Glenn unpacks the most recent population projections from the ABS, explaining how projections are different to forecasts, and why there is a role for both in planning.

Last week, the ABS released their latest population projections for Australia and for each state and territory.

These were last done in 2012, and are normally produced after each Census, to give an idea of where the nation is headed. This set goes out to 2066, a 50-year timeframe from the previous Census.

The heading to this blog post and all the ABS data projections have one thing in common: they produce projections in several “series”. 3 of them to be precise, based on a different set of assumptions; A, B and C – a high, medium and low series.

Some of the key takeaways

• Australia’s population should hit 30 million (currently about 25 million by 2028/29 based on the high series, and 2032/33 on the low series).
• The middle series, series B, shows the population hitting 30 million in 2030/31, and 40 million in 2058/59.
• Series B still has quite a high migration assumption, with a continuation of the current rate of migration about 225,000 per year (net) throughout the time period.
• New South Wales would hit 10 million people around 2036, while Victoria is expected to continue to be the fastest growing state and hit 9 million by 2040. Victoria’s projected growth is between 1.0% and 1.7% based on the highest and the lowest assumptions. Notably, even the highest assumptions have a growth rate for Victoria below that experienced over the last few years, of 2.3-2.4%.
• Western Australia’s growth is highly variable between the series – the current growth rate of 0.7% p.a. is equivalent to the lowest series, while the high one has it at 1.7% – still well below the 3.3% experienced a few years ago at the peak of the mining construction boom.
• Series C, the lowest growth series projects a population decline in Tasmania to 453,000 by 2066. But the highest growth series, Series A projects 744,000 people in the island state, almost 50% more than today.
• Based on the high growth Series A, Melbourne would equal Sydney’s population in 2029 and overtake it after that to become Australia’s largest city, based on current Greater Capital City boundaries. If the current rate of growth continues, this will occur sooner than that.

What’s the difference between a forecast and a projection?

The ABS is at pains to point out that these projections are not in any way a forecast or a prediction. They are what might happen to Australia’s population based on an arbitrary set of parameters around migration, birth and death rates.

Migration is the main driver as this accounts for around 2/3rds of Australia’s population growth at the moment. And it’s the most susceptible to change.

So a projection is based on extending and modifying past growth rates into the future. And the ABS projections have 3 series, plus a number of other supplementary series based on varying assumptions. So it’s likely that at least one of them will come close to reality. The difficulty is if you’re planning for future populations, which would you choose?

A forecast is quite different. .id’s future population calculations are forecasts. We produce forecasts of the most likely future population, age structure and household type, with a single series over the next 25 years.

Our forecasts are still based on assumptions, but the assumptions take quite a different form. And we don’t always get it exactly right either, but we rely on regularly monitoring and reviewing the underlying assumptions that we use to model our forecasts to keep them relevant.

• Our forecasts are locally based, and look at the key drivers of change at a micro-level – what is happening in each suburb/town or district.
• Our forecasts are workshopped with Local Government so they represent the outcome of both current trends and the council’s housing policy.
• They contain a specific set of assumptions for each area on the rate of residential development, with each of the larger developments separately apportioned.
• They contain a migration by age rate which is based on the amenity and attractiveness of the area to different age groups and the future dwelling construction.
• They are reviewed and updated regularly rather than having multiple scenarios. So at any point in time, you know which figures to use, based on the most likely outcome of the current trends. And when new information comes to light, they are reviewed.

So the ABS population projections are interesting as a “What if?” scenario. But they don’t give you the detailed evidence to make decisions about your future population, certainly not at a local level.

I sometimes liken our population forecast to a weather forecast.

A “Weather projection” would say: It was 25 degrees and sunny today. The last time it was 25 degrees and sunny on a day, it was 27 degrees and sunny the next day, and 21 degrees and raining the following day, so that becomes the prediction.

A “Weather forecast” which is what we’re more familiar with looks at the underlying drivers of weather – eg. cold fronts and low-pressure systems and tracks these to provide a local forecast, which is usually provided a week out, but then tracked and updated every day. This is more like our population forecasts.

But we do look at the overall projections for the country. They can certainly give you an insight into larger areas and also longer timeframes than our forecasts can produce. And of course, they are a great talking point about the future of population in this country.