Against ‘Affordable’ Housing

The building that sparked the controversy is Watermans Residences, a 30-storey tower completed in 2024. It is located right on Sydney Harbour, in Barangaroo, where the median rent for a two-bedroom unit is currently a staggering $2,400 a week, or $124,800 annually. So-called ‘affordable’ housing in the building rented for less than that: an estimated $1,800 a week; $93,600 annually. As part of the trade for reduced rents, ‘affordable’ housing tenants are unable to access the pool, gym, or even, yes, the main entrance. And for all this they’re still paying almost the entirety of a median Australian salary in rent. 

Watermans Residences is not an isolated case. Shelter NSW research from the same period highlighted an ‘affordable’ 1-bedder in Bondi for $925/week, and an 'affordable' 3-bedder in Coogee for $1,400/week. New units typically cost 20-30% more than the median for the suburb, so generally, all the ‘affordable’ housing discount does is bring the cost of a new apartment down to the same level as an apartment that got built a decade or so earlier. 

Put simply, rents for new-build apartments in Australia’s major cities are so high that, even after a discount, the only people who could possibly benefit from a moderate subsidy are those already earning relatively high incomes. 

'Affordable' housing gives politicians a great announceable, but it really is nothing more than a subsidy for those on average incomes. And it is not a small subsidy—the amount we are spending on ‘affordable’ housing, both per-unit and across the nation, is much larger than most politicians, policy makers, and commentators realise. 

The reality is that ‘affordable’ housing is deeply, intrinsically flawed. It is a wasteful use of public funds that redirects government support away from the people who need it most. It provides those who do benefit with help they don’t particularly value, and the criteria and rules set up to administer it result in bizarre outcomes. Most importantly, the policy is wholly ineffective at delivering the outcome Australians actually want: not ‘affordable’ housing, but inexpensive housing. 

It’s time for Australia to move on from ‘affordable’ housing.  

If you don’t know what ‘affordable’ housing actually is, you’re not alone: it’s a nebulous term, hard to pin down, and used to refer to a vast array of housing types. In fact, defining ‘affordable’ housing is so confusing that the National Housing Supply and Affordability Council dedicated a whole section of their 2025 State of the Housing System report to attempting to establish a clear definition.¹¹1

Before we unpack what ‘affordable’ housing is, it’s important to explain what it is not.

Most importantly, ‘affordable’ housing is not social housing. Social housing refers to both public housing—homes owned by state governments and rented to people on low incomes for around 25-30 per cent of their income—and community housing, which is similar to public housing, but where homes are managed by a non-profit Community Housing Provider (CHP).²²2

In Australia, social housing is targeted at those with very low incomes: at the last census, the typical household living in public housing earned around $500-650/week—around a third of the median household income.³³3

These figures have been generated using ABS Tablebuilder, TENLLD, HIND. They excludes negative income and income not reported. At the same census the national median household income was $1,746.

Social housing is an important form of social insurance. Around 80% of people moving into social housing are “in greatest need”, meaning that they are currently homeless, living in unsafe or inappropriate housing, or paying extremely high rental costs. A robust supply of social housing keeps people out of homelessness when they fall on tough times and provides a stable home for many who would struggle to afford even basic accommodation in the private market.  

Like social housing, ‘affordable’ housing is discounted housing subsidised by the government.⁴⁴4

Unlike social housing, it’s explicitly targeted at higher earners, and rents are generally set at a discount in relation to market rates, as opposed to tenant incomes. 

This is where the inconsistencies and difficulties defining ‘affordable’ housing come in. Some programs require only a 10% discount, while others require 20%-30%.⁵⁵5

Some states require ‘affordable’ housing to be operated by a CHP, while others let any private housing provider do it themselves. Eligibility and income limits are opaque and vary hugely. 

There’s no centralised reporting or data collection: remarkably, for a sector that receives billions in government subsidy each year, nobody knows how many ‘affordable’ housing units there actually are. 

Nobody designed Australia’s ‘affordable’ housing system to look like this. This is simply the accumulation of uncoordinated state, federal and local government programs, optimised for the sole purpose of giving politicians a ribbon to cut, and a press release that says “affordable housing, coming soon”. 

Because of these inconsistent definitions across the country, for clarity in this piece we are going to use the following generalised definition:

‘affordable’ housing is housing which the government requires to be rented, at a discount to market-rates, to households on moderate incomes. 

