Same question, every model

Six questions put to every macro model, each answer next to a sourced benchmark where one exists. A model that cannot answer says why.

Pick a question

Models are in the hub order, not ranked. The sourced benchmark sits in the first row; each model row gives its peak effects and a verdict, and Details holds the run, full units and caveats.

How to read the verdicts

Each verdict starts with one of six words: plausible, off, questionable, unresolved, not tested or no instrument; some add a qualifier (such as “off, sign right”), and each carries its reason and the open issue. These words grade one experiment; each model’s overall standing uses the status vocabulary on its model card. Results file version 2026-11-19.1; how it is checked and regenerated.

Where and when each row was run

The battery was first put on 6 Nov 2026; each row’s Details give its exact call and commit.

  • OBR emulator: rate, spending and tax rows re-run locally on 19 Nov 2026 at the current pin 0e44ec8 (October 2026 EFO); every quarter converges. The rate-row peaks are unchanged from 0fe04c9; the spending multiplier is now 1.01–1.02 (was 1.01–1.03) and the q12 tax ratio 0.37 (was 0.38). The energy row is an earlier committed output.
  • FRB/US: the +100bp, spending, tax and energy rows re-run on 18 Nov 2026 at the current pin 162cfe6 (October 2026 LONGBASE), with shocks from 2026Q3, the first quarter after LONGBASE history; identical to the 846eeb5c runs of 16 Nov, which the −100bp and +25bp rows still are.
  • US HANK: the +100bp and spending rows re-run on 18 Nov 2026 at the current pin 58353e5, now storing the policy-rate path; output and inflation are identical to the hosted runs of 6 Nov, which the −100bp and +25bp rows still are.
  • BoE SVAR and DEFINE-UK: committed model-repo outputs, with no solve; the SVAR +100bp medians come from its fast test fixture (see that row’s Details).

What does a 100bp rise in the policy rate do?

A 1pp rise in the policy rate. The models do not share a rate path, so each row states its own and the size of the path in pp-quarters (annualised rate × quarters).

Four of five models answer. Off: OBR emulator and BoE SVAR; plausible: FRB/US; plausible in-model: US HANK. DEFINE-UK has no instrument for it.

