Why Housing is (Primarily) a Supply Problem

Australia's Population Growth is Extraordinary – and Hard to Justify

The Australian housing debate has a favourite villain: immigration. Too many people, not enough homes. It's an intuitively appealing argument, and the population data is genuinely striking. But it's the wrong diagnosis – and wrong diagnoses lead to wrong policy.

Let's start with what the data actually shows. By any international comparison, Australia's population growth rate is remarkable. Since 2000, Australia's population has grown to an index of 142 against a base of 100 – well above the OECD mean of 117 and median of 115. Since 2022 alone, Australia has grown at around 2.2% per year, second only to Saudi Arabia and Canada among OECD-monitored nations, and roughly double the OECD average.

The majority of that growth is immigration-driven, especially in recent years.

I'll be direct: I can't construct a compelling economic justification for running population growth at that pace. That volumetric debate deserves to be had honestly. 

But here's the thing: that population growth is not new. It has fluctuated within a broad one to two and a half per cent band for sixty years, sitting at the upper end since the mid-2000s and spiking after 2022, but the upper end is not unprecedented, the late 1980s ran nearly as high. It has been a dominant feature of Australia's demographic landscape for decades. And that changes the argument entirely.

The many drivers of new dwelling prices

Before focusing on a primary driver, it is worth slowing down, because the debate's besetting habit is to seize one cause and run. How many homes get built, and the price they sell at, is set by a long list of forces, not one. My rough working list follows:

Demand side (sets the price ceiling)

• Population growth, including immigration
• Household formation / average household size
• Capacity to pay: incomes, interest rates, credit and macroprudential settings
• Location value: amenity, transport accessibility to the city core and natural features
• Foreign investment and capital flows
• Tax/subsidy settings that lift buyer demand (negative gearing, CGT, first-home incentives)
  

 

Cost side (construction costs that set the price floor)

• Land cost
• Construction input costs and supply chains (materials and wages)
• Financing cost (interest rates again)
• Required profit margin to ensure ongoing business viability
• Regulatory/building-code compliance costs (energy, fire, accessibility standards)
• Zoning caps on yield (density, height limits)
• Holding/transaction taxes (land tax, stamp duty)

 

Friction and capacity (time and money costs: how fast and reliably builds happen)

• Planning and approval processes, and rezoning delay and risk
• Infrastructure and servicing (water, sewerage, power, roads, drainage, gutters, pavements)
• Land availability and rate of release
• Construction labour and trades supply
• Construction sector capacity and productivity

 

Market structure and behaviour (who builds and how they play it)

• Industry concentration / developer and land market power (market competition)
• Land banking and strategic release
• Builder solvency and contract risk
• Substitutes: renovation, knock-down-rebuild, conversions

 

Expectations (the lens over everything above)

• Builders build against expected prices, costs, and demand at completion
• Expected future demand and policy capitalised into land prices now
• Policy and sovereign risk / market uncertainty

A few of these matter more than the rest, and a few do double duty: interest rates sit on both the demand ceiling and the cost floor, which is why monetary policy moves building so violently, lifting what buyers can pay and lowering what it costs to build at the same time.

Within that picture, supply works on housing in two distinct ways, and it is worth keeping them apart because they call for different fixes. The first is responsiveness. The building pipeline is slow, years from approval to completion, and its capacity grows slowly, so when demand jumps the system cannot lift output to match. The result is not fewer homes in the long run but a price and rent spike while supply catches up, the glitch you see when demand peaks against a system that can only respond at its own pace. The second is the level of building. A standing regulatory and cost burden, the dollar costs and time costs above, can hold the baseline rate of construction below what it would otherwise be, year in and year out, independent of any surge. The first is a timing failure; the second is a level effect.

Both are real and both matter. But the list is the reason neither is the whole story. Supply responsiveness and the regulatory burden are powerful terms in the equation, not the equation itself, and a piece that treats either as the sole cause makes the same mistake as the immigration-only account it is criticising. With that said, supply is where the most persistent and least cyclical of these forces sit, which is why it earns the scrutiny that follows.

Why look at supply?