One of the things we know from examples of ‘affordable’ housing in Barangaroo, Bondi, and across our entire country is that it is, in many cases, far from affordable. Much of the nation’s ‘affordable’ housing stock consists of new or nearly-new units in high-cost locations, meaning that even after the government subsidy, many ‘affordable’ units are still well out of the reach of those on low incomes. 

For example, St George Community Housing, who manages the ‘affordable’ housing in Barangaroo, reveals in their annual report that their ‘affordable’ units rent for almost two and a half times more than their social housing units do.⁶⁶6

 

The now-defunct precursor to modern ‘affordable’ housing programs was plagued by similar problems. The National Rental Affordability Scheme (NRAS) was extremely poorly targeted, with an eligibility criteria so broad that half of renters in Australia were eligible for a place.

Some schemes even have minimum income eligibility criteria. In Victoria, “rental stress” definitions create a regulatory ceiling, prohibiting individuals from renting ‘affordable’ units if the cost exceeds 30% of their gross income. The result is ‘affordable’ housing that the people most in need are legally barred from living in.

Consider the above ‘affordable’ housing unit in Melbourne, listed at $561 per week—$29,172 annually. To avoid the 30% rental stress ceiling, a tenant must earn at least $97,240. However, the maximum income a single person can earn in order to qualify for the unit is $73,530. Because the minimum income required by one rule is higher than the maximum income allowed by the other, a single person can never legally qualify for this home. 

Even a single parent with one child—exactly the kind of household for which a 2 bedroom unit is well-suited—is unable to legally rent the unit, because the maximum income for a single parent is $93,873, just $3,367 below the rent-stress ceiling. 

The result is that the only people eligible to rent this unit are singles with two kids, or couples, earning roughly average incomes—not too much more and not too much less than the average income.

Absurd situations like this are the outcome of a thicket of well-meaning but arbitrary thresholds and rules of thumb that comprise ‘affordable’ housing in Australia. In many cases, there’s only a vanishingly small slice of the income distribution that is eligible for ‘affordable’ housing, and those who do benefit are generally on average rather than low incomes. 

And if at the end of all this you happen to be eligible for ‘affordable’ housing, you’re not guaranteed a place: ‘affordable’ housing sometimes operates as a lottery. Rather than being needs-based, like the social housing system, whether or not an eligible party receives an ‘affordable’ housing unit is often effectively random.⁷⁷7

Now that we’ve established what ‘affordable’ housing is, and who gets to live in it, we need to explain how ‘affordable’ housing units are provided in Australia. Generally, there are two main sources of ‘affordable’ housing: explicitly government-funded projects, and units built by developers as a requirement of planning regulation. 

In the first case, the government’s financial contribution is clear: a payment or other subsidy, such as a zero-interest loan or tax concession, is made to an operator who then builds, buys, or leases homes at below-market rents. 

This is where a large share of the money from the Housing Australia Future Fund (HAFF) has been spent. For example, Evolve Housing, a major ‘affordable’ housing provider, received grants to build 1,253 social and affordable housing units from the first round of the HAFF. Of these, 1,083, or 86%, are ‘affordable’ rather than social housing. Overall, across its first two rounds, the HAFF has funded 9,366 ‘affordable’ housing dwellings—slightly more than the number of social housing units it has funded (9,284). 

The second source of ‘affordable’ housing, units delivered as part of planning requirements, might on first glance appear free to governments. After all, the homes are built by private developers without any payments being made by the government. Indeed, the financial cost to the government of these units is zero. However, what really matters is the opportunity cost: what the government could have done instead of requiring the units be leased out at a below-market rate. 

We can demonstrate this by returning to the Barangaroo apartments. A typical 2-bedroom unit in this part of Sydney leases for an eye-watering $2,400/week, meaning that the 25% discount for each ‘affordable’ unit represents a cost to the developer of around $600/week. 

The developer, Lendlease, agreed to this as part of a deal with the state government to secure approval. Instead of imposing this requirement, we could instead have taken the equivalent amount in cash, and spent it on more targeted housing programs.⁸⁸8

In effect, ‘affordable’ housing requirements are a tax on developers, which is then immediately spent on buying down rent for a few lucky recipients of discounted units.

