PolicyEngine Macro

Open economic models · public evidence

Open UK and US economic models, scored in public.

Run the models yourself. Every result carries its assumptions, official data vintage, and validation — and the UK forecasts are scored in public.

Run a hosted model Score a reform
boe-svar: UK real GDP growth forecast, median with 68% and 90% credible bands Fan chart. Posterior forecast of UK year-on-year real GDP growth from the boe-svar model, 2026Q2 to 2029Q2, conditioned through 2026Q1. The median path runs from +1.14% to +1.41%. The 90% credible band is widest at 2029Q2, spanning -0.67% to +3.77%; its lower edge is positive for the first 3 quarters and dips below zero from 2027Q1 onward. -1% +0% +1% +2% +3% 2026 2027 2028 2029
Current UK GDP growth forecast. Explore the model →
00 — start with the work

One path from economic question to checkable result.

Score a reformUK

Score a tax or benefit change

The microsimulation computes the direct cost and who pays; the OBR emulator returns the economy-wide feedback — one reform, scored end to end.

score a reform →
Macro scenariosUKUS

Test a policy or shock

Choose among seven open household and macro models, then inspect each model’s assumptions, evidence, and limits.

choose a macro model →
Current economyUKUS

See what changed

Follow activity, prices, labour, rates, upcoming releases, and the official-data vintages behind every figure.

see the economy →
Forecast recordUK

Check forecasts against what happened

Archive forecast rounds before results arrive, preserve official-data vintages, and score every realised period. A US record follows once a US forecasting model is hosted.

inspect the record →

Explore all seven models or compare their assumptions side by side.

01 — at a glance

Latest outturns beside the model's near-term view.

indicatorlatest outturnmodel near-term forecast
Real GDP growth y/y 0.9% 2026Q1 → 1.1% 2026Q268% range 0.6%–1.7%
CPI inflation y/y 2.8% 2026Q2 → 2.5% 2026Q368% range 1.8%–3.3%
Unemployment rate 5.0% 2026Q1 → 5.0% 2026Q268% range 4.9%–5.0%

ONS outturns (as of 2026-07-26) beside archived forecast rounds. Full horizon →

02 — validation

What the evidence shows—and what it doesn't.

Each model is checked against the strongest available benchmark. Replication, forecast performance, statutory correctness, and calibration are different kinds of evidence—and where no independent check exists, we say so. Compare the evidence →

obr-macroAnchored replication

Tracks the March 2026 EFO to 0.15% GDP MAPE with published anchors; free-running error is 5.75%.

boe-svarForecast evaluation

Beats a random-walk baseline on CPI across 49 rolling origins; measured band coverage runs below nominal.

frb-usSoftware replication

Matches the Fed’s pyfrbus to within the Fed’s own two releases’ disagreement (~1×10⁻⁸); no predictive claim.

us-hankPublished replication

Hits every Auclert et al. (2021) calibration target; responses are first-order and stylized.

pe-microsimStatutory checks

Rules are tested against statute; population totals inherit survey uncertainty.

psl-ogCalibration only

Targets are met by construction; no independent outcome benchmark exists.

define-ukPartial replication

Baseline macro block replicates the manual; scenario deltas are design-gated and paper-anchored — no published scenario numbers exist. Deltas only, never levels.

03 — start using it

Run hosted models—or use the code directly.

Connect the public MCP server with no PolicyEngine account or API key, use the shared CLI, or call each Python package directly. Connect and try a model →

Claude ChatGPT