Count farm rent and leave out passive income that is also earned in the EITC investment income test (26 U.S.C. 32(i)(2)(E))

changelog.d/eitc-passive-earned-overlap-farm-rent.fixed.md

@@ -0,0 +1 @@
+The EITC investment income test now includes farm rental income and losses in the passive basket, and leaves out passive partnership income or loss that is also earned income (26 U.S.C. 32(i)(2)(E)), entered through the new eitc_passive_income_also_in_earned_income input.

policyengine_us/tests/core/test_dependent_eitc_investment_income_invariants.py

@@ -1,37 +1,44 @@
-"""Invariants for dependents' income in the filer's EITC investment income test.
+"""Invariants for the EITC investment income test.
 
 26 USC 32(i)(1) denies the credit when the "disqualified income of the
 taxpayer" exceeds the limit. A tax unit dependent's interest, dividends,
 capital gains, rents and passive income are on the dependent's own return, and
 Publication 596 Worksheet 1 picks up a child's interest and dividends only
 through a Form 8814 election, which the model does not implement. So
 `eitc_relevant_investment_income` counts the head and spouse only, as
-`irs_gross_income` does for AGI.
+`irs_gross_income` does for AGI. 32(i)(2)(E) determines passive income and
+losses without regard to amounts included in earned income, so the passive
+basket removes the signed `eitc_passive_income_also_in_earned_income`, and it
+includes farm rental income, which Worksheet 1 lines 11 and 12 name.
 
 Hypothesis draws batches of tax units (single, head of household with
 dependents, joint with and without dependents), and a seeded population of 200
 such units adds breadth. Dependents' capital gains are sometimes $16 million
 or more, where float32 rounding of a tax unit total would show. Each batch runs
 as one vectorized simulation three times: as drawn; with the dependents'
 investment income inputs set to zero; and with $1,000 more of the head's
-interest. For every tax unit:
+interest and different partnership self-employment earnings for the head and
+spouse. For every tax unit:
 
 1. The dependents' investment income, of any size or sign, never changes the
    filer's `eitc_relevant_investment_income`, `eitc_investment_income_eligible`,
    `eitc`, `filer_loss_limited_net_capital_gains` or AGI.
 2. Differential: `eitc_relevant_investment_income` equals an independent numpy
    computation of Worksheet 1 over the head and spouse: interest, tax-exempt
    interest and dividends, plus their Schedule D gain with capital gain
-   distributions floored at zero (line 7), plus their net rental and passive
-   income floored at zero (line 13).
+   distributions floored at zero (line 7), plus their net rental, farm rental
+   and passive income less the signed part also in earned income, floored at
+   zero (line 13).
 3. Bounds: it is at least the filers' interest and dividends (drawn
    nonnegative), so never negative, and $1,000 more of the head's interest
-   raises it by exactly $1,000.
+   raises it by exactly $1,000 even though the filers' partnership
+   self-employment earnings change too: only the explicit overlap input moves
+   passive income out of the basket.
 4. For tax units without dependents and without a negative distributions
-   input, it equals the previous all-member formula, so leaving dependents out
-   changes nothing for them. (With filer gains in the drawn range. Past about
-   $16.8 million the previous formula's float32 tax unit total could be off by
-   a few dollars, which the direct sum corrects.)
+   input, it equals the same worksheet summed over every member, so leaving
+   dependents out changes nothing for them. (With filer gains in the drawn
+   range. Past about $16.8 million the previous formula's float32 tax unit
+   total could be off by a few dollars, which the direct sum corrects.)
 
