible. Keep territories without a guideline out of the right-closed lookup. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
hua7450
left a comment
There was a problem hiding this comment.
Thanks, David. The policy and code both check out. 26 U.S.C. 36B(c)(1)(A) says "does not exceed 400 percent", and Form 8962 Worksheet 2 compares household income with four times the poverty line in dollars before dropping decimals, so income exactly at 400% qualifies while 400.003% to 400.99% stays out. aca_ptc_income_eligible does exactly that, can't remove anyone's eligibility, and the territory guard keeps the infinite fraction out of the top bracket. The 100% floor is unaffected, the Basic Health Program edit only reuses the new variable, and the property test is justified. Approving.
Optional:
- Affected years. The changelog and PR body say the fix applies to 2026 and later, but it applies in every year the cap applies, 2014 to 2020 as well. The behavior is right; only the wording.
- Cap-year reference. The variable cites only the 2025 Form 8962 instructions, a year without the cap. The 2020 instructions state the ceiling as an eligibility rule ("at least 100% but no more than 400%"): https://www.irs.gov/pub/irs-prior/i8962--2020.pdf#page=8
- Tests. Cases 6 and 9 and the
@exampleblock cover the same scenario. - PR body. The coverage-gap test is in
is_aca_marketplace_enrollee, notpays_aca_premium.
Series-wide (follow-up, not for this PR). The exact ratio max_(aca_magi, 0) / aca_fpg is now computed inline in each PR of this series, plus two California copies on main, and only 10312 guards a zero poverty line (territories, where aca_fpg is 0; the formula runs before the state mask, so a mixed simulation emits NumPy warnings, though outputs are unaffected). One guarded tax-unit variable for the untruncated ratio, read by every gate, would be a good follow-up PR once these merge.
Done: the fix, YAML cases, a Hypothesis differential test, and the update onto current
main(below). Still pending: full-file impact, Axiom parity line, GPT-6.1 Sol review.Update onto current main (2026-11-18)
Merged
main(no force-push; the original commit stays). Since this PR was written, #10041 moved the credit's income conditions intois_aca_ptc_eligible_before_premiumand madeaca_magi_fractiondivide byaca_fpg, and #10860 added the Alaska and Hawaii guidelines. So:aca_ptc_income_eligiblenow divides byaca_fpg, in float64, the same poverty line asaca_magi_fraction.is_aca_ptc_eligible_before_premiumreadsaca_ptc_income_eligible;is_aca_ptc_eligiblekeepsmain's form.aca_fpginstead oftax_unit_fpg.Run on Modal at
f3543f2d4b: every YAML file undergov/acaandgov/hhs/basic_health_program, every state YAML file that reads the credit or ACA MAGI (140 files), and the property test all passed.Bug
26 U.S.C. 36B(c)(1)(A) makes an applicable taxpayer one whose household income "equals or exceeds 100 percent but does not exceed 400 percent" of the poverty line. Form 8962 Worksheet 2 asks "Is the amount on line 1 more than the amount on line 3?" (income against 4 x the poverty line, in dollars) and only then drops the decimals for line 5. The 2020 edition, the last year the cap applied, has the same worksheet.
gov.aca.ptc_income_eligibilityis asingle_amountscale whose 400% bracket (false from 2026) starts at 4.00, evaluated on the truncatedaca_magi_fraction. So a household exactly at 400% fell into the ineligible bracket. Above 400% the truncation does no harm here: 400.003% truncates to 4.00, which is already ineligible.Executed on
main(645fcd3; these files are unchanged at 740e1eb), single filer in Texas, 2026, 2025 one-person guideline $15,650:aca_magi_fractionaca_ptcat $62,600 goes from $0 to $5,765.04 on a $12,000 benchmark (9.96% applicable percentage).Fix
aca_ptc_income_eligible(tax unit, bool) carries the income test. It evaluates the scale on the truncated fraction, as before, and additionally right-closed when the exact ratiomax(aca_magi, 0) / aca_fpgdoes not exceed the truncated fraction (truncation dropped nothing). That reproduces Worksheet 2: a ceiling ("does not exceed") keeps income exactly at it; the floor ("equals or exceeds") already did.is_aca_ptc_eligible_before_premiumand the Basic Health Program payment (basic_health_program, which reused the scale) read the new variable. The coverage-gap test inis_aca_marketplace_enrolleereads only below the 100% floor, where nothing changes, so it is untouched.aca_ptc_700_fpl_cliff,aca_ptc_additional_bracket,aca_ptc_simplified_bracket) replaceis_aca_ptc_eligiblewith their own scales and are untouched.A supplied
aca_magi_fractionwith no income is read as a whole percentage, so tests that supply only the fraction keep their results, except a supplied fraction exactly at 4.00, which is now eligible.Invariants
For every input (exact arithmetic on the float32 values the model receives):
aca_ptc_income_eligibleholds if and only if1 <= r <= 4(2026, 2027) orr >= 1(2025, 36B(c)(1)(E)), wherer = max(MAGI, 0) / prior-year guideline. Tested by a Hypothesis differential test against afractions.Fractionreading of the statute (bounds transcribed from the statute, not the parameters), with incomes in cents within $2 of each edge.Tests
aca_ptc_income_eligible.yaml(new): 99.994%, 100%, 399.994%, 400%, 400.003%, 400.994% and 401% in 2026; 400% and 400.003% in 2027; 400.003% in 2025 (no cap); a two-person unit at $84,600 (400% of $21,150) and $84,601; supplied fractions 3.99 and 4.01.is_aca_ptc_eligible.yaml,aca_ptc.yaml: end-to-end cases at $62,600 (eligible, $5,765.04 credit) and $62,600.47 (ineligible, $0).tests/unit/test_aca_ptc_income_eligible_properties.py: the scale matches the statute, and the model matches the statute's exact band for 60 generated batches of 64 tax units per year (2025-2027), one reused simulation per year. Local cost (macOS arm64, not CI): 6 tests in 5.2 s after a ~16 s model import that the Rest Pythonremaininggroup already pays; 1.14 GB peak resident memory.pays_aca_premium.yaml, 73 passed; the property test, 6 passed.Impact
Pending: full-file microsimulation via the US hub impact queue, 2026 and 2027.
Coordination
Part of an audit of every gate that compares the truncated
aca_magi_fractionwith an income limit, which began with #10305 (California); the table and the sibling PRs are in #10329.Resolves #9264 with a corrected scope. That issue proposes moving the threshold to 4.01 on the premise that a household at 400.x% is eligible under Form 8962. Worksheet 2 enters 401 for any income "more than" 4 x the poverty line, so 400.x% is ineligible and main already got it right; only exactly 400% was misclassified, which this PR fixes.
axiom: pending
🤖 Generated with Claude Code