thesisPrototype

Government data

UK PAYE Payrolled Employees, June 2026

What will the ONS/HMRC first-print early estimate of UK seasonally adjusted PAYE payrolled employees be for June 2026, in millions?

Overdue · was due Jul 21, 2026

The resolver has no route for this target's reference. The target was registered with a generic source link rather than a resolver binding.

NO_ROUTE_GENERIC_URL · resolver status as of Oct 7, 2026

Forecast

30.3M
30.2M–30.3M
80% prediction interval
gpt-5.5Run record ↗

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%30.1M30.3M30.4MForecast value

Shaded band: 80% interval (30.2M–30.3M). Dashed line: point estimate (30.3M).

Derived from the point estimate and 80% interval; the agent did not report a full distribution.

Analysis

§UK PAYE Payrolled Employees June 2026 First Print

Framing and exact resolver: forecast the ONS/HMRC PAYE RTI seasonally adjusted UK payrolled employees early estimate for June 2026, in millions, first print only. The ledger source URL is the ONS June 2026 bulletin even though that page is already the May 2026 data release; the official next-release date for the dataset is 21 July 2026, so I keep the registered target but flag the URL-period discrepancy.

▸ Reported tool use: official.lookupmodel report
Opened ONS June 2026 PAYE RTI bulletin and release header.
↳ Fetched release date 18 June 2026, next release 21 July 2026, May 2026 payrolled employees 30.3 million, May month change +2,000, May annual change -119,000.
▸ Reported tool use: official.lookupmodel report
Opened ONS June 2026 PAYE RTI payrolled-employees section and main points.
↳ Fetched April 2026 growth revised from -100,000 to -53,000; May 2026 was +2,000 month on month; May 2026 annual percentage change was -0.4%.
▸ Reported tool use: official.lookupmodel report
Opened ONS May 2026 PAYE RTI bulletin for the prior flash and recent reference points.
↳ Fetched April 2026 early estimate 30.2 million, April flash month change -100,000, April annual change -210,000, and March 2026 revised month change -28,000.
▸ Reported tool use: official.lookupmodel report
Opened ONS seasonally adjusted dataset and revision triangle pages to verify schedule and source variant.
↳ Fetched seasonally adjusted dataset release date 18 June 2026, next release 21 July 2026, data coverage July 2014 to May 2026, and revision triangle next release 21 July 2026.

Variant consistency: all anchors are the same ONS/HMRC PAYE RTI payrolled employees, seasonally adjusted, UK series used in the bulletin figure/table, not non-seasonally adjusted or regional/sector breakdowns.

Reference class and base rate: recent month-to-month first/recent print changes around the target are March -28,000, April -53,000 after revision, and May +2,000. The base rate is a soft labour-market drift of roughly -25,000 to -30,000 per month, but May's flat print and April's upward revision argue against extrapolating the April flash shock one-for-one.

Prior/update/interval: the chosen time-series prior is a local persistence-plus-momentum model rather than a broader historical model, because the target is a first-print PAYE flash in a current soft labour-market regime with early-tax-year revision noise. Persistence prior is May 2026 level 30.300 million; historical sample for successive changes is intentionally local at [-0.028, -0.053, 0.002] million from recent ONS prints; adjustment components are level 30.300 plus momentum -0.026 million, softened by May stabilization and early-tax-year upward revision risk to -0.020 million, giving point 30.280 million. For interval, sigma = 0.0275 million from those successive changes; 1.28*sigma = 0.035 million. I widen to 0.055 million, about 1.6x, because ONS states early tax-year flash estimates have greater uncertainty and only about 85% of information is available initially, versus 98% to 99% next month. Final implied bounds: 30.280 - 0.055 = 30.225 and 30.280 + 0.055 = 30.335 million.

Upside risk: a continued payrolling rebound after April revisions, especially if administrative and support services strength persists, would land above the interval with a June move more than about +55,000 versus the adjusted May base. Downside risk: renewed job cuts in accommodation, retail, or small employers after wage and employer-cost pressure would land below the interval with a June move worse than about -55,000 versus the adjusted May base. Outside the interval would require a June move materially larger than roughly +/-55,000 from the adjusted May base.

Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

Key drivers

  • May level was broadly flat at 30.3 million
  • April weakness was revised from -100,000 to -53,000
  • Annual growth remained negative at -119,000
  • Early tax-year PAYE flashes have elevated revision and imputation uncertainty
  • Vacancy and sector signals point to a soft but not collapsing labour market

Tool evidence

This run has no captured tool responses. “Reported tool use” in the analysis is the model’s account, not an archived tool response.

Sources and resolution

Official source
ONS/HMRC Earnings and employment from Pay As You Earn Real Time Information, UK
Resolution date
July 21, 2026· outcome not recorded
Resolution rule
Resolve to the first official ONS/HMRC print of seasonally adjusted UK PAYE payrolled employees for June 2026, in millions, using the headline payrolled employees series in the Earnings and employment from Pay As You Earn Real Time Information bulletin/table. Use the first published estimate only, with no later revisions. The JSON resolutionDate of 2026-07-21 is the expected first-print publication date from the official next-release schedule. The registered ledger URL points to the June 2026 bulletin, which ONS has already used for May 2026 data; this discrepancy is preserved for audit rather than silently changing the target.
Run details

The analysis is the model’s written report. Tool-use descriptions in that report are model claims; the activity artifacts contain the execution record.

thesis.analyst · gpt-5.5 · v2.2.0

pre-submit review · completed

Draft is publishable with minor warnings: it handles the ledger/source discrepancy transparently and keeps units, point, interval, and first-print rule coherent.

