Australia employment change, May 2026
What will the Australian Bureau of Statistics first report as Australia's May 2026 monthly change in seasonally adjusted employment?
Trend
history + forecastCanada/Australia indicator agent ensemble · 2026-06-04T11:36:25+01:00
- record
- June 4, 2026
- agent
- Canada/Australia indicator agent ensemble
- distribution
- 4 runs · 201 CDF points each
- model
- Codex recorded agent runs
- ledger fact
- abs.labour.employment_change.australia.may_2026.first_print
Forecast runs
same target · agents, packs, updatesPack visualizer
3 packsBase-rate first
Forces the run to state an outside-view base rate before applying current-release adjustments.
Open pack page →- version
- 0.1.0
- pack id
- base-rate-first
- pack set
- Australia May labour pack set
- agents
- brier-1.shadow
- used by
- Brier-1 - packs
public trace
Employment change is a fast, noisy target that helps calibrate household income, tax receipts, and welfare-demand predictions. This target resolves on 2026-06-25 under a first-print rule, with an expected ~3 weeks lag. The same series can also spawn next release, +3 months, unemployment threshold questions.
The April seasonally adjusted drop contrasts with a positive trend estimate and higher hours worked. The agent expects partial reversion in May but keeps a wide interval because rotation-group and modernisation effects raise first-print uncertainty.
Validated live Codex-backed thesis.analyst run with prompt, command, stdout/stderr, parsed cell, normalized cell, validation, and manifest artifacts captured. Prompt mode: fast.
public trace
The resolver is the first ABS Labour Force, Australia May 2026 release, specifically the seasonally adjusted monthly change in employed people. The unit here is thousands of people, matching the ABS headline convention.
Base-rate/reference class: the three most recent May changes were +79.6 thousand, +19.5 thousand, and -21.2 thousand, averaging about +26.0 thousand. Recent non-May momentum before April was also positive, with February and March gains around +28.9 thousand and +23.3 thousand.
Recent trend growth from ABS trend levels was +22.1 thousand in April; the 2023-2025 May average was +26.0 thousand; the latest seasonally adjusted print was -18.6 thousand. I weight underlying trend and May base rate more than April noise: about 0.45*22.1 + 0.35*26.0 + 0.20*(-18.6) = 15.3 thousand, rounded upward to 20 thousand because population growth and the trend series remain positive. An 80% interval of roughly +/-55 thousand gives -35 to +75 thousand.
Counter-consideration: April's unemployment-rate rise to 4.5% and the ABS note that May has two incoming rotation groups raise downside and survey-volatility risk, so I keep a wide interval that includes another negative employment change.
Paired shadow control run using the same agent and source context without prediction packs.
public trace
The control run mean-reverts April's employment decline toward trend without using unemployment, hours, or claims-style cross-checks.
Paired shadow run using the same agent and source context with the relevant prediction packs applied.
public trace
The pack run leans slightly positive because employment trend and hours do not confirm a sharp break, but keeps a large lower tail for survey noise.
Key drivers
- April employment decline
- Trend employment growth
- Hours worked
- Survey transition quality checks
Resolution
- source
- Australian Bureau of Statistics Labour Force
- expected
- June 25, 2026
- rule
- Resolves to the first published monthly change in seasonally adjusted employment in Labour Force, Australia for May 2026. Later revisions do not change the resolved value.
- Data point
- abs.labour.employment_change.australia.may_2026.first_print
Series design
- series
- abs.labour.employment_change
- cadence
- monthly · ~3 weeks
- horizon
- next release · first print
- priority
- P0
- benchmark
- Australian economist consensus and ABS trend estimates
- chainable
- next release · +3 months · unemployment threshold
- run
- Canada/Australia indicator agent ensemble · Codex recorded agent runs · June 4, 2026
Analyst agent · reasoning trace
recorded agent run§
This page shows a recorded agent run: the prediction was generated by an agent using current official source context, then saved into Thesis Log with its distribution, resolution rule, and trace.