What does a 100bp rise in the policy rate do?: every model's answer, with the sourced benchmark first
ModelPeak GDP% of GDPPeak inflationpp, annual rateTimingGDP per pp-quarter% of GDPVerdictDetails
BoE benchmark UK−0.2 to −0.35−0.2 to −0.4After about a year0.05–0.09%Reference, sourced
Benchmark
Bank of England, 1pp rise held one year (UK). Output −0.2% to −0.35% after about a year; inflation −0.2 to −0.4pp a year or so after that. 1pp × 4 quarters = 4 pp-quarters, so 0.05–0.09% of GDP per pp-quarter.
Source
Bank of England, The transmission mechanism of monetary policy, Quarterly Bulletin 1999 Q2
Fed benchmark US−0.486 (output gap)−0.0345 (core PCE)Trough at q8not publishedReference, sourced
Benchmark
Fed FRB/US, 100bp funds-rate shock (VAR expectations), current model (US). The Board's own simulation of the current (post-2018) model: output gap trough −0.486pp, core PCE inflation trough −0.0345pp and unemployment peak +0.219pp, all at quarter 8. The previous model gives −0.459pp, −0.062pp and +0.215pp. The −0.4pp and −0.08pp in the 2014 FEDS Note describe the pre-2018 model.
Source
Laforte, Overview of the changes to the FRB/US model (2018), FEDS Notes, 7 Dec 2018, doi:10.17016/2380-7172.2306, Fig. 2 (accessible data table)
OBR emulator UK−0.058at q40no responseGDP peak q40.014%OffGDP 3–6× below the BoE range; prices flat
Run
local run at obr-macroeconomic-model 0e44ec8 (October 2026 EFO), 2026-11-19obr_shock(var="R", shock=1, periods=4)
Rate path
+1pp held 4 quarters, then zero; 4 pp-quarters
Timing
GDP peaks at q4; no price response
Peak GDP
−0.058, % of GDP, peak, at q4
Peak inflation
0, pp, annual rate, CPI does not respond
GDP per pp-quarter
0.014% of GDP per pp-quarter
Verdict
Off. GDP is 3–6× below the BoE 1999 range (−0.2% to −0.35%), through consumption only; inflation and unemployment are exactly zero.
Solver
converged in all 12 quarters (nominal-GDP pin)
Open issue
obr-macroeconomic-model #50
BoE SVAR UK−1.6h = 8−2.5CPI level, h = 20Builds, no reversion0.2%Off, sign right4.6–8× the BoE peak; 2–4× per pp-quarter
Run
committed output sourcecommitted fig3 fixture + PR #17: shock="1 s.d. UK monetary-policy shock (+25bp on impact), scaled ×4"
Rate path
+25bp on impact, still +0.17pp at h = 12 (a unit root: the rate does not revert); about 1.9 pp-quarters to h = 8 per 25bp
Timing
GDP −0.14% per s.d. on impact, building to −0.39% by h = 8; no reversion within 5 years
Peak GDP
−1.6, % of GDP, at h = 8, scaled from −0.39% per 25bp, at h = 8
Peak inflation
−2.5, % on the CPI level at h = 20, scaled from −0.63% per 25bp, CPI level, still falling
GDP per pp-quarter
0.2% of GDP per pp-quarter
Verdict
Off, sign right. At h = 8 GDP is −1.6% (scaled to 100bp) against the BoE's −0.2% to −0.35% after about a year: 4.6–8× the peak, or 2–4× per pp-quarter once the SVAR's longer rate path is allowed for. It builds to h = 8 and does not revert within 5 years (the rate path has a unit root), against the BoE's fullest effect after about a year.
Draws
The medians are from the repository's fast test fixture (600 posterior draws, about 50 accepted, unweighted; tests/conftest.py), not the published replication run (10,000 draws, 765 accepted, effective sample size 368.0; results/summary.md at c827ab8). An earlier sense check found bands wider than the paper's (186 accepted draws, effective sample size 91, against the paper's 10,000; results/sense_check.md).
Open issue
boe-var-model #11
FRB/US US−0.552at q8 (2028Q2)−0.034at q8Hump, peak q80.15%PlausibleCore PCE matches; unemployment 1.04× on the Board's design
Run
local run at frbus 162cfe6 (October 2026 LONGBASE), 2026-11-18, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="rffintay_aerr", shock=1.0, periods=1, horizon=12, start="2026Q3", policy_rule="inertial_taylor")
Rate path
+1.00pp on impact, decays under the inertial rule, crosses zero at q8, undershoots; 3.6 pp-quarters, q1–q8
Timing
Hump: GDP and inflation peak at q8
Peak GDP
−0.552, % of GDP, peak, at q8 (2028Q2)
Peak inflation
−0.034, pp, annual rate, core PCE, at q8
GDP per pp-quarter
0.15% of GDP per pp-quarter
Verdict
Plausible; within 10% of the Board on its own design. The Board's simulation of the current model gives an output-gap trough of −0.486pp, core PCE −0.0345pp and an unemployment peak of +0.219pp at quarter 8 (Laforte 2018). This run gives core PCE −0.034pp at quarter 8 and an output-gap trough of −0.50pp, and an unemployment peak of +0.28pp, 1.28× the Board's. That ratio depends on the start date: on the Board's own 2020Q1 design the model gives +0.228pp, 1.04×, with core PCE 0.96× and the output gap 0.97× (us-frb-model results/inflation_transmission.md at 162cfe6). The −0.08pp in earlier Fed notes describes the pre-2018 model.
US HANK US−0.361at impact−1.34at impactPeak on impact0.16%Plausible in-modelInflation peaks on impact, not after two years
Run
local run at us-hank-model 58353e5, 2026-11-18hank_shock(kind="monetary", size=0.0025, persistence=0.9, horizon=12, variant="two_asset")
Rate path
Taylor-rule intercept shock (rstar) +0.25pp a quarter, AR(1) ρ = 0.9: not a policy-rate rise of that size. The nominal policy rate the rule sets falls 0.25pp a quarter on impact (about 1.0pp annualised), because inflation falls by more than the intercept rises (i = rstar + 1.5π), and is still 0.11pp below steady state at q11. Realised real rate +1.36pp annualised on impact; 2.3 pp-quarters, real rate, 12 quarters
Timing
Both peak on impact; no hump
Peak GDP
−0.361, % of GDP, peak, at impact
Peak inflation
−1.34, pp, annual rate (−0.336pp a quarter × 4), at impact
GDP per pp-quarter
0.16% of GDP per pp-quarter
Verdict
Plausible in-model, off on timing; an intercept shock, not a 100bp rate rise. Inflation is about 40× FRB/US and 3–7× the BoE range, and peaks on impact rather than after about two years, which is consistent with the model's forward-looking Phillips curve. The shock is to the rule's intercept: the nominal policy rate falls while the real rate rises.
Open issue
macro #211
DEFINE-UK UKNo instrumentDEFINE-UK has no monetary-policy instrument.

