MaxGhenis · Oct 20, 2026
diff --git a/changelog.d/ma-connectorcare-exact-fpl-ratio.fixed.md b/changelog.d/ma-connectorcare-exact-fpl-ratio.fixed.md
@@ -0,0 +1 @@
+Decide Massachusetts ConnectorCare's income ceiling and Plan Type bands on the exact ratio of ACA MAGI to the prior-year poverty guideline instead of the truncated `aca_magi_fraction`. 956 CMR 12.04(3)(b) puts income "at or below 150%" in Plan Type 2A, "over 150% and not over 200%" in 2B, and so on up to 3C, "not over 400%"; truncation put incomes from 150.003% to 150.994% (and the same slivers above 200%, 250% and 300%) in the cheaper band below. The 400% ceiling changes results only under a reform that extends the federal credit above 400%.
diff --git a/policyengine_us/parameters/gov/states/ma/hic/connector_care/enrollee_premium.yaml b/policyengine_us/parameters/gov/states/ma/hic/connector_care/enrollee_premium.yaml
@@ -16,41 +16,44 @@ metadata:
       href: https://www.mahealthconnector.org/wp-content/uploads/amended-956CMR12.00.pdf#page=13
 
 brackets:
-  # Household income measured as ACA MAGI as a share of the federal poverty line.
-  # Band boundaries are lower-exclusive / upper-inclusive per 956 CMR 12.04(3)(b),
-  # so thresholds use the "x.0001" convention matching the published "150.1%"
-  # banding. The schedule is a flat dollar step function decoupled from the
-  # affordability percent-of-income schedule (Board-set, 956 CMR 12.12(9)); every
-  # band is $0 before the PY2026 effective date.
-  # Plan Type 2A, 100-150% FPL -> $0 per person per month.
+  # Household income measured as ACA MAGI as a share of the prior-year federal
+  # poverty guideline, the exact ratio. Band boundaries are lower-exclusive /
+  # upper-inclusive per 956 CMR 12.04(3)(b) ("at or below 150%", "over 150% and
+  # not over 200%"), so each threshold is the band's lower edge and
+  # ma_connector_care evaluates the scale with right-closed brackets: a ratio
+  # exactly at a threshold stays in the band below it. The schedule is a flat
+  # dollar step function decoupled from the affordability percent-of-income
+  # schedule (Board-set, 956 CMR 12.12(9)); every band is $0 before the PY2026
+  # effective date.
+  # Plan Type 2A, 100% to 150% FPL -> $0 per person per month.
   - threshold:
       2026-01-01: 1
     amount:
       0000-01-01: 0
       2026-01-01: 0
-  # Plan Type 2B, 150.1-200% FPL -> $53.
+  # Plan Type 2B, over 150% to 200% FPL -> $53.
   - threshold:
-      2026-01-01: 1.5001
+      2026-01-01: 1.5
     amount:
       0000-01-01: 0
       2026-01-01: 53
-  # Plan Type 3A, 200.1-250% FPL -> $103.
+  # Plan Type 3A, over 200% to 250% FPL -> $103.
   - threshold:
-      2026-01-01: 2.0001
+      2026-01-01: 2
     amount:
       0000-01-01: 0
       2026-01-01: 103
-  # Plan Type 3B, 250.1-300% FPL -> $152.
+  # Plan Type 3B, over 250% to 300% FPL -> $152.
   - threshold:
-      2026-01-01: 2.5001
+      2026-01-01: 2.5
     amount:
       0000-01-01: 0
       2026-01-01: 152
-  # Plan Type 3C, 300.1-400% FPL -> $235. This tier is a pilot extended through
+  # Plan Type 3C, over 300% to 400% FPL -> $235. This tier is a pilot extended through
   # FY2026 in the state budget and may lapse in out-years; it is carried forward
   # without end-dating as a current-law assumption.
   - threshold:
-      2026-01-01: 3.0001
+      2026-01-01: 3
     amount:
       0000-01-01: 0
       2026-01-01: 235
diff --git a/policyengine_us/tests/policy/baseline/gov/states/ma/hic/connector_care/integration.yaml b/policyengine_us/tests/policy/baseline/gov/states/ma/hic/connector_care/integration.yaml
@@ -1,12 +1,14 @@
 # Integration tests for Massachusetts ConnectorCare (PY2026 premium wrap).
 #
-# These exercise the full chain: aca_magi_fraction (income gate) and
+# These exercise the full chain: the exact ratio of ACA MAGI to the prior-year
+# guideline (income gate and Plan Type band) and
 # is_aca_ptc_eligible (member gate) -> ma_connector_care_member_eligible ->
 # ma_connector_care_eligible -> ma_connector_care (residual subsidy) ->
 # healthcare_benefit_value (household resource aggregation).
 #
-# In Scenarios 1-5 slcsp, aca_ptc, and aca_magi_fraction are supplied directly
-# (slcsp as an annual value) so every expected value is fully hand-verifiable.