  • warning model_prior: The model prior is effectively a persistence-plus-three-month momentum rule, but it does not explicitly say whether a broader time-series model was considered or ruled out.
  • warning interval: The interval uses only three recent monthly changes for sigma, which is a very small realized-volatility sample even though it is widened for flash uncertainty.
  • info optional_suggestion: State explicitly that the JSON `resolutionDate` of 2026-07-21 is the expected first-print publication date despite the retained ledger URL discrepancy.
  • info optional_suggestion: The upside/downside tails are useful; tying them to approximate monthly changes in thousands would make them easier to audit.

disposition accepted: Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

disposition accepted: Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

disposition not applicable: Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

disposition not applicable: Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

Activity artifacts

Complete original trace

§UK PAYE Payrolled Employees June 2026 First Print

Framing and exact resolver: forecast the ONS/HMRC PAYE RTI seasonally adjusted UK payrolled employees early estimate for June 2026, in millions, first print only. The ledger source URL is the ONS June 2026 bulletin even though that page is already the May 2026 data release; the official next-release date for the dataset is 21 July 2026, so I keep the registered target but flag the URL-period discrepancy.

▸ Reported tool use: official.lookupmodel report
Opened ONS June 2026 PAYE RTI bulletin and release header.
↳ Fetched release date 18 June 2026, next release 21 July 2026, May 2026 payrolled employees 30.3 million, May month change +2,000, May annual change -119,000.
▸ Reported tool use: official.lookupmodel report
Opened ONS June 2026 PAYE RTI payrolled-employees section and main points.
↳ Fetched April 2026 growth revised from -100,000 to -53,000; May 2026 was +2,000 month on month; May 2026 annual percentage change was -0.4%.
▸ Reported tool use: official.lookupmodel report
Opened ONS May 2026 PAYE RTI bulletin for the prior flash and recent reference points.
↳ Fetched April 2026 early estimate 30.2 million, April flash month change -100,000, April annual change -210,000, and March 2026 revised month change -28,000.
▸ Reported tool use: official.lookupmodel report
Opened ONS seasonally adjusted dataset and revision triangle pages to verify schedule and source variant.
↳ Fetched seasonally adjusted dataset release date 18 June 2026, next release 21 July 2026, data coverage July 2014 to May 2026, and revision triangle next release 21 July 2026.

Variant consistency: all anchors are the same ONS/HMRC PAYE RTI payrolled employees, seasonally adjusted, UK series used in the bulletin figure/table, not non-seasonally adjusted or regional/sector breakdowns.

Reference class and base rate: recent month-to-month first/recent print changes around the target are March -28,000, April -53,000 after revision, and May +2,000. The base rate is a soft labour-market drift of roughly -25,000 to -30,000 per month, but May's flat print and April's upward revision argue against extrapolating the April flash shock one-for-one.

Prior/update/interval: the chosen time-series prior is a local persistence-plus-momentum model rather than a broader historical model, because the target is a first-print PAYE flash in a current soft labour-market regime with early-tax-year revision noise. Persistence prior is May 2026 level 30.300 million; historical sample for successive changes is intentionally local at [-0.028, -0.053, 0.002] million from recent ONS prints; adjustment components are level 30.300 plus momentum -0.026 million, softened by May stabilization and early-tax-year upward revision risk to -0.020 million, giving point 30.280 million. For interval, sigma = 0.0275 million from those successive changes; 1.28*sigma = 0.035 million. I widen to 0.055 million, about 1.6x, because ONS states early tax-year flash estimates have greater uncertainty and only about 85% of information is available initially, versus 98% to 99% next month. Final implied bounds: 30.280 - 0.055 = 30.225 and 30.280 + 0.055 = 30.335 million.

Upside risk: a continued payrolling rebound after April revisions, especially if administrative and support services strength persists, would land above the interval with a June move more than about +55,000 versus the adjusted May base. Downside risk: renewed job cuts in accommodation, retail, or small employers after wage and employer-cost pressure would land below the interval with a June move worse than about -55,000 versus the adjusted May base. Outside the interval would require a June move materially larger than roughly +/-55,000 from the adjusted May base.

Review disposition: accepted the reviewer warnings by explicitly labeling the persistence-plus-momentum calculation as the chosen local time-series prior, justifying the three-observation volatility sample as regime-local, clarifying that 2026-07-21 is the expected first-print publication date, and expressing the tail scenarios in approximate thousands.

calibrated forecast · 80% CI
30.3M[30.2M · 30.3M]
Target metadata

Data point: ons.hmrc.paye_payrolled_employees.june_2026.first_print

More government data forecasts