 The formula reads the head's and spouse's person-level gains, never the tax
 unit's `net_capital_gains`; a YAML case pins that a supplied tax unit amount is
@@ -57,7 +64,13 @@
 DISTRIBUTIONS = "non_sch_d_capital_gains"
 # passive_partnership_s_corp_income is the passive subset of
 # partnership_s_corp_income; both are drawn together below.
-PASSIVE_INPUTS = ["rental_income", "passive_partnership_s_corp_income"]
+PASSIVE_INPUTS = [
+    "rental_income",
+    "farm_rent_income",
+    "passive_partnership_s_corp_income",
+]
+OVERLAP = "eitc_passive_income_also_in_earned_income"
+SE_EARNINGS = "partnership_self_employment_net_earnings"
 # The filer's amounts, which a dependent's income must not change.
 FILER_OUTPUTS = [
     "eitc_relevant_investment_income",
@@ -68,8 +81,10 @@
 ]
 # Sums every member, for the previous formula in property 4.
 OUTPUTS = FILER_OUTPUTS + ["net_capital_gains"]
-# More of the head's interest, in the third run.
+# More of the head's interest, and different partnership self-employment
+# earnings for the head and spouse, in the third run.
 INTEREST_SHIFT = 1_000.0
+SE_EARNINGS_SHIFT = 5_000.0
 
 nonnegative = st.one_of(st.just(0.0), st.integers(1, 8_000).map(float))
 signed = st.one_of(st.just(0.0), st.integers(-12_000, 12_000).map(float))
@@ -89,8 +104,20 @@ def investment_amounts(draw, dependent=False):
         # the model floors each person's amount at zero.
         DISTRIBUTIONS: draw(st.one_of(nonnegative, st.integers(-3_000, -1).map(float))),
         "rental_income": draw(signed),
+        "farm_rent_income": draw(signed),
         "partnership_s_corp_income": passive,
         "passive_partnership_s_corp_income": passive,
+        # None, all, half or an unrelated signed amount of the passive share
+        # is also earned income; the formula removes exactly this amount.
+        OVERLAP: draw(
+            st.one_of(
+                st.just(0.0),
+                st.just(passive),
+                st.just(float(round(passive / 2))),
+                signed,
+            )
+        ),
+        SE_EARNINGS: draw(signed),
     }
 
 
@@ -126,13 +153,22 @@ def capital():
         return some(-12_000, 12_000)
 
     passive = some(-12_000, 12_000)
+    overlap = rng.choice(["none", "all", "half", "other"])
     return {
         **{name: some(1, 8_000) for name in PORTFOLIO_INPUTS},
         **{name: capital() for name in CAPITAL_INPUTS},
         DISTRIBUTIONS: some(-3_000, 8_000),
         "rental_income": some(-12_000, 12_000),
+        "farm_rent_income": some(-12_000, 12_000, 0.3),
         "partnership_s_corp_income": passive,
         "passive_partnership_s_corp_income": passive,
+        OVERLAP: {
+            "none": 0.0,
+            "all": passive,
+            "half": float(round(passive / 2)),
+            "other": some(-12_000, 12_000),
+        }[overlap],
+        SE_EARNINGS: some(-12_000, 12_000),
     }
 