GDP per pp-quarter divides the peak GDP effect by the rate path's size to the peak. HANK's figure uses the realised real-rate path; the other rows use the nominal policy rate, which is close to the real rate in FRB/US because its inflation barely moves. On HANK's rstar path (10 pp-quarters at ρ = 0.9) the figure would be about 0.04%, so HANK's 0.16 and FRB/US's 0.15 are not like-for-like.

HANK reports inflation as a quarterly rate; this page multiplies it by 4 to give an annual rate.

How the response unfolds, quarter by quarter. The paths behind the peaks and timing in the table; each panel has its own scale.

Source: each row’s stored path (FRB/US and US HANK local runs of 18 Nov 2026 at 162cfe6 and 58353e5) · results.json

FRB/US

GDP, % deviation

FRB/US: GDP, 100bp rate rise GDP, % deviation, quarters 1 to 12: 0.001, −0.184, −0.286, −0.433, −0.48, −0.527, −0.544, −0.552, −0.546, −0.528, −0.501, −0.467. Shaded: Benchmark (US): pp output gap, trough, current and previous model, −0.486 to −0.459.
  • Benchmark (US): pp output gap, trough, current and previous model

Inflation, pp, annual rate (picxfe, core PCE)

FRB/US: inflation, 100bp rate rise Inflation, pp, annual rate (picxfe, core PCE), quarters 1 to 12: 0, −0.01, −0.021, −0.025, −0.031, −0.033, −0.034, −0.034, −0.034, −0.033, −0.032, −0.032.

US HANK

GDP, % deviation

US HANK: GDP, 100bp rate rise GDP, % deviation, quarters 1 to 12: −0.361, −0.273, −0.216, −0.175, −0.144, −0.121, −0.102, −0.087, −0.074, −0.063, −0.054, −0.046.

Inflation, pp, annual rate (quarterly rate × 4)

US HANK: inflation, 100bp rate rise Inflation, pp, annual rate (quarterly rate × 4), quarters 1 to 12: −1.342, −1.282, −1.201, −1.11, −1.018, −0.928, −0.843, −0.764, −0.691, −0.624, −0.563, −0.508.

No stored path for OBR emulator, BoE SVAR: those rows are peaks from the battery or committed outputs.

Is a 100bp cut the mirror image of a 100bp rise?

The same experiment with the sign flipped. A linear model gives exactly the opposite answer; the gap is the asymmetry.

Four of five models answer. Plausible in-model: OBR emulator; not tested (linear): BoE SVAR and US HANK; plausible: FRB/US. DEFINE-UK has no instrument for it.