Although the financial cost to the government is zero, the real cost – the money the government could have had instead of requiring discounted rent – is almost $400,000 for every single unit.⁹⁹9

Not all areas are as expensive as Barangaroo, but the typical ‘affordable’ housing unit in New South Wales still receives six figures of implicit government subsidy. Recent research by Shelter NSW estimated the value of this foregone revenue between $89,000 and $172,000 per unit in exchange for 15 years of discounted rent.

Development rights are only valuable because our planning systems have choked the supply of new housing in the most desirable parts of our cities, driving up prices for everyone. Because of this, large-scale planning changes can unlock huge amounts of value, and charging developers a modest proportion of this uplift can make sense in the short-term.¹⁰¹⁰10

In this context, taking the value capture as discounted units rather than cash forgoes revenue that could do far more good elsewhere. Wasteful, but not actively harmful. 

When there is no corresponding planning uplift, these charges represent a tax on new homes, reducing supply and driving up rents and prices.¹¹¹¹11

In this case, not only is ‘affordable’ housing wasteful, it makes the housing crisis worse for everyone else.

As an example of ‘affordable’ housing policy going wrong, the Commonwealth recently passed build-to-rent (BTR) legislation that provided substantial tax concessions to developers building new homes for rent rather than for sale, in exchange for the provision of ‘affordable’ housing across their projects. 

What are these tax concessions getting us? Assemble, one of the leading BTR developers, backed by superannuation, has been delivering a spree of housing projects across Melbourne. Their units typically rent at a substantial premium over market rates, though with few of the additional amenities you might see in BTR from other developers trying to justify the premiums they charge.

A recently completed project in Kensington offers a one-bedroom apartment for rent starting at $588 per week. The median rent for a similar apartment in the same area is only $450 per week—almost 25% less.¹²¹²12

Under the Commonwealth’s requirements, Assemble would need to offer ‘affordable’ housing one-beds in their building at $441 per week, only $9 below the suburb’s median rent.

Is subsidising the rent of brand new units—which still cost more than 80% of a typical social housing resident's entire income—truly the best use of $7,000 a year in government money?

Financial support makes a bigger difference to someone’s wellbeing the less money they have. If you’re on a very low income, extra cash means that you can afford more nutritious food, new clothes, or to run air conditioning more often on hot days: things that meaningfully improve your quality of life. To someone on a higher income, the same amount of money makes less of a difference; it might mean a slightly nicer holiday or an extra meal eaten at a restaurant rather than at home. 

When researchers try to quantify this effect, they find that the same amount of money has an exponentially smaller impact as the recipient gets richer. For example, providing a person on jobseeker¹³¹³13

with the $7,000 of subsidy implicit in each of Assemble’s units would improve their wellbeing by more than three times as much as someone on $71,000– the lowest income someone could have to be eligible for one of these ‘affordable’ units.¹⁴¹⁴14

This is based on Killingsworth's 2021 study on experienced well-being.

 

This is the classical utilitarian case for redistribution, and carries a clear implication for policymakers: if you’re trying to improve people’s wellbeing, you can do far more good with your dollars by providing support to those on the lowest incomes first. 

With almost 40,000 people homeless and 169,000 households now on the waitlist for social housing, there are many, many extremely vulnerable households waiting for support. But governments across Australia are choosing instead to subsidise ‘affordable’ housing units for those who need the money far less. 

By now, we’ve hopefully convinced you that we should target public subsidy to the lowest earners, and that ‘affordable’ housing does not do this. However, even if you reject this analysis and your explicit policy ambition is to improve the housing situation for moderate-to-high earning households, we regret to inform you that ‘affordable’ housing is also very bad at doing this. 

This is because by tying the subsidy to housing units, rather than to the people living in those units, the money provides help in a form that many recipients don’t really value. 

Let’s go back to the $600/week subsidy that the government is providing to each Affordable unit in the Barangaroo building. If offered the choice between receiving $31,000 in income support a year or the right to rent a 2-bedroom apartment in Barangaroo for $1,800/week, which would the tenants prefer? 

It’s almost certain that they would prefer the cash, with which they would likely rent a different, cheaper apartment and increase their spending on food, entertainment, bills, travel and all their other, non-housing costs. If we grant this as true, then these residents are almost certainly deriving less than $600/week of value from the subsidy they receive. 

To borrow a turn of phrase from the great economist Arthur Okun, ‘affordable’ housing is a ‘leaky bucket’. 

