PolicyEngine Macro
UNEMPLOYMENT RATE
5.0%
percent
- Observation
- 2026Q1
- Vintage
- 2026-07-26
- Source
- ONS · MGSX
UK topic · labour market
What is happening to work, pay, and hiring?
Unemployment is 5.0% in 2026Q1 and vacancies 712.0 thousand in 2026-05; the satellite forecast runs 4 quarters and then stops.
01 — where it stands
The numbers, as the snapshot stores them.
UK VACANCIES
712.0
thousands, seasonally adjusted three-month average
- Observation
- 2026-05
- Vintage
- 2026-07-26
- Source
- ONS · AP2Y
AVERAGE WEEKLY EARNINGS
749.0
£ per week, seasonally adjusted
- Observation
- 2026-05
- Vintage
- 2026-07-26
- Source
- ONS · KAB9
Vacancies are a seasonally adjusted three-month average in thousands and earnings are £ per week in cash terms — neither is deflated here, and neither is annualised.
How those readings have moved, derived from the same stored observations rather than typed in: Unemployment rate -0.2pp against the prior quarter (2025Q4); UK vacancies -18.0 thousand against a year earlier (2025-05); Average weekly earnings +3.5% against a year earlier (2025-05).
02 — what the models see
The model view, and the same breath its limits.
The labour market is covered by a satellite, not a model of its own: the svar-unemployment satellite maps the boe-svar GDP forecast onto the ONS unemployment rate through a fitted Okun relation. Round 2026-07-28 reads GDP from archived round forecasts/rounds/2026-07-21/boe-svar.json (data edge 2026Q1) and unemployment from ONS MGSX through 2026Q1 (u=5.0, du=-0.2), vintage 2026-07-26, and puts 2026Q2 at 5.0%, 68% 4.9%–5.0%.
The round file's own description of the mapping, verbatim: Okun's-law satellite: deterministic mapping of the boe-svar UK GDP growth forecast to the ONS unemployment rate (MGSX). Fit du_t = alpha + beta*g_t + gamma*du_{t-1} by OLS on 1992Q1-2025Q1 with 2020Q1-2021Q2 dummied out (furlough decoupled unemployment from GDP): alpha=0.0611, beta=-0.0474, gamma=0.1871, sigma=0.172pp, n=123, R2=0.368.
The round states its own limits, and they are strict:
- Horizon capped at 4 quarters: the rolling-origin validation (boe-var-model results/okun_validation.json) shows the end-to-end SVAR->Okun forecast beats a random walk at horizons 1-4 ex-COVID (RMSE ratios 0.82-0.99) and is clearly worse beyond 4 quarters.
- The improvement over no-change is not statistically significant at 73 overlapping origins (Diebold-Mariano p >= 0.14).
- Bands carry GDP-forecast uncertainty only, mapped through the deterministic Okun equation; Okun residual uncertainty (sigma 0.17pp per quarter) is NOT sampled, so the stated bands understate total uncertainty.
The satellite has no entry in the capability registry that backs the other model pages, and no CLI command or MCP tool of its own: its limits are the ones the archived round file states, quoted above. See the archived rounds →
Vacancies and average weekly earnings above have no model view at all. They are outturns, shown because they move before the unemployment rate does — not because anything here forecasts them.
03 — run it yourself
Every number above is a command away.
The satellite is a deterministic mapping applied to the GDP forecast, so the runnable half is the forecast itself; the mapping's coefficients are in the archived round file, not behind a command.
pe-macro forecast --horizons 4 # the GDP path the satellite maps
pe-macro model-status boe-svar # supported uses, access, and stated limitationsOver MCP: forecast_uk for the GDP path and get_model_status for the limitations as structured data. The archived round itself is a committed JSON file under forecasts/rounds/, readable with no client at all. Connect a client →
04 — the data behind it
Source, coverage, and the immutable file.
| Series | Publisher | Coverage | Snapshots | This vintage | Next release |
|---|---|---|---|---|---|
| Unemployment rate | ONS · MGSX | 1971Q1 – 2026Q1 | 2 | 2026-07-26.json | 18 August 2026 |
| UK vacancies | ONS · AP2Y | 2001-05 – 2026-05 | 1 | 2026-07-26.json | 18 August 2026 |
| Average weekly earnings | ONS · KAB9 | 2000-01 – 2026-05 | 1 | 2026-07-26.json | 18 August 2026 |
The labour market series are survey based and revised; the unemployment rate is the only one of the three that the satellite forecast touches.
The snapshot files are append-only and never edited, so a number published here can be reproduced against the data as it stood — browse the store, its release calendar, and the as-of recipe →