Is a 100bp cut the mirror image of a 100bp rise?: every model's answer, with the sourced benchmark first
ModelAgainst the risepeak GDP, % of GDP: cut vs riseVerdictDetails
BenchmarkNone sourced for this question.
OBR emulator UK+0.059 vs −0.058Plausible in-modelClose to symmetric; size off, see +100bp row
Run
local run at obr-macroeconomic-model 0e44ec8 (October 2026 EFO), 2026-11-19obr_shock(var="R", shock=-1, periods=4)
Against the rise
Peak +0.059 against −0.058 for the rise (2.8% apart); at q12 +0.0060 against −0.0055.
Peak GDP
+0.059, % of GDP, peak, at q4
Verdict
Plausible in-model; size off. Close to symmetric. The size is judged in the +100bp row, where it is off.
Solver
converged in all 12 quarters (nominal-GDP pin)
BoE SVAR UKExact mirrorNot testedLinear VAR: exact mirror by construction
Run
committed output sourcePR #17 identity test: shock="−1 s.d. against +1 s.d."
Against the rise
Exact mirror (tested in PR #17).
Verdict
Not tested (linear). A linear VAR: the cut is the exact negative of the rise. Its size is off; see the +100bp question.
Open issue
boe-var-model #11
FRB/US US+0.568 vs −0.552 (2.8% apart)PlausibleClose to symmetric, as a nonlinear model allows
Run
local run at frbus 846eeb5c (October 2026 LONGBASE), 2026-11-16, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="rffintay_aerr", shock=-1.0, periods=1, horizon=20, start="2026Q3", policy_rule="inertial_taylor")
Against the rise
+0.568 against −0.552 for the rise: 2.8% apart.
Peak GDP
+0.568, % of GDP, peak, at q8
Verdict
Plausible. Close to symmetric; FRB/US is nonlinear, so a small gap is expected.
US HANK USExact mirrorNot testedLinear by construction
Run
hosted run, 2026-11-06hank_shock(kind="monetary", size=-0.0025, persistence=0.9)
Against the rise
Exact mirror: responses are first-order around the steady state.
Verdict
Not tested (linear). Linear by construction, so this tests nothing beyond the +100bp run.
DEFINE-UK UKNo instrumentDEFINE-UK has no monetary-policy instrument.

Is a 25bp rise a quarter of a 100bp rise?

A 25bp rise, compared with the 100bp result divided by four. A linear model gives the same GDP effect per 100bp at both sizes.

Four of five models answer. Plausible in-model: OBR emulator; not tested (linear): BoE SVAR and US HANK; plausible: FRB/US. DEFINE-UK has no instrument for it.

Is a 25bp rise a quarter of a 100bp rise?: every model's answer, with the sourced benchmark first
ModelAgainst 100bp ÷ 4peak GDP per bpVerdictDetails
BenchmarkNone sourced for this question.
OBR emulator UK0.9% apartPlausible in-modelClose to proportional; size off, see +100bp row
Run
local run at obr-macroeconomic-model 0e44ec8 (October 2026 EFO), 2026-11-19obr_shock(var="R", shock=0.25, periods=4)
Against 100bp ÷ 4
Peak GDP per 100bp: 0.058 at 25bp, 0.0575 at 100bp, 0.057 at 200bp. 25bp × 4 is 0.9% from the 100bp peak.
Verdict
Plausible in-model; size off. Close to proportional across 25bp to 200bp. The size is judged in the +100bp row, where it is off.
Solver
converged in all 12 quarters (nominal-GDP pin)
BoE SVAR UKExactNot testedLinear VAR: exact by construction
Run
committed output sourcePR #17 identity test: shock="0.25 s.d. against 1 s.d."
Against 100bp ÷ 4
Exact (tested in PR #17).
Verdict
Not tested (linear). A linear VAR. Its size is off; see the +100bp question.
Open issue
boe-var-model #11
FRB/US US1.1% apartPlausibleClose to proportional
Run
local run at frbus 846eeb5c (October 2026 LONGBASE), 2026-11-16, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="rffintay_aerr", shock=0.25, periods=1, horizon=20, start="2026Q3", policy_rule="inertial_taylor")
Against 100bp ÷ 4
25bp × 4 = −0.558% against −0.552% at 100bp: 1.1% apart.
Verdict
Plausible. Close to proportional.
US HANK USExactNot testedLinear by construction
Run
hosted run, 2026-11-06hank_shock(kind="monetary", size=0.000625, persistence=0.9)
Against 100bp ÷ 4
Exact: responses scale with the shock size.
Verdict
Not tested (linear). Linear by construction, so this tests nothing beyond the +100bp run.
DEFINE-UK UKNo instrumentDEFINE-UK has no monetary-policy instrument.

