Full Time Employment Chg

Full Time Employment Change is the monthly change in Australia's seasonally adjusted count of full-time employed people (figures with regular seasonal patterns removed). It sits inside Labour Force Australia (catalogue number 6202.0), the Australian Bureau of Statistics (ABS) monthly jobs report from the Monthly Labour Force Survey (LFS).

Full-time means usually working 35 hours or more weekly across all jobs, with reference-week hours (hours in the survey week) deciding borderline cases. It covers resident civilians aged 15 years and over. The survey counts people by main job, not jobs, so a second job never changes the headcount.

Interviews generally run during the 2 weeks starting on the Sunday between the 5th and the 11th, covering the prior reference week (survey week). Release usually comes 39 days after interviews start, usually the third Thursday after the reference month, except December at 46 days, at 1130am Canberra time. Each release publishes seasonally adjusted figures (seasonality removed) and trend estimates (smoothed direction) for employment, unemployment, underemployment, participation rate (labour force share) and hours worked. A Detailed file adds state detail, while concurrent seasonal adjustment (monthly factor refresh), quarterly rebenchmarking (population updates) and Census rebasing (Census realignment) revise past values.

The publisher names no weighted components because the series is a single headcount with analytical splits only.

• Full-time employed people: headcount of employed people meeting the 35 hours rule across all jobs.
• Male and female full-time levels: separate headcounts showing how each sex contributes to change.
• State and territory levels: full-time headcounts for each region that sum toward national movement.
• Derived ratios: full-time share of employment and employment to population ratio giving context only.
• Survey mechanics: rotation overlap, sampling variability, reference-week effects and benchmark updates shifting monthly moves.

The December 2023 example published January 18 2024 shows seasonally adjusted full-time employment down by 106600 to 9791200 while part-time rose by 41400 to 4409900. Total employment fell to 14201100 with unemployment steady at 3.9% and participation at 66.8%. Trend estimates showed full-time down only 1800 to 9857100 and part-time up 20600 to 4388900, with total employment up 18800 to 14246000. The lesson is that a sharp adjusted fall offset by part-time gains can leave trend almost flat, the steadier guide to direction.

1. Classify by hours: use usual hours across all jobs against the 35 hours cutoff, with survey-week hours deciding borderline cases. Count people working 1 hour or more as employed.
2. Composite estimation: combine current month with previous 6 months of overlapping rotation responses (repeated dwelling groups) to estimate the level, exploiting respondent correlation. Use 8 rotation groups with 7 of 8 parts overlapping.
3. Weight and calibrate: apply sampling fractions as initial weights, then align 7 months of data to population benchmarks (official population totals) projected forward 3 quarters. Example: New South Wales 1 in 445 means each respondent initially represents 445 people.
4. Seasonal adjustment: decompose each original series as O = T * S * I where O is original, T trend-cycle, S seasonal and I irregular. Use concurrent factors refreshed monthly with ARIMA forecasts discarded after use, removing supplementary-survey effects and adjusting January starts and Easter.
5. Trend: apply a 13-term Henderson weighted moving average (smoothing formula) to the seasonally adjusted series, using surrogate weights for the latest 6 months.
6. Change: subtract the prior month full-time level from the reference month level using unrounded values, published in persons and quoted by calendars in thousands. Formula: Chg(t) = FT_SA(t) - FT_SA(t-1) where FT_SA is seasonally adjusted full-time.
7. Illustrative example: 9801200 persons in month A and 9791800 in month B gives Chg = 9791800 - 9801200 = -9400 persons, published as -9.4K.