 
@@ -161,7 +197,7 @@ def _seeded_units(n=200):
     return units
 
 
-def _situation(units, year, *, zero_dependents=False, interest_shift=0.0):
+def _situation(units, year, *, zero_dependents=False, shift=False):
     people, groups = {}, {"tax_units": {}, "households": {}, "marital_units": {}}
     for i, u in enumerate(units):
         head = f"head_{i}"
@@ -171,10 +207,10 @@ def _situation(units, year, *, zero_dependents=False, interest_shift=0.0):
             "is_tax_unit_dependent": False,
             "employment_income": u["head_wages"],
             **u["head"],
-            "taxable_interest_income": (
-                u["head"]["taxable_interest_income"] + interest_shift
-            ),
         }
+        if shift:
+            people[head]["taxable_interest_income"] += INTEREST_SHIFT
+            people[head][SE_EARNINGS] += SE_EARNINGS_SHIFT
         members, couple = [head], [head]
         if u["spouse"] is not None:
             spouse = f"spouse_{i}"
@@ -184,6 +220,8 @@ def _situation(units, year, *, zero_dependents=False, interest_shift=0.0):
                 "is_tax_unit_dependent": False,
                 **u["spouse"],
             }
+            if shift:
+                people[spouse][SE_EARNINGS] -= SE_EARNINGS_SHIFT
             members.append(spouse)
             couple.append(spouse)
         groups["marital_units"][f"couple_{i}"] = {"members": couple}
@@ -247,7 +285,7 @@ def _unit_sum(run, values):
 def _check(units, year):
     run = _run(units, year)
     zeroed = _run(units, year, zero_dependents=True)
-    shifted = _run(units, year, interest_shift=INTEREST_SHIFT)
+    shifted = _run(units, year, shift=True)
     filer = ~_by_person(units, "is_dependent")
     assert np.array_equal(~run["is_dependent"], filer)
     has_dependent = _unit_sum(run, (~filer).astype(float)) > 0
@@ -267,8 +305,23 @@ def filer_sum(names):
         run, filer * np.maximum(0, _by_person(units, DISTRIBUTIONS))
     )
     capital = np.maximum(0, filer_sum(CAPITAL_INPUTS) + distributions)
-    passive = np.maximum(0, filer_sum(PASSIVE_INPUTS))
+    passive = np.maximum(0, filer_sum(PASSIVE_INPUTS) - filer_sum([OVERLAP]))
     expected = portfolio + capital + passive
+    # Units whose eligibility turns on farm rent alone, or on the overlap input
+    # alone, so the seeded population shows each term is exercised.
+    eligible = expected <= run["limit"]
+    without_farm_rent = (
+        portfolio
+        + capital
+        + np.maximum(
+            0,
+            filer_sum([n for n in PASSIVE_INPUTS if n != "farm_rent_income"])
+            - filer_sum([OVERLAP]),
+        )
+    )
+    without_overlap = portfolio + capital + np.maximum(0, filer_sum(PASSIVE_INPUTS))
+    farm_rent_decides = eligible != (without_farm_rent <= run["limit"])
+    overlap_decides = eligible != (without_overlap <= run["limit"])
     investment_income = run["eitc_relevant_investment_income"]
     np.testing.assert_allclose(investment_income, expected, atol=TOLERANCE)
     np.testing.assert_array_equal(
@@ -277,7 +330,8 @@ def filer_sum(names):
     )
     assert not (run["eitc"][expected > run["limit"]] > 0).any()
 
-    # 3. Bounds and an exact shift in the filer's interest.
+    # 3. Bounds, and an exact shift in the head's interest while the filers'
+    # partnership self-employment earnings change too.
     assert (investment_income >= portfolio - TOLERANCE).all()
     assert (investment_income >= -TOLERANCE).all()
     np.testing.assert_allclose(
@@ -286,8 +340,8 @@ def filer_sum(names):
         atol=TOLERANCE,
     )
 
-    # 4. No dependents and no negative distributions input: the previous
-    # all-member formula, unchanged.
+    # 4. No dependents and no negative distributions input: the same worksheet
+    # summed over every member, unchanged.
     no_negative_distributions = (
         _unit_sum(run, (_by_person(units, DISTRIBUTIONS) < 0).astype(float)) == 0
     )
@@ -299,13 +353,18 @@ def filer_sum(names):
             run["net_capital_gains"] + _unit_sum(run, _by_person(units, DISTRIBUTIONS)),
         )
         + np.maximum(
-            0, _unit_sum(run, sum(_by_person(units, n) for n in PASSIVE_INPUTS))
+            0,
+            _unit_sum(
+                run,
+                sum(_by_person(units, n) for n in PASSIVE_INPUTS)
+                - _by_person(units, OVERLAP),
+            ),
         )
     )
     np.testing.assert_allclose(
         investment_income[unchanged], all_member[unchanged], atol=TOLERANCE
     )
-    return has_dependent, unchanged
+    return has_dependent, unchanged, farm_rent_decides, overlap_decides
 
 
 # A batch's cost is mostly per-variable overhead, so each example is a large
@@ -326,7 +385,11 @@ def test_dependent_investment_income_stays_off_the_filers_eitc_test(units):
 
 @pytest.mark.parametrize("year", [2021, 2025])
 def test_seeded_population(year):
-    has_dependent, unchanged = _check(_seeded_units(), year)
+    has_dependent, unchanged, farm_rent_decides, overlap_decides = _check(
+        _seeded_units(), year
+    )
     # The population exercises both branches of each property.
     assert has_dependent.sum() > 50
     assert unchanged.sum() > 20
+    assert farm_rent_decides.sum() > 2
+    assert overlap_decides.sum() > 2