+# In Scenarios 1-5 slcsp and aca_ptc are supplied directly (slcsp as an annual
+# value), and income is aca_magi against a prior-year tax_unit_fpg pinned to
+# $10,000, so every expected value is fully hand-verifiable.
 # slcsp is a MONTH variable; supplying it at the YEAR period sets the annual sum.
 # Scenario 6 instead derives aca_magi_fraction, aca_ptc, and is_aca_ptc_eligible
 # from raw employment_income (only slcsp is supplied), per the Code Integrity note
@@ -20,12 +22,12 @@
 #
 # Amount formula recap:
 #   n_members      = count of ma_connector_care_member_eligible (= is_aca_ptc_eligible)
-#   per_person_mo  = enrollee_premium.calc(aca_magi_fraction)   # single_amount bracket
+#   per_person_mo  = enrollee_premium.calc(ACA MAGI / prior-year FPG, right=True)
 #   contribution   = 12 * per_person_mo * n_members
 #   subsidy        = max(0, slcsp_annual - aca_ptc - contribution)
 #   amount         = where(ma_connector_care_eligible, subsidy, 0)
-# Bracket (per person per month): [1.0,1.5001)->0, [1.5001,2.0001)->53,
-#   [2.0001,2.5001)->103, [2.5001,3.0001)->152, [3.0001,inf)->235.
+# Bracket (per person per month, right-closed): [1.0,1.5]->0, (1.5,2.0]->53,
+#   (2.0,2.5]->103, (2.5,3.0]->152, (3.0,inf)->235.
 
 - name: Scenario 1, single adult at 175% FPL, band 2B ($53/mo), positive residual.
   # n_members=1, band 2B -> per_person 53. contribution = 12*53 = 636.
@@ -40,7 +42,9 @@
     tax_units:
       tax_unit:
         members: [person1]
-        aca_magi_fraction: 1.75
+        aca_magi: 17_500
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 7_200
         aca_ptc: 5_000
         assigned_aca_ptc: 5_000
@@ -68,7 +72,9 @@
     tax_units:
       tax_unit:
         members: [person1]
-        aca_magi_fraction: 1.4
+        aca_magi: 14_000
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 6_000
         aca_ptc: 5_500
         assigned_aca_ptc: 5_500
@@ -100,7 +106,9 @@
     tax_units:
       tax_unit:
         members: [person1, person2]
-        aca_magi_fraction: 1.75
+        aca_magi: 17_500
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 7_200
         aca_ptc: 5_000
         assigned_aca_ptc: 5_000
@@ -129,7 +137,9 @@
     tax_units:
       tax_unit:
         members: [person1]
-        aca_magi_fraction: 3.5
+        aca_magi: 35_000
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 8_000
         aca_ptc: 4_000
         assigned_aca_ptc: 4_000
@@ -159,7 +169,9 @@
     tax_units:
       tax_unit:
         members: [person1]
-        aca_magi_fraction: 3.5
+        aca_magi: 35_000
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 5_000
         aca_ptc: 3_000
         assigned_aca_ptc: 3_000
@@ -221,7 +233,9 @@
     tax_units:
       tax_unit:
         members: [person1]
-        aca_magi_fraction: 1.75
+        aca_magi: 17_500
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 7_200
         aca_ptc: 5_000
         assigned_aca_ptc: 5_000
@@ -261,7 +275,9 @@
     tax_units:
       tax_unit:
         members: [person1, person2, person3, person4, person5]
-        aca_magi_fraction: 3.5
+        aca_magi: 35_000
+        tax_unit_fpg:
+          2025: 10_000
         slcsp: 20_000
         aca_ptc: 0
         assigned_aca_ptc: 0
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		Decide Massachusetts ConnectorCare's income ceiling and Plan Type bands on the exact ratio of ACA MAGI to the prior-year poverty guideline instead of the truncated `aca_magi_fraction`. 956 CMR 12.04(3)(b) puts income "at or below 150%" in Plan Type 2A, "over 150% and not over 200%" in 2B, and so on up to 3C, "not over 400%"; truncation put incomes from 150.003% to 150.994% (and the same slivers above 200%, 250% and 300%) in the cheaper band below. The 400% ceiling changes results only under a reform that extends the federal credit above 400%.