What does government spending of 1% of GDP do to output?

Higher government purchases worth about 1% of GDP. The multiplier is the GDP change divided by the spending change, both in % of GDP. Financing and the monetary rule differ by model and drive most of the spread.

Four of five models answer. Off: OBR emulator; plausible: FRB/US; questionable: US HANK; unresolved: DEFINE-UK. BoE SVAR has no instrument for it.

What does government spending of 1% of GDP do to output?: every model's answer, with the sourced benchmark first
ModelMultiplierGDP per unit spentPeak GDP% of GDPPeak inflationpp, annual rateVerdictDetails
Published range US0.45–1.82 (CBO); 0.6–1 (Ramey); ≈0.95 (Coenen, no accommodation)——Reference, sourced
Benchmark
Purchases multiplier, year 1 (US). CBO 0.45–1.82 (Fed responding); Ramey 2019 0.6–1; Coenen et al. 2012 FRB-US, no accommodation, ≈0.95. OBR uses 0.6 for current spending on impact, tapering to zero over five years (OBR, Nov 2023).
Source
Whalen and Reichling, CBO Working Paper 2015-02 (converted); Ramey, Journal of Economic Perspectives 2019; Coenen et al., AEJ: Macroeconomics 2012; OBR, Dynamic scoring of policy measures in OBR forecasts, Nov 2023
OBR emulator UK1.01–1.02+1.014at q1–q40no responseOffAn accounting identity; the OBR's own figure is 0.6
Run
local run at obr-macroeconomic-model 0e44ec8 (October 2026 EFO), 2026-11-19obr_shock(var="CGG", shock=7000, periods=4)
Experiment
CGG +£7,000m a quarter (about 1.0% of GDP) for 4 quarters
Multiplier
1.01 on impact, 1.02 by q4. In the model, borrowing rises by the cash cost (obr #75); obr_shock does not return it
Peak GDP
+1.014, % of GDP, peak (+1.007% to +1.014% over q1–q4, +0.01% at q5, fading), at q1–q4
Peak inflation
0, pp, annual rate, prices do not respond
Verdict
Off. 1.01→1.02 comes from an identity, not estimated behaviour. It falls inside CBO's range but sits above Ramey's upper bound of 1 and the OBR's own 0.6.
Solver
converged in all 12 quarters (nominal-GDP pin); the tool's own caveat calls the result an upper bound
See also
OBR emulator limits
BoE SVAR UKNo instrumentThe BoE SVAR identifies no fiscal shock.
FRB/US US0.81 on impact; 0.75 year 1+0.037per 0.046% of GDP+0.0008at q2Plausible0.81 impact, 0.747 yr 1: inside both ranges
Run
local run at frbus 162cfe6 (October 2026 LONGBASE), 2026-11-18, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="egfe_aerr", shock=0.01, periods=1, horizon=12, start="2026Q3", policy_rule="inertial_taylor")
Experiment
Federal purchases +1% (about 0.046% of GDP), one quarter
Multiplier
about 0.81 on impact; 0.747 in year 1 for a committed sustained run from 2026Q1 (inertial rule; us-frb-model results/benchmark_multipliers.md)
Peak GDP
+0.037, % of GDP, peak, per 0.046% of GDP spent, at impact
Peak inflation
+0.0008, pp, annual rate, core PCE, at q2
Verdict
Inside CBO and Ramey ranges. Impact 0.81 and year-1 0.747 sit inside CBO's 0.45–1.82 and Ramey's 0.6–1. Both are below Coenen's ≈0.95, which is consistent: Coenen assumes no accommodation, while this run uses an inertial Taylor rule that responds.
Open issue
macro #212
US HANK US0.175 (ρ = 0.8)+0.175at impact+0.84at impactQuestionableSign flips while spending is still high
Run
local run at us-hank-model 58353e5, 2026-11-18hank_shock(kind="fiscal_spending", size=0.01, persistence=0.8, horizon=12, variant="two_asset")
Experiment
G +1% of GDP on impact, AR(1) ρ = 0.8, financed by the labour tax
Multiplier
0.175 (ρ = 0.8); 0.075 at ρ = 0.9. Output turns negative from q3 while spending is still 51–73% of impact
Peak GDP
+0.175, % of GDP, peak, at impact
Peak inflation
+0.84, pp, annual rate (+0.21pp a quarter × 4), at impact
Verdict
Questionable. Impact 0.175 at ρ 0.8 is below the CBO floor of 0.45 and Ramey's floor of 0.6. Labour-tax financing differs from the benchmark setups, so it cannot be ruled off.
Open issue
us-hank-model #5
DEFINE-UK UK2.25 nominal, 2027 to 2040not reportednot reportedNo benchmarkAnnual, investment-led; not comparable
Run
committed output sourcecommitted green public investment scenario: scenario="green public investment", frequency="annual"
Experiment
Green public investment, annual, investment-led
Multiplier
2.25 nominal per £ of the power-sector spending the policy sets, from the 2027 policy start to 2040; back-loaded: 0.47 after one year, 0.79 to 2031, 1.40 to 2034. Over the full run from 2023: 2.30 nominal, and 1.78 per £ of all non-wage spending. Parameter mismatch included: 2.25 if its effect stops in 2027, as low as 2.08 if it persists (1.73–1.75 for the secondary)
Verdict
Unresolved. The headline impulse is now the policy instrument, government power-sector spending (define-uk-model #16, decided 15 Nov 2026). An annual, investment-led scenario cumulated to 2040 is still not comparable to a 4-quarter purchases shock, and the published comparators (the OBR's impact multiplier of about 1.0; Batini et al. 2021, IMF WP/21/87: 1.1–1.5 for renewable investment) are short-horizon, where this model sits below them (0.47 after one year). Both figures include a Baseline/policy parameter mismatch (define-uk-model #20).
See also
define-uk-model #16 · DEFINE-UK limits

