When it comes to rental affordability, the two main ways Australian governments provide assistance to those who would struggle most (or miss out on housing entirely) are through the provision of public housing, and through the payment of Commonwealth Rent Assistance (CRA) to households with Centrelink incomes in the private rental market.

Background

These programs are generally the responsibility of different levels of government. State governments provide public housing (although sometimes with federal funding), while the federal government provides CRA (although state governments also provide some cash-based assistance to renters). But the two schemes also represent very different philosophical or political economy approaches to housing support. Public housing is a direct intervention involving government provision of goods and services, while CRA is premised on the primacy of the market and private rental, with the government role simply to assist those on the margins.

To put it crudely, the former is a social democratic approach, the later a neoliberal one.

Unsurprisingly given this broader political economy, since the 1980s we have seen an increased role for rent assistance at the expense of investment in public housing. (The parliamentary library has provided a nice summary of the shift in the 1980s and 1990s). A comparison of how the two approaches to rental affordability assistance stack up (and for who) therefore not only has relevance for the policies themselves, but also has broader political economic implications.

The following comparison does not deal with important issues of renters’ rights, landlord behaviour, or the maintenance and condition of properties, but simply focuses on rental affordability – “following the money” at a number of levels. While the data is South Australian, the patterns are likely to be similar in other states.

Results: Rental Affordability and Costs to Government

The results can be summarised in a simple graph, noting that it refers only to public housing (i.e. government-owned and managed housing) and does not include community or other social housing.

The numbers are based on calculations made primarily from Productivity Commission data, supplemented by the SA Housing Trust financial statements, and are summarised in the table below. The rationale and detail are set out in the methodology section below, along with some caveats.

Based on these figures it is clear that public housing provides a higher level of support to tenants and gets renters out of housing stress. Further, public housing costs the government/taxpayer about the same as CRA in year-to-year expenses, and slightly less when capital costs and changes in asset valuations are taken into account.

These figures do not include, on the one hand, the opportunity cost of the upfront investment in public housing, and on the other hand, the full capital gains and land tax implications. There are some estimates of these below, but the figures get rubbery and are incomplete.

In summary though, the conclusion from the more concrete rental affordability comparison is simple and stark: public housing is better for tenants and for tax-payers.

Commonwealth Rent Assistance to private renters is important (and is currently too low), but even an increased CRA is no replacement for public housing. The fact that recent years have seen a chronic under-investment in public housing and a preference for support in the private market is another neoliberal own-goal.

Method and Explanation

Comparative Level of Support

The Productivity Commission (PC) data shows that in June 2021, some 26,804 SA public housing tenants were paying less than market rent (Table 18A.5). This was around 89% of all SA public housing tenants. That is, 11% of public housing tenants received no financial support for renting: they were paying the full market rate (and in theory, with efficient management the government should have been receiving a return on investment from these renters).

For the rest, there is a simple calculation of the level of financial support. PC Table 18A.5 estimates that for the week of 30 June 2021 the difference between total market value of rent for all public housing and the rent actually collected in South Australia was $3,483,000. Based on the number of tenant households paying less than market rent, this equates to an average subsidy of $130 per household (per week). (With higher average market rents, the figure for the whole of Australia is $192 per week).

By comparison, the maximum CRA at that time was $70.40 per week for an individual household (or $82.81 per week for a household with 2 children, going up to $93.52 for three of more children).

There are of course great intricacies in eligibility criteria and rates for CRA, and there are fundamental difficulties accessing public housing with long waiting lists, but on the basic comparison above, provision of public housing provides far greater level of support per household than Commonwealth Rent Assistance.

Implications for Rental Affordability

The provision of public housing is not only a more generous benefit to those who could not afford market rental, it makes a significant difference to the household budget. Public housing rent in South Australia is capped at 25% of household income, so by definition no public housing tenant should be in housing stress (based on the widely-used 30/40 rule: housing stress = being in the bottom two income quintiles and paying more than 30% of income for housing).

By comparison, calculations I did for a recent SACOSS Cost of Living Report showed that even with CRA, key Centrelink recipients would still be in significant housing stress. The rental affordabilty table from that report is reproduced below and is based on SA government data for median rental prices in the cheapest Adelaide suburbs of $300 per week for a 2-bedroom unit, and $400 for a 3-bedroom house.

Overall, ACOSS estimates that 45.7% of CRA recipients nationally are paying more than 30% of their income on housing as CRA has failed to keep up with increasing rents over recent years. Accordingly, advocates are calling for a 50% increase in CRA (alongside significant investments in public housing). However, on the figures above, even this would not be enough to get many households out of housing stress (as other income support increases are also urgently required).

Again, CRA fares badly in the comparison to public housing.

Cost to Government

The figures above might suggest that public housing is more expensive than CRA (because the support is more generous). However, this is not necessarily the case due to the rent paid and capital gains (which, as I have pointed out elsewhere in relation to private housing, is a game-changing income stream which is often ignored).