On its face, ‘affordable’ housing appears to be a zero-sum transfer from taxpayers to recipients of subsidised housing. But it’s actually much worse than this. Because the residents receiving the support don’t value it at its full cost to government, they are worse off than they would have been with simple financial support. It’s like we’re paying for pints, but only giving people pots.¹⁵¹⁵15

 

Supporters of ‘affordable’ housing also argue that it’s necessary to ensure that teachers, nurses and other so-called ‘key workers’ are able to live a reasonable distance from their work. While this is an admirable goal, ‘affordable’ housing is an inefficient way of achieving it and may end up harming those it aims to help.

Firstly, defining ‘key worker’ quickly becomes a nightmare. To demonstrate the complexity of the definition, the City of Melbourne won a research award from the Planning Institute of Australia for even attempting to quantify what a “key worker” is. 

Many low-to-medium income jobs fundamental to our society’s functioning—such as cleaners, shelf stackers and delivery drivers—are often left out in the “key worker” definitions. But any narrowing or expanding of the definition will create a different set of arbitrary winners and losers. 

The stricter the definition, the more intraprofessional inequality exists—why should one nurse receive the multi-thousand dollar government subsidy, and not another? But the looser the definition, the more the scheme becomes bloated, and the lottery less targeted.

The City of Ryde in NSW encountered this exact issue when they one day randomly decided to exclude public transport drivers, delivery drivers, retail workers, and cleaners from their new definition of ‘key workers’. In a Torres Shire Council submission to the Productivity Commission, the Council highlighted how ‘key worker’ housing was primarily allocated to government employees sent to remote regions, while the predominantly Aboriginal and Torres Strait Islander locals were ineligible because their jobs weren’t on the list.

A second problem with linking someone’s housing to their job is that we make it far harder for them to change either their job or their home. Consider a nurse or teacher who wants to start a family and needs more space, but has a deal on an Affordable 1-bedroom unit that’s too good to give up. Or a train driver considering further engineering study, but who knows that taking time away from their career means losing their home. In much the same way that US-style employment-linked health insurance results in worse health and employment outcomes, job-linked housing can act as ‘golden handcuffs’, trapping people into suboptimal situations. 

As has been the case for every example in this essay, key workers living in ‘affordable’ housing would be better off if they received their subsidy as cash instead. Linking workers’ housing to their job is yet another way in which the ‘affordable’ housing bucket leaks.

Even putting aside the welfare of the workers, if our sole objective was to maximise schools’ and hospitals’ ability to attract and retain staff, then ‘affordable’ housing is still an inefficient way of doing so. The typical worker would prefer a $200/week raise over a $200/week discount on their rent, so offering them the raise rather than the subsidy would be more effective in convincing them to join or stay.

Key workers should be able to live in better homes, closer to their work. It would improve the lives of people who contribute huge amounts to the country, and would make it easier to attract and retain their services, benefitting us all. But ‘affordable’ housing is a bad way of trying to achieve this goal. 

To make the lives of all workers and residents in our country, we need to confront the actual problems that exacerbate the cost of housing. 

Using government funds to simply buy down market rents is not a sustainable solution to the housing crisis. Rents in Sydney have increased by 20% in just the last three years, so the lucky few households that land an ‘affordable’ housing unit are simply back to where they were a few years ago.¹⁶¹⁶16

There is no solution to the housing crisis that does not involve the level of market rents falling substantially. 

'Affordable' housing reflects a misdiagnosis by policymakers, managing the symptoms of the housing crisis rather than its cause. This idea that the market is untameable and uncontrollable is a common one in housing policy. The urban planning lobby—the people who control where and how much housing can be built—famously argues that their profession somehow doesn’t impact the supply of housing.

If you believe that there’s nothing that can be done to influence the private market, then the logic for ‘affordable’ housing becomes clearer. There’s nothing to be done about market rents, and so the best we can do is shield as many people from those rents as possible. 

To be sure, there are factors in the housing market that are largely out of Australian policymakers’ control. Elevated interest rates are a global phenomenon, as are high energy, lumber, and other raw material prices. However, this doesn’t mean that we should simply throw our hands up and accept that market rents will be what they will be. Policymakers control levers that can meaningfully impact supply, demand and rents. They should use them.¹⁷¹⁷17

There are also non-government stakeholders who explicitly benefit from ‘affordable’ housing schemes. CHPs advocate for ‘affordable’ housing because it enables them to provide more housing and help more people. For these providers, ‘affordable’ housing also happens to be both easier and more lucrative than social housing, enabling them to rent out units at prices that subsidise their more costly—and more important—operations. We cannot fault them for this, as remaining financially viable is a key condition placed on them by state-level regulators. CHPs have a singular mission: to directly provide housing assistance to Australians. But just because it helps with their mission, it doesn’t mean it’s a good policy, or the right choice for governments to make. 