FRB/US's hosted tool cannot run a sustained +1%-of-GDP experiment: the stored run is a one-quarter impulse of about 0.046% of GDP, and the year-1 figure comes from a committed sustained run (macro #212).

How the response unfolds, quarter by quarter. The paths behind the peaks and timing in the table; each panel has its own scale.

Source: each row’s stored path (FRB/US and US HANK local runs of 18 Nov 2026 at 162cfe6 and 58353e5) · results.json

FRB/US

GDP, % deviation

FRB/US: GDP, spending shock GDP, % deviation, quarters 1 to 12: 0.037, 0.025, 0.025, 0.019, 0.018, 0.015, 0.012, 0.009, 0.007, 0.004, 0.002, 0.

Inflation, pp, annual rate (picxfe, core PCE)

FRB/US: inflation, spending shock Inflation, pp, annual rate (picxfe, core PCE), quarters 1 to 12: 0, 0.001, 0.001, 0.001, 0, 0, 0.001, 0.001, 0.001, 0.001, 0.001, 0.001.

US HANK

GDP, % deviation

US HANK: GDP, spending shock GDP, % deviation, quarters 1 to 12: 0.175, 0.069, 0.011, −0.018, −0.032, −0.036, −0.036, −0.033, −0.029, −0.025, −0.022, −0.019.

Inflation, pp, annual rate (quarterly rate × 4)

US HANK: inflation, spending shock Inflation, pp, annual rate (quarterly rate × 4), quarters 1 to 12: 0.835, 0.79, 0.709, 0.616, 0.524, 0.439, 0.365, 0.301, 0.247, 0.203, 0.166, 0.136.

