Extend current $2,000 CTC beyond 2025 sunset. Without extension, reverts to $1,000.
Ten-year deficit effect, conventional+$600.0Badds to the deficit
With macroeconomic feedback and debt service+$676.9BFRB/US-Lite (Federal Reserve calibrated)
Checked against
Published score +$600.0B; this estimate is 0.0% away (calibrated reference). Calibrated to reproduce $600B (CBO 2024) — agreement is by construction, not an independent test.
No out-of-sample band: no pre-registered row scores a refundable credit. The 44 pre-registered Tier 1 rows are the only tier that measures this model's accuracy, and none of them scores a policy of this kind.
Approximate bracket method: this policy cannot run through the return-level microsimulation, so read the shape as directional rather than as a validated table. Groups are ranked by adjusted gross income per return. Average change is in dollars per return; positive means the group pays more tax.
Economic effects
difference from the CBO baseline
GDP, summed over the window+0.79percent-years
Jobs at the largest effect+107K
Deficit effect of the GDP change-$55.1Bmore tax collected
Deficit effect of interest on the debt+$132.0Bmore debt, more interest
Fiscal year
GDP level
Jobs
Unemployment
10-year rate
GDP change, deficit
Interest, deficit
2026
+0.149%
+95K
-0.07 pt
+0.01 pt
-$10.4B
+$2.4B
2027
+0.168%
+107K
-0.08 pt
+0.01 pt
-$11.7B
+$4.8B
2028
+0.143%
+91K
-0.07 pt
+0.02 pt
-$10.0B
+$7.2B
2029
+0.109%
+69K
-0.05 pt
+0.03 pt
-$7.6B
+$9.6B
2030
+0.078%
+50K
-0.04 pt
+0.03 pt
-$5.5B
+$12.0B
2031
+0.054%
+35K
-0.03 pt
+0.04 pt
-$3.8B
+$14.4B
2032
+0.037%
+24K
-0.02 pt
+0.04 pt
-$2.6B
+$16.8B
2033
+0.025%
+16K
-0.01 pt
+0.05 pt
-$1.7B
+$19.2B
2034
+0.016%
+10K
-0.01 pt
+0.06 pt
-$1.1B
+$21.6B
2035
+0.011%
+7K
-0.01 pt
+0.06 pt
-$0.7B
+$24.0B
FRB/US-Lite (Federal Reserve calibrated). Demand effects fade over five to seven years as monetary policy responds; there is no supply-side channel. The dynamic score is the conventional score, plus the deficit effects of the GDP change and of interest on the debt: +$600.0B - $55.1B + $132.0B = +$676.9B.
Other scoring models
exploratory
Model
Ten-year deficit effect
Method
CBO-Style
+$600.0B
Static + ETI behavioral response + optional FRB/US dynamic feedback
TPC-Microsim Pilot
Cannot score this
CTC Permanent Extension (TaxCreditPolicy) does not map onto current microsim pilot reforms (income-tax rates, CTC, EITC, SALT, std deduction, AMT exemption).
No pilot model returned distributional output; the comparison is aggregate-only.
CBO-Style is the model behind the headline above. The TPC-style microsimulation is a pilot: where the two disagree, the gap tells you how much the answer depends on method. It is not an official range.
Limitations
holdout labels follow the locked post-2026-05-02 regression protocol; they are not retroactive historical out-of-sample claims.
No pre-registered out-of-sample row scores a policy of this class, so this model's accuracy on one has not been measured.
Credit eligibility and refundability are modeled with synthetic tax units rather than CPS ASEC microdata.
Interactions with SALT, AMT, and filing-status heterogeneity remain approximated in the current household tax module.
TAUTOLOGICAL FITTED ROW (owner Decision 5). The module's annual for this benchmark is -$160.0B, the carried $1,600B target divided by exactly ten, so the by-construction error tests x/10 x 10 == x and nothing else. Read the leave-one-out row instead: run_loo.py derives this case per unit from the CPS microdata and never touches the annual. The row is kept rather than deleted so the tautology stays visible - deleting a benchmark to improve a tier mean is the failure mode pre-registration exists to forbid.
The $1,600B is a rounded secondhand figure attributed to 'CBO/JCT 2021' with no transcribed table row behind it, so the residual on the derived side is measured against a target of unknown precision.
The current credit module extrapolates from bracket-level aggregates rather than return-level household data.
TAUTOLOGICAL FITTED ROW (owner Decision 5). The module's annual for this benchmark is -$60.0B, the carried $600B target divided by exactly ten, so the by-construction error tests x/10 x 10 == x and nothing else. Read the leave-one-out row instead: run_loo.py derives this case per unit from the CPS microdata and never touches the annual. The row is kept rather than deleted so the tautology stays visible - deleting a benchmark to improve a tier mean is the failure mode pre-registration exists to forbid.
Examined and deliberately not moved in Wave 4 (target_revisions.EXAMINED_NOT_REVISED). Two published figures score a child-credit extension and neither replaces $600B. JCT's JCX-35-25 scores P.L. 119-21's child credit at +$816.8B over FY2025-2034, but that is a $2,200 indexed credit against this benchmark's $2,000 flat one and it is already carried here as the `pl119_21_child_tax_credit` benchmark, so adopting it would score one JCT row as two benchmarks. CRS R48286 Table 1, transcribing CBO, prints $735.3B for 'Increase and Modification of Child and Dependent Credit', which CRS itself says includes the credit for other dependents that this module does not score. Both sit *above* the module's design rather than bracketing it, so a range would assert a containment neither publisher supports.
NO LONGER A TAUTOLOGICAL FITTED ROW. The module's annual for this benchmark is -$17.8B, which is the *superseded* $178B target divided by exactly ten; the target is now Treasury's published $162.6B, so the reported error is a real 9.5% rather than the by-construction 0.0% it used to be, and no constant was retuned to close it. The leave-one-out row is still the better read: run_loo.py derives this case per unit from the CPS microdata and never touches the annual.
The derived side under-predicts this row for two reasons in the survey file rather than the rules: CPS ASEC carries no self-employment earnings, which the EITC counts, and the tax-unit builder folds 19-to-23-year-olds with a parent pointer into the parent's unit - which is most of the population an age expansion from 25 to 19 is about.