Individuals who benefit from ‘affordable’ housing seem to overwhelmingly love it and are likely better off than they were before. But this is expected. A winner of a lottery is glad for the lottery to exist. Good policy should look not just at the winners, but the losers too.  

The most significant player to blame is the Commonwealth Government. ‘affordable’ housing results from the bad incentive structures they created and their historic lack of leadership in housing policy. 

Within this context, State and Territory Governments, strapped for cash and desperate to look like they’re doing something to help solve the housing crisis, do the most low-cost and low-risk political thing possible—‘affordable’ housing. ‘affordable’ housing provides an easy answer when the pressure mounts to look like you’re doing a lot with the smallest amount of cash possible. 

This is in part because policies that fund ‘affordable’ housing—being cheaper than social housing—results in a bigger number that looks better on a press release. 

As Travis Jordan, fellow Inflection Points author, wrote:

affordable housing—and inclusionary zoning, its planning counterpart—have become all the rage because it doesn’t impact public finances at all. No new revenue, no new expenditure, no new liabilities, no new assets. No public finance impact means no political consequences. It’s convenient and expedient. But that doesn’t mean there isn’t a cost. Someone pays—whether that’s another tenant, the state indirectly, or our whole society through reduced building completions.

Delegation of responsibility, cost and political exposure is rampant in Australia’s housing policy landscape. Government-subsidised ‘affordable’ housing is the natural result of  decades of policy cowardice. And it doesn’t even solve the problem.

We’re card carrying YIMBYs: Dominic is secretary of Sydney YIMBY and Ethan works full-time for YIMBY Melbourne, so of course our first policy prescription is for governments around Australia to reform the outdated and inflexible planning systems that are the fundamental cause of our housing crisis. A great place to start would be to allow terraces and townhouses to be built on all residential land in our major cities. 

Inflection Points has already published a lengthy discussion of issues with planning, so instead our policy prescriptions are going to focus on actions outside of the planning system that governments can take to provide support to those struggling with the housing crisis.

By highlighting the failings of ‘affordable’ housing in this regard, we’ve come up with a good set of principles for housing policymakers.

1. Aim low. The most good is done by helping those on the lowest incomes.
2. Trust people. Outside of those most vulnerable, providing people with monetary support, and letting them choose how to spend it, is more efficient than in-kind subsidies.
3. Be explicit. Accounting tricks and policies that hide and obscure costs don’t hide or obscure their effects. Being explicit about costs and tradeoffs lets you measure and manage them properly.

What does a housing policy based on these principles look like? 

First, we should redirect sources of explicit government funding for ‘affordable’ housing, such as the HAFF or the Affordable Housing Bond Aggregator, to social housing instead, providing a deeper subsidy to households on lower incomes. This would deliver much greater value for money, by providing support to those who benefit from it most. While this would result in the HAFF funding a lower number of homes, insufficient government subsidies is not the barrier to new construction, planning is.

Second, we should reform policies that require or encourage developers to provide ‘affordable’ housing, like the NSW Infill Affordable Housing scheme, to instead take contributions in cash.¹⁸¹⁸18

This would reduce absurd outcomes like ‘affordable’ housing that goes for $1,500 a week, or for which practically no one is eligible. Taken as cash, these dollars can do more good elsewhere.¹⁹¹⁹19

We only advocate for this in the context of fully funded Inclusionary Zoning.

Third, what support we do provide for moderate-income renters and key workers should be provided in cash. This should be done through direct payments and schemes such as Commonwealth Rental Assistance (CRA) which should be increased and expanded. This would also address the inefficiencies that come from providing subsidies to units rather than people – empowering people to choose the housing that best suits them. 

By making these common-sense changes, we can make sure the funds we do choose to spend on subsidised housing goes as far as it can possibly go, for the benefit of the greatest number of people. And then governments can focus on the real policy task ahead of them: not just making housing ‘affordable’, but making it inexpensive.

This essay is written in Dominic's personal capacity and should not be attributed to Sydney YIMBY or his employers.