No stored path for OBR emulator, DEFINE-UK: those rows are peaks from the battery or committed outputs.

What does a 1pp rise in the personal tax rate do?

A rise in the effective personal income-tax rate of about 1pp.

Two of five models answer. Off: OBR emulator; unresolved: FRB/US. BoE SVAR, US HANK and DEFINE-UK have no instrument for it.

What does a 1pp rise in the personal tax rate do?: every model's answer, with the sourced benchmark first
ModelMultiplierGDP per unit of taxPeak GDP% of GDPPeak inflationpp, annual rateVerdictDetails
Published range US and UK0.3–1.5 (CBO, 2-yr lower/middle); 0.1–0.6 (CBO, 1-yr higher income); OBR 0.3, decaying——Reference, sourced
Benchmark
Tax multiplier (US and UK). CBO, Fed not tightening: two-year lower/middle-income cut 0.3–1.5; one-year higher-income cut 0.1–0.6; no central value. OBR household tax 0.3 on impact, decaying.
Source
Whalen and Reichling, CBO Working Paper 2015-02; OBR, Dynamic scoring of policy measures in OBR forecasts, Nov 2023
OBR emulator UK0.21 at q4 → 0.37 q12, current pricesnot reportednot reportedOffRises where the OBR's 0.3 decays
Run
local run at obr-macroeconomic-model 0e44ec8 (October 2026 EFO), 2026-11-19obr_shock(var="HHDI_ADDFACTOR", shock=1637.5, periods=12)
Multiplier
0.21 at q4, rising to 0.37 by q12, in current prices, the basis of the OBR's 0.3 and of the Overview (0.19 and 0.33 in real GDP)
Verdict
Off: wrong profile. Right order of magnitude, but it rises where the OBR's 0.3 decays.
See also
OBR emulator limits
BoE SVAR UKNo instrumentThe BoE SVAR identifies no tax shock.
FRB/US US0.23 year 1; 0.32 year 2−0.115at impact−0.003at q4UnresolvedYr 1 0.23 unmatched; yr 2 0.32 between rows
Run
local run at frbus 162cfe6 (October 2026 LONGBASE), 2026-11-18, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="trp_aerr", shock=0.01, periods=1, horizon=12, start="2026Q3", policy_rule="inertial_taylor")
Experiment
+1pp on the effective personal rate on impact, decaying to +0.05pp by q8
Multiplier
0.23 in year 1, 0.32 in year 2 (committed tax-cut run from 2026Q1; us-frb-model results/benchmark_multipliers.md)
Peak GDP
−0.115, % of GDP, peak, at impact
Peak inflation
−0.003, pp, annual rate, core PCE, at q4
Verdict
No matched CBO cell. CBO publishes no central value and no cell matching a 1pp rise in the effective rate. Year 1 (0.23) has no matched cell. Year 2 (0.32) falls between the two CBO rows, whose regimes differ. The earlier 'about 0.3, band 0.15–0.45' was not sourced.
Open issue
macro #212
US HANK USNo instrumentThe labour tax is endogenous in US HANK, so there is no tax-rate instrument.
DEFINE-UK UKNo instrumentNo income-tax instrument in the committed scenarios.

FRB/US with periods = 4 stacks the tax add-factor to +2.37pp rather than +1pp, and the tool's label does not warn (macro #212). The stored run uses periods = 1.

How the response unfolds, quarter by quarter. The paths behind the peaks and timing in the table; each panel has its own scale.

Source: each row’s stored path (FRB/US and US HANK local runs of 18 Nov 2026 at 162cfe6 and 58353e5) · results.json

FRB/US

GDP, % deviation

FRB/US: GDP, income-tax rise GDP, % deviation, quarters 1 to 12: −0.115, −0.087, −0.104, −0.079, −0.078, −0.067, −0.057, −0.046, −0.034, −0.023, −0.012, −0.002.

