Below is another excellent post from MB reader Erin Rolandsen, CEO of Angelassist and author of beyondtheragemachine.substack.com:
High levels of short-term immigration have made official population measures unfit for purpose.
Australia’s population statistics were designed for a low-churn world, where most migrants arrived permanently and most residents stayed put. That is no longer the economy we are running.
Today, Australia deliberately operates a high population churn model, with millions of temporary residents cycling through the country on student, working holiday and short-term skilled visas.
Despite this, we have kept our old conceptual machinery.
The population figure produced by the Census is not a headcount of people physically present in Australia. It measures the usual resident population, defined as people who have been present in Australia for at least 12 of the last 16 months.
Between censuses, the ABS updates this figure by adding births, subtracting deaths, and adjusting for Net Overseas Migration, again using the usual residence definition. The result is the Estimated Resident Population: a measure designed to track long-term residence, not real-time physical presence.
The usual residence definition made sense when migration was overwhelmingly permanent. It makes little sense in an economy increasingly dependent on short-duration, high-turnover migration.
There are currently more than 2.9 million people in Australia on temporary visas. Many of these are on visas of one year or less. If they leave one day before 12 months, they do not fulfil our usual residence definition.
This means a significant proportion of working holiday, student, PALM scheme, training and skilled visas are not counted within our Net Overseas Migration numbers.
The result is that hundreds of thousands of people who require housing, transport and services immediately are statistically invisible at the point those pressures arise.
After accounting for visa duration and residency definitions, there are over one million people who are physically present but who are not captured in the population figures most often used for housing and infrastructure planning.
This is not a statistical quirk. It is a systemic planning failure.
Housing supply responds to physical presence, not narrow statistical definitions.
One person staying 365 days in one house for one year is directly equivalent to 365 people staying one night each in that same house for one year.
Consider for a second what this means when on any given night Australia has over 300,000 international tourists staying in private rentals.
It is the cumulative impact of the sustained constant volume that matters. The usual residence definition is irrelevant.
Stop reading this article right now. Go search the major Australian government websites, and ask yourself this one simple question:
How many people are in Australia, right now?
If your source mentions the Census count, has it added back in every short-term visitor? If it mentions the Estimated Resident Population, has it made allowance for every visitor staying less than 12 out of the last 16 months? If they aren’t using the term “population” to describe our total number of current residents, what term are they using?
With all our sophistication as a society, with all the complicated statistical tools at our disposal, we have somehow forgotten how to do something as simple as a headcount.
The horrifying reality is that the housing shortages, infrastructure bottlenecks, transport overload, hospital congestion, and service strains Australia is experiencing are not simply due to a failure to forecast demand. They are the result of measurement problems that have failed to consider the impact of those who are already here.
We are arguing about future supply responses while flying blind on existing demand.
This failure to accommodate over a million people actually living in our country is one of the most consequential technocratic failures in recent Australian history.
The problem is not that the data does not exist. Arrivals, departures, and visa stocks are all recorded.
The problem is that Australia lacks a single official, transparent measure of how many people are actually in the country at any given time. We rely on population labels that imply a headcount when none exists and migration labels that systematically exclude a significant proportion of our migrants.
At a minimum, the statistical series should be relabelled to reflect what they actually measure. “Estimated Resident Population” should be renamed Estimated Usual Resident Population. “Net Overseas Migration” should be described as Net Overseas Longer-Term Migration.
More fundamentally, Australia needs a commonly cited indicator that approximates a physical headcount, so that housing and infrastructure planning reflects lived reality rather than statistical abstraction.
In a high-churn economy, planning without knowing how many people are here is no longer just imprecise.
It is disastrous.