Current Yearly Costs

The Productivity Commission data (Table 18A.43) shows a net current expenditure of $10,361 per public housing dwelling (a figure in line with the South Australian Housing Trust financial statements (SAHT) for 2020-21 when capital costs are subtracted from expenses). However, calculating from PC Table 18A.5 which shows total rent collected in the week ending 30 June as $4,169,000, the rent collected equates to $6,730 annually per property (rent collected x 52 weeks, divided by 32,212 properties as per Table 18A.43). Deducting this rental income from the annual expenditure, the net current expenditure is $3631 per household.

The expenditure for CRA is simply the weekly rate of $70.40 multiplied by 52 weeks for the year. The total is $3,661, which means the recurrent cost of assisting one household through CRA is about the same as the public housing current costs above.

Including Capital Costs

The Productivity Commission data (Table 18A.43) shows an annual depreciation of $1,578 per property in South Australia – which equates to nearly $51m for the whole public housing estate (again, broadly in line with SAHT data). However, the Productivity Commission does not account for any capital gain.

By contrast, Note 5.2 to the SAHT financial statements includes the periodic revaluation of public housing rental properties with increases in 2020-21 of $69.6m in land value and $24.4m in buildings. This $94m equates to $2,799 per household when allocated across the 33,590 public houses (including State-Owned and Managed Indigenous Houses [which are included in the SAHT data, but not in the PC data used above]).

The difference between this capital gain of $2,799 and the depreciation of $1,578 per household results in a net capital gain of $1,221 per public housing household. This is not cash income, but represents a fair accounting of capital income based on the official data.

Subtracting this capital income from the net current costs to government gives the bottom line of public housing costs of $2,410 per household. Again, this is cheaper than the $3,661 yearly cost of Commonwealth Rent Assistance to one household.

However, this capital accounting is not complete – although in going further the figures get rubbery. For this reason, they are not included in the summary table above, but are discussed here.

The big drawback of public housing is the cost of the upfront investment to build the houses, or as the Productivity Commission describes it: “the cost of the funds tied up in the capital used to provide social housing”. While the PC suggests caution in interpreting the data, at Table 18A.43 it estimates this “Indicative user cost of capital” at $18,933 per year for each SA public housing dwelling – an imputed figure which dwarfs all other public housing costs (and would make CRA much more attractive to government).

Yet this accounting is one-sided and does not complete the picture because while capital costs are imputed, the capital gains are not included – and not even fully recognised in the SA Housing Trust revaluations of its rental properties.

The SAHT (Note 5.2) estimate of a total value of rental land and buildings of $7,121m at the beginning of the 2020-21 year, and so the $94m revaluation noted above represents only a 1.3% annual increase in the capital value of public housing assets. This was at a time when market analysts CoreLogic were suggesting a 17.9% increase in Adelaide housing prices. It is not clear why the capital revaluation was so low, but is perhaps based on accounting practice based on depreciated asset (book) value rather than changes in current market price.

That said, market prices are very volatile and uncertain, and both the Productivity Commission’s indicative user cost of capital, and any market valuations, require bold assumptions about interest rates and market behaviour. But if, or to the extent that the Productivity Commission’s imputed cost of capital represents a market value of public housing assets, it is equivalent to about 8.5% of the $7,121m public housing estate valuation. That is, when housing prices go up by more than about 8.5%, the government makes a net capital gain on public housing (or loss when less than that) even with the inclusion of the cost of capital.

Again though, I am not confident of this figure as a true market costing or representation of capital gain. If the market value of public housing properties is significantly higher than the book value, then a smaller capital gain is needed to balance the cost of capital.

Final Caveats

There are a two final caveats to the calculations above.

Firstly, there is no accounting for differences in land tax revenue to government from public or private ownership of rental properties. The SA Housing Trust financial statements show $139m in land tax equivalent expenses – which was 42% of their rental property expenses. While this is the government paying money to itself, it is legitimate accounting and necessary for a comparison to private market support costs – not least because if rental properties are left to market provision with government subsidy through CRA, the land tax take may be considerably less.

For instance, apportioned evenly across all public housing rentals, averaging the total land tax paid across all public housing properties gives a land tax bill of $4,179 per property. By contrast, if a property with the average (book) value of a public housing dwelling ($215,000 based on the $7,121m total valuation) were a private landlord’s only investment property, they would pay no land tax. Even if the private landlord owned 4 such properties, they would only pay $2,535 land tax – about 15% of the housing trust land tax bill for four properties.

It is impossible to calculate the extent of the difference in land tax income for government because it depends on private landlords’ individual circumstance, but rental housing in public hands clearly maximises land tax revenue. Accordingly, in the comparison of public housing costs vs CRA, public housing costs to government should theoretically take account of this land tax gain.

Finally, while the comparison is not a support for existing tenants or a cost to government, there is a difference in the impact of public housing and CRA on rental affordability more generally. CRA payments create no new housing, and while they assists renters into the market they also allow them to bid up rental prices because those renters have (slightly) better rental affordability. By contrast, public housing brings new supply to the market and puts downward pressure on market rents.

Given the above caveats, my calculations are incomplete, but even without precise cost-benefit calculations, the figures above that are available and robust suggest a strong case for public housing.