Inflation, pp, annual rate (picxfe, core PCE)

FRB/US: inflation, income-tax rise Inflation, pp, annual rate (picxfe, core PCE), quarters 1 to 12: 0, −0.003, −0.002, −0.003, −0.002, −0.002, −0.002, −0.002, −0.002, −0.002, −0.002, −0.002.

No stored path for OBR emulator: those rows are peaks from the battery or committed outputs.

What does a rise in oil and energy prices do?

A rise in the oil price of about $10 a barrel. No sourced magnitude benchmark exists for any model, so only the sign and the transmission are judged.

Three of five models answer. Off: OBR emulator; plausible: BoE SVAR and FRB/US. US HANK and DEFINE-UK have no instrument for it.

What does a rise in oil and energy prices do?: every model's answer, with the sourced benchmark first
ModelPeak GDP% of GDPPeak price effect% on the CPI levelPolicy rateVerdictDetails
BenchmarkNone sourced for this question.
OBR emulator UK−0.020no responsenot reportedOffThe price block does not respond to oil
Run
committed output sourcetransmission audit: shock="oil +$10/bbl"
Peak GDP
−0.02, % of GDP
Peak price
0, % on the CPI level, CPI does not respond
Verdict
Off. The price block does not respond to the oil price.
See also
OBR emulator limits
BoE SVAR UKnot reportednot reportedBank Rate: no significant responsePlausible on textBank Rate response matches the paper; size not benchmarked
Run
committed output sourcecommitted world energy shock: shock="world energy shock"
Experiment
World energy shock
What else moves
Bank Rate does not significantly respond: the 68% set straddles zero at h = 0–8, as the paper's text says.
Verdict
Plausible on the paper's text. Magnitudes are not benchmarked.
Open issue
boe-var-model #11
FRB/US US−0.036at q4+0.093CPI level, impactFunds rate eases 0.014ppPlausible on signNo sourced size benchmark; core PCE dip undiagnosed
Run
local run at frbus 162cfe6 (October 2026 LONGBASE), 2026-11-18, shocks from 2026Q3 (the first quarter after LONGBASE history)frbus_shock(var="poil_aerr", shock=10, periods=4, horizon=12, start="2026Q3", policy_rule="inertial_taylor")
Experiment
Oil +$10 a barrel (+9.9% at the 2026Q3 LONGBASE price) for 4 quarters
What else moves
The funds rate eases by 0.014pp, because the rule targets core inflation; everything reverts within a quarter of the shock ending, with no second-round effects.
Peak GDP
−0.036, % of GDP, peak, at q4
Peak price
+0.093, % on the CPI level, on impact, CPI level; core PCE −0.002pp, at impact
Verdict
Plausible on sign. No sourced benchmark for the size. Core PCE falls 0.002pp after the oil rise; the cause is not yet diagnosed.
Open issue
macro #212
US HANK USNo instrumentUS HANK has no energy-price instrument.
DEFINE-UK UKNo instrumentNo energy-price shock in the committed scenarios.

FRB/US's tool labels the oil lever's units 'model units'; they are dollars a barrel, so +10 is about +15% at the tool's default 2026Q1 start and +9.9% at this run's 2026Q3 start (macro #212).

How the response unfolds, quarter by quarter. The paths behind the peaks and timing in the table; each panel has its own scale.

Source: each row’s stored path (FRB/US and US HANK local runs of 18 Nov 2026 at 162cfe6 and 58353e5) · results.json

FRB/US

GDP, % deviation

FRB/US: GDP, oil-price shock GDP, % deviation, quarters 1 to 12: −0.016, −0.026, −0.033, −0.036, −0.015, −0.009, −0.005, −0.006, −0.012, −0.01, −0.009, −0.006.

CPI level, % deviation

FRB/US: prices, oil-price shock CPI level, % deviation, quarters 1 to 12: 0.093, 0.097, 0.103, 0.11, −0.001, −0.001, −0.001, −0.001, −0.001, −0.001, −0.001, −0.001.

No stored path for OBR emulator, BoE SVAR: those rows are peaks from the battery or committed outputs.