The jump-off year is the last year with estimated data from the ABS (population, births, deaths, and migration). The jump-off year marks the line between estimates and projections. Figures up to and including the jump-off year are estimates. Figures after it are projections. The jump-off year population is the base for all future figures. Every projected birth, death, and migration event starts from this base. So how precise the data is directly affects all projected numbers.
The cohort-component method uses the jump-off year population as its starting base. Any error in the jump-off year carries forward for that cohort. For example, if the estimate of women aged 25 is too high, then projected births will be too high.
The jump-off year moves forward when new population data comes out. After each Census, the ABS rebases the data. It goes back and fixes all figures since the last Census. It lines them up with the new Census count. This can change the jump-off year numbers. The change can be large for small areas, where fixes make up a bigger share.
When the jump-off year moves forward, the previous projected data for that year is replaced with estimated data. This improves the accuracy of short-term projections. All projections should be re-run from the new base. Place Forecast does this through its workflow. The system finds the jump-off year from the data at hand. It works out rates from the estimation period. Then it projects forward. This makes sure the projections always use the latest data.
Place Forecast shows the jump-off year with a red dot on charts. This dot marks where estimated data ends and modelled data begins. Everything to (and including) the left of the red dot comes from ABS population estimates. Everything to the right is projected by Place Forecast.
The red dot helps users see which numbers are from real-world data and which are modelled. When the jump-off year moves forward — by pulling in new ABS data — the modelled period gets shorter. This makes near-term figures more sound.
Each dataset has its own jump-off year based on the latest data it holds. When a figure is worked out from two or more datasets, its jump-off year is the earliest of the input jump-off years. For example, average household size depends on both enumerated population and dwellings. If one has data up to 2021 and the other up to 2022, the jump-off year for average household size is 2021.