So given the long list, why look at supply? Two reasons, and they are two views of one shift. First, across forty-five years, most of the periods where building fell short of what the population required come after the mid-2000s. Second, completions per thousand people have been trending down over the same period; I come to that chart below. Built differently, the two measures agree, and that agreement is what makes the change hard to wave away as an artefact of either one. What changed most is the requirement: it rose and grew volatile. But a requirement the building system cannot match is a supply story as much as a demand one. The rising line merely exposed how flat, and how slow to lift, the supply response is, which is why supply is where the rest of this piece looks. There is a sharper version of the point. Population growth reached similar heights in the late 1980s, and building broadly kept pace then. That it did not after the mid-2000s, at comparable rates of growth, is the clearest sign the change is on the supply side: the same demand that was once absorbed now produces a shortfall.

The second view normalises by population. The chart above shows dwelling completions per thousand people, read through the building cycle rather than at any single point in it, drifting down over the same period. Some of the recent low is the denominator: population ran ahead while completions held in their band, so the rate falls even when building does not. But the decline predates the surge and shows through the cycle, which a fast denominator alone does not explain. It is consistent with a build rate that has stayed roughly flat while the population it must house has grown, the same flatness the first chart showed, seen here relative to people rather than relative to need.

Known Demand Is Not an Excuse for Supply Failure

A well-functioning market anticipates and responds to known, foreseeable demand. Australian population growth has been exactly that – visible for thirty years, rising to the top of its historical band since the mid-2000s, and extensively modelled and projected by government agencies throughout.

The post-COVID acceleration is real – the charts show it clearly, and it did contribute to the acute rental crisis of 2022-23. But that spike sits on top of a supply side that was already slow to respond, well before the surge arrived. Even without the post-COVID surge, Australian housing supply struggled to lift when demand rose.

When demand rose to its recent highs, supply could not lift to meet it. That is not a demand problem. It is a supply failure. Immigration contributes to demand pressures, but it does not explain why supply failed to respond. In a functioning market, that demand would have been met with higher supply rather than higher prices.

The analogy is simple: if a city consistently can't build enough supermarkets despite decades of population growth, you don't blame the people for eating food. You ask why nobody can get a planning permit.

Tokyo is the proof of concept: a city of 37 million, long one of the world's most populous cities, that has avoided the sustained price escalation seen in cities like Sydney – not because demand was low but because supply was legally permitted to respond. Auckland provides a sharper test still: following major upzoning reforms in 2016 and 2021, rents and prices softened relative to the rest of New Zealand – same immigration settings, different planning rules, different outcome. The variable that differs between these cities and Sydney isn't demand. It's whether local governments can block construction.

Worker Immigration Expands Capacity – But Housing Adjusts Slowly

There's a further reason to be cautious about blaming immigration for housing stress: for the dominant category – working-age migrants – the long-run impact on housing supply and demand is roughly offsetting.

A working migrant needs somewhere to live. That's a unit of demand. But that same worker also participates in the economy, pays taxes, works in construction trades, provides services, and contributes to the productive capacity that builds and maintains housing. The mainstream economics literature broadly supports this: working-age immigration is roughly neutral for housing affordability in the long run, in economies where supply can respond.

The qualification is timing. Housing supply adjusts slowly – approvals, financing, and construction take years, not months. So in the short run, a surge in working-age arrivals can tighten markets even if the long-run effect is neutral. That's a real dynamic, but it's an argument for smoother immigration intake management, not for treating immigration as the root cause.

And even setting the housing ledger aside, there is a deeper objection to setting migration by the housing numbers. Immigration is not a housing instrument. It bears on the labour market, on skills, on the budget, on inflation, on the ageing of the population, on education exports, on a great deal at once. To set it by the building breakeven line is to subordinate all of that to one downstream effect, to let housing govern a lever that the whole economy relies on. The level of migration is a legitimate debate, but it has to be argued on its own full terms, not settled by the housing shortfall. Cutting it to fix housing makes the same error this piece warns against everywhere else: it pulls a many-purpose lever for a single purpose.

Student Immigration is Different – But It's Also an Export Industry

The more legitimate concern is international students. They are rental-heavy, urban-concentrated demand – arriving in large numbers into exactly the suburbs and dwelling types already under most pressure. And unlike working migrants, they only make a limited labour supply contribution to the domestic economy in the short run. They are also a large part of the recent surge in migration numbers.

But international education is one of Australia's largest export industries, worth tens of billions of dollars annually. Any honest accounting needs to include that benefit alongside the housing pressure.

Even so – even if you dramatically cut student visas tomorrow – the underlying supply problem would remain. You'd be treating a symptom, not the disease.

