Wrong Schedule R initial amount for MFS filers

Open Beginner friendly
  • Is something wrong?
#10,552 0 comments 0 reactions 0 assignees View on GitHub

Maintainers usually reply within 2 days

Nobody has claimed this yet.

Assessment

Difficulty
2/5
Estimated time
1-3 hours
Newbie friendliness
84/100
Issue type
Bug
Clarity
Clearly specified
Activity status
Active
Tech stack
python
Domain
backend

Research direction

Start with policyengine_us/variables/gov/irs/credits/elderly_and_disabled/section_22_income.py, named in the issue, and check the existing tests for Section 22 income and Schedule R. Run the relevant tests, then verify that an MFS filer uses the $3,750 initial amount while single and joint filers retain their amounts. Done when the reported MFS example no longer receives the erroneous credit and the relevant tests pass.

Written by the indexing model from the issue text.

Description

bug

This bug report was generated by Claude Code 2.1.314 running Opus 5.5
with medium effort. The response below was generated after reading
this validation summary.

Summary

PolicyEngine-US (version 2.87.0, used by policyengine-taxsim 3.3.0 at
commit c0d565948af3896e3a6e4d34970f7502d17cdf34) gives a married
filing separately (MFS) filer who is 65 or older a Schedule R
(credit for the elderly or the disabled) initial amount of $5,000
instead of the statutory $3,750. This overstates the credit, and
so understates income tax, for low-income elderly MFS filers.

The law

IRC section 22(c)(2)(A) sets the initial amount at:

  • (i) $5,000 for a single individual or a joint return where only one
    spouse is a qualified individual,
  • (ii) $7,500 for a joint return where both spouses are qualified
    individuals, and
  • (iii) $3,750 for a married individual filing a separate return.

IRC section 22(d) reduces the initial amount by one-half of AGI above
$7,500 (single), $10,000 (joint), or $5,000 (MFS), and the credit is
15% of the reduced amount. The Schedule R instructions (Part I, box
8, and Part III, line 10) give the same $3,750 and $5,000 MFS amounts.

In addition, IRC section 22(e)(1) allows the credit to a married
individual filing separately only if the spouses lived apart for the
whole year. Tax-Calculator records do not record whether MFS spouses
lived apart, so Tax-Calculator (like PE) allows the credit to every
MFS filer aged 65 or older. That limitation is therefore not part of
the difference described here.

Why PE is wrong

PE's section_22_income variable
(policyengine_us/variables/gov/irs/credits/elderly_and_disabled/section_22_income.py)
chooses the initial amount with:

initial_amount = select(
    [
        num_qualifying_individuals == 1,
        num_qualifying_individuals == 2,
        filing_status == filing_status.possible_values.SEPARATE,
    ],
    [
        elderly_disabled.amount.one_qualified,   # 5,000
        elderly_disabled.amount.two_qualified,   # 7,500
        elderly_disabled.amount.separate,        # 3,750
    ],
    default=0,
)

select returns the value for the first condition that is true. An
MFS tax unit has only one person, so an elderly MFS filer satisfies
num_qualifying_individuals == 1 first and gets $5,000; the
SEPARATE branch is never reached. PE's phase-out threshold
parameter is correct (phase_out.threshold.SEPARATE = 5,000).

Putting the SEPARATE condition first would fix the bug.

Note that with the $5,000 MFS threshold, PE's result,
5,000 - (AGI - 5,000) / 2, equals 3,750 - (AGI - 7,500) / 2. So the
output looks as if PE used the correct $3,750 initial amount with the
single-filer $7,500 AGI threshold, but the cause is the initial
amount.

Example tax unit

A 2021 MFS filer aged 72 with $14,000 of wages and no other income,
in the TAXSIM input format used by the PolicyEngine-TAXSIM tool:

taxsimid,year,mstat,page,pwages,idtl
1062,2021,6,72,14000,2

This is unit 1062 of the a21 validation sample in
taxcalc/validation/pe_taxsim/samples/a21.in.csv.gz.

Item Tax-Calculator PolicyEngine
AGI 14,000 14,000
Taxable income (v18) 100 100
Income tax before credits (v19) 10 10
Section 22 amount after phase-out 0 500
Schedule R credit, before tax limit 0 75
Schedule R credit 0 10
fiitax -2,526.33 -2,536.33

The two models agree on every other part of fiitax (the EITC of
$1,136.33 and the $1,400 recovery rebate credit).

Tax-Calculator: 3,750 - (14,000 - 5,000) / 2 = -750, so there is no
credit. PolicyEngine: 15% x (5,000 - (14,000 - 5,000) / 2) = $75,
limited to the $10 tax. At $15,000 of wages both models give no
credit, because PE's amount, 5,000 - 5,000, is also zero.

The PE results were confirmed by running the pinned PE version
directly (elderly_disabled_credit_potential = 75 and
elderly_disabled_credit = 10 at $14,000 of wages; both zero at
$15,000).

cc: @MaxGhenis @PavelMakarchuk

Dominant language
Python
Stars
171
Forks
220
Avg merge
4d 15h
Merged PRs (30d)
283

Getting set up

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

More from PolicyEngine/policyengine-us

All issues in PolicyEngine/policyengine-us

Similar issues

More Python issues

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.