DEFINE-UK ecological stock-flow
DEFINE-UK
Scores how a UK climate-policy mix moves output, emissions, unemployment and consumption, as deltas against the current-policies baseline.
Good for
- Climate-policy scenarios from the papers’ published switches
- Output, emissions, unemployment, consumption as deltas from the S1 baseline
- Scenario income deltas for the incidence bridge (experimental)
A worked example
Green public investment (green_public_investment), minus the current-policies baseline, from the cached pinned run (upstream commit 846081a).
Extra public spending on green power, financed with green bonds
About £15–16bn a year in 2027–31, tapering to 2035; about £91bn net over the run.
Effects on public debt and interest costs are not among the served outputs, so the fiscal side of the green-bond financing is not shown here.
- By 2040 (full run) the GDP gain fades to +0.25% and the emissions cut to 23 MtCO2e a year
- The gain is back-loaded: the cumulative multiplier is 2.25 (nominal, 2027 policy start to 2040), below published estimates at short horizons (by horizon)
- In 2035, 46 of the 54 Mt come from outside the power sector (2030: 23 of 49): electricity becomes much cheaper (−52% in 2030), which moves other sectors off fossil fuels; this channel is read from the outputs, not traced equation by equation (limits)
Read with care. Every delta includes a parameter mismatch between the baseline and policy runs, about −0.16% of GDP (limits). No published v1.1 result exists to check these against.
Several checks do not pass (limits):
- Baseline emissions run below the manual’s table and above DESNZ actuals
- Baseline growth is far above outturn
- Two scenarios are refused as not stock-flow consistent
- Subsidy scenarios lower GDP, including cumulative real GDP over 2027–40, so far unexplained (detail)
How far to trust it
Three checks, weakest included; each opens its evidence.
- Baseline macro blockManual v1.1 Table 4, 2025/2030/2040inside every gate (the gates were set after the run); two by only 0.01pppartial
- Baseline vs outturnsONS 2025 growth; levels never served4.66% vs 1.31%fail
- Clean-room reimplementationUpstream run as oracle; milestone 2 not passedGDP up to 44% below upstream by 2040fail
The macro block matches Table 4 within every gate (the 2025 growth gate was set at that miss); 2025 growth (−0.30pp, gate ±0.31) and 2040 unemployment (−0.19pp, gate ±0.2) clear by only 0.01pp.
Deltas only, never levels. No published v1.1 scenario result exists to check a delta against (the ceiling).
Run it
The worked example above, three ways, after a one-off local R run. Defaults return 2027–2040, from the policy start; 2023–26, the parameter mismatch alone, is left out.
Using DEFINE-UK, run the green public investment scenario and show real GDP, emissions and unemployment as deltas from baseline, year by year.DEFINE-UK is local-only: the hosted tools (define_scenario) return run instructions, never results. Install the adapter, then run it locally or through a local MCP server. Connect
pip install git+https://github.com/PolicyEngine/define-uk-model
pe-macro define-scenario green_public_investmentfrom define_uk import scenarios
r = scenarios.run_scenario("green_public_investment")
print(r["years"], r["variables"]["GDP_R"]["delta_pct"])With the microsimulation
Connected experimentally: deltas in, incidence out.
define_scenario_incidence carries only the numbers, never the upstream code (every join in the suite).
Model cardinputs, outputs, horizon, data, access
- Answers
- How a UK climate-policy mix moves output, emissions, unemployment and consumption against the current-policies baseline
- Inputs
- A curated scenario name from the upstream scenario set
- Outputs
- Scenario minus baseline: real GDP, emissions, unemployment, real consumption
- Horizon
- Annual scenario deltas, 2027–2040 (from the policy start); the validation checks read the full run from 2023
- Data
- DEFINE-UK 1.1 upstream at pinned commit
846081a(April 2026 manual) - Uncertainty
- None quantified: deterministic deltas from one calibration; every delta also carries a Baseline/policy parameter mismatch (limits)
- Access
- Local CLI and Python package; hosted tools return run instructions, never results (Code)
- Runtime
- Instant from the cached upstream run; a fresh run needs local R
- Joins the microsimulation
- Household income deltas become a UK earnings overlay (the join above, experimental)
- Status
- Partial replication of the DEFINE-UK manual: baseline macro block matches Table 4 inside every gate, two by only 0.01pp (the 2025 growth gate was set at that miss); scenario deltas gated on the pinned oracle run, the published scenario definitions, and two paper anchors (no numeric v1.1 scenario results are published); experimental; unlicensed upstream is never hosted, so hosted calls return run instructions
- Cannot answer
- Reform scoring (no statute mapping); forecasts or baseline levels (deltas only)