Prefix state ACA spending calibration label with state/ (#779)

Line 820 labeled the per-state ACA spending target
`nation/irs/aca_spending/{state}` — starting with `nation/` even
though every row is specific to one state. `reweight()` classifies
targets via `columns.str.startswith("nation/")` and divides the
normalisation weight by the count of national versus state
targets, so this label was counted as a national target, diluting
the national-target pool with what are actually 50 state rows.

Sibling state labels use `state/` prefixes (e.g. line 849
`state/irs/aca_enrollment/{state}`, state AGI, state SNAP). Fix
the ACA spending label to match.

This was flagged during independent review of PR #778 (state
Medicaid label prefix) as an analogous but higher-severity bug
(real calibration-weighting impact vs naming-consistency for #778).

Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: MaxGhenis

changelog.d/fix-state-aca-spending-label-prefix.fixed.md

@@ -0,0 +1 @@
+Re-prefix state ACA spending calibration label with state/ (was nation/) so reweight() correctly classifies it as a state target.

policyengine_us_data/utils/loss.py

@@ -818,8 +818,10 @@ def build_loss_matrix(dataset: type, time_period):
             "aca_ptc", map_to="household", period=time_period
         ).values
 
-        # Add a loss-matrix entry and matching target
-        label = f"nation/irs/aca_spending/{row['state'].lower()}"
+        # Add a loss-matrix entry and matching target. Prefix `state/`
+        # so `reweight()` correctly classifies this as a state-level
+        # (non-national) target via `startswith("nation/")`.
+        label = f"state/irs/aca_spending/{row['state'].lower()}"
         loss_matrix[label] = aca_value * in_state
         annual_target = row["spending"]
         if any(loss_matrix[label].isna()):