The Real Culprit: Planning, Zoning, and NIMBY Government

So what is the disease? Planning did not suddenly worsen after the mid-2000s. What it does is hold the supply response flat and slow, so that when demand rose to meet it, the shortfall appeared. Of the forces in that list, planning is the one most worth focusing on, because it is the part government can actually change. A fundamental failure in Australian housing is that supply cannot respond to demand signals, even when demand has risen clearly and predictably. The reasons are well understood:

Zoning restrictions prevent higher-density development in established suburbs, protecting the asset values of existing homeowners at the expense of future residents who need somewhere to live.

NIMBY local governments systematically obstruct, delay, and water down development approvals. Objection processes, heritage overlays, height limits, setback requirements, and design review panels all add time and cost to housing delivery.

Approval delays themselves are a significant cost. A project that takes three years to approve instead of one carries three years of financing costs, holding costs, and opportunity costs – all of which flow through to the price of the finished dwelling.

Infrastructure cost-shifting loads the cost of roads, water, sewerage, and parks onto new developments, making housing more expensive to build and therefore less of it gets built.

Other constraints matter, but planning is the one that most determines how readily supply can respond.

These are not isolated frictions. They are the predictable output of a system that gives existing residents the power to veto new housing. The cumulative effect is a market that cannot clear. Demand arrives. Supply cannot respond. Prices rise. Renters and first home buyers absorb the pain while established homeowners quietly appreciate the outcome.

The standard economic test for which side of the market is driving a price rise is simple: watch quantity. If prices rise and quantity rises, demand shifted. If prices rise and quantity stagnates, supply failed to respond. Australia's housing market has delivered the second pattern when demand surged – prices up, completions slow to lift to meet it. In a functioning market, rising prices are supposed to induce more supply. The fact that ours didn't is not an immigration story. It is a planning story.

The failure is largely one of responsiveness, not of a long-term chronic under-building against a steady demand. Through the late 2010s, completions roughly tracked the dwellings the population required; supply was not failing in level. What it cannot do is accelerate. The pipeline runs years from approval to completion and the industry's capacity grows slowly, so a fast demand impulse opens a gap that takes years to close. That reframes immigration rather than dismissing it. A supply system this slow to respond is precisely the one a rapid migration surge overwhelms, so the pace of population growth is a genuine accelerant of the acute squeeze, even though it is not the structural cause and the gap reverts once the pulse passes and building catches up. Supply is primarily the problem in the sense that it responds too slowly and weakly, not in the sense that it has chronically under-delivered, and that is the sense in which immigration is a real short-run pressure but not the underlying reason housing is expensive.

The chart below shows the dynamics in the data. Approvals, commencements and completions have cycled through boom and bust since 1984, with completions trailing approvals by the years a project takes to build. That lag is not itself the failure; it is the pipeline working as it must. The failure shows in the cycle's behaviour: when a demand impulse arrives, the system cannot lift output quickly to meet it, and the gap closes only slowly as the pipeline grinds through. The government's target of 1.2 million well-located homes over five years from July 2024, the red dashed line, sits well above current approvals and completions, which is less a measure of chronic past under-building than of how far short the system falls when asked to accelerate.

Cutting immigration would likely provide some rental relief over two to three years. But the planning system would still be broken. The next demand impulse – domestic household formation, internal migration, whatever it is – hits the same constrained supply and you're back to the same problem within a decade. You've bought time without using it.

What Would Actually Help

Getting serious about housing affordability means getting serious about planning reform: rezoning established suburbs near transport corridors for genuine medium and high density; streamlining and time-limiting approval processes; removing the ability of local governments to indefinitely obstruct state-approved development; and reforming the tax treatment of investment property, where negative gearing and the capital gains discount inflate demand for existing dwellings rather than new supply.

None of this is easy. All of it involves overriding the preferences of existing homeowners and the councils that represent them. That's precisely why it hasn't happened – and precisely why the housing crisis has persisted through multiple governments of both persuasions.

The Bottom Line

Australia's unusually high immigration rate is a legitimate policy question. The pace is hard to justify on purely economic grounds and the volumetric debate is overdue. That debate should be had on its own terms – labour markets, skills shortages, infrastructure, social cohesion, wages. These are legitimate questions. But they are different questions from the housing question, and conflating them produces bad answers to both.

Immigration is not the structural cause of Australia's housing crisis. Decades of visible, foreseeable demand have simply exposed a supply system that was never allowed to function properly. Fix the planning system, and Australia can better absorb its population growth. Leave it broken, and no immigration target will solve the problem.

Don't blame the people needing homes. Ask why nobody can get a planning permit.

Updates

I edited in June 2026 to add nuance to the argument being made.