Department of the Treasury's Design and Delivery of the Housing Australia Future Fund and the National Housing Accord Facility
SCALE
- A $10 billion fund and a $350 million facility target 40,000 new social and affordable homes by mid-2029.
- 1,432 homes were completed by April 2026; half the target is forecast by mid-2028.
- Average government support per dwelling is $770,387 nominal over 25 years ($392,518 in net present value).
COMPLIANCE
- A dedicated statute establishes the fund, its credit, and the minimum it must disburse each year.
- A whole-of-government risk-management policy requires entities to embed, document and periodically review risk management, and to assign clear ownership of risks shared between entities.
- A whole-of-government evaluation policy is the standard the department invoked when it cancelled its planned process evaluation.
ECONOMY
- Concessional lending costs government around $221 million a year in foregone interest on $4.491 billion of budgeted loans - a cost the audit had to calculate, not one the headline figures surface.
- Later funding decisions added $2.6 billion for concessional loans and $3.1 billion over 15 years for availability payments running to the 2050s.
- Support per dwelling averages $770,387 nominal over 25 years, with no benchmark against which to judge it.
EFFICIENCY
- No measure exists for cost per dwelling, cost of administering the program, process speed, or timeliness of actual against planned delivery.
- Governance, risk-management and evaluation arrangements were each finalised more than a year - in one case 17 months - after the programs began.
- The shared risk register was meant to be reviewed quarterly and the two risks rated high were meant to carry documented actions; neither happened.
EFFECTIVENESS
- 1,432 of the 40,000 homes had been completed by April 2026, with 20,000 forecast by mid-2028.
- Reporting against outcomes such as improved housing access and increased private investment is not available at all.
- The planned process evaluation was cancelled, so the design will not be tested early enough to correct it.
1. No minutes and no decisions register leave governance action untraceable
Documentation Coordination
The program board produced no minutes between October 2024 and July 2025, and the steering committee 'has not documented a review of its terms of reference' since 2024 (para 3.40). The audit concludes that 'the absence of minutes or a functional decisions register... means it is not apparent that the board has acted upon reporting received' (para 3.43). A Documentation gap that undermines Coordination: the governance bodies exist, but their deliberations leave no trace.
- Function: Governance
- Stakeholders: Governing body
- Quality: Functioning oversight and governance
2. The risk management plan was approved a year after the program began
Responsibility Monitoring
The department 'approved the HAFF program risk management plan on 24 February 2025, more than a year after the program was established' (para 3.52). A Responsibility gap in the initial phase: the program ran for over a year with delivery already under way before its formal risk-management structure existed, weakening Monitoring of that whole interim period.
- Function: Governance
- Quality: Sound risk management
3. The risk register looks five years ahead at a twenty-five-year commitment
Goal-setting Monitoring
Although availability payments run for 25 years, 'the delivery risk included in the risk register refers only to a failure to deliver 30,000 homes in five years, and not long-term objectives' (para 3.64). A Goal-setting mismatch between the program's actual time horizon and what its own Monitoring arrangements are built to see.
- Function: Planning
- Quality: Clear objectives and goal-setting Sound risk management
4. Risks shared with the delivery agency are named but not managed
Responsibility Coordination
The memorandum of understanding between the department and the delivery agency requires that risks rated high get documented actions and regular reporting, but 'this has not occurred for the two risks rated as high risk' (para 3.73); the register itself 'was intended to be reviewed quarterly... This has not happened' (para 3.74). A Responsibility gap at exactly the point - risk shared across entities - where the whole-of-government risk policy demands the clearest ownership, weakening Coordination between the two bodies.
- Function: Governance
- Value: Regulatory system
- Stakeholders: Delivery partner
- Quality: Sound risk management
5. No measure exists for the efficiency the program promises
Monitoring
Despite a stated outcome that the program is 'improved and streamlined with efficiency gains', 'there are no relevant measures related to the cost per dwelling, cost of administering the program... process speed and the timeliness of actual delivery against planned delivery' (para 3.92). A Monitoring blind spot with a direct Finance consequence: cost per dwelling can be reported after the fact but never benchmarked against a plan.
- Function: Finance
- Value: Finance
- Quality: Cost control and value for money Streamlined, standardized processes
6. The review meant to catch problems early was cancelled
Results review Guidance
'Treasury has decided that the process evaluation will no longer occur' (para 3.107), even though its own assurance plan, reflecting on an earlier program, records that a late review 'limit[ed] incorporation of earlier learnings in program implementation' (para 3.109) and that earlier reviews would have been beneficial. A Results review cancelled against the department's own documented Guidance on why early evaluation matters.
- Function: Governance
- Quality: Functioning oversight and governance
7. Public reporting still does not show whether the program is working
Reporting
The audit's overall conclusion finds 'insufficient transparency on program delivery, costs and impact', and that reporting against outcomes such as improved housing access and increased private investment 'is currently not available' (para 9). A Reporting gap that leaves the parliament and taxpayers unable to judge whether a multi-billion-dollar, 25-year commitment is delivering - the subject of the audit's fifth recommendation.
- Function: Governance
- Value: Community
- Stakeholders: Lawmaker
- Quality: Reliable, integrated information base
8. The people the fund exists to house were never consulted
Design
While the department consulted industry and government stakeholders extensively, 'there was no direct consultation with the intended beneficiaries of the HAFF' (para 2.88) - the tenants the program exists to house. A Design process missing the very people it is designed for: those most affected had no channel through which to shape it.
- Function: Planning
- Value: Community
- Stakeholders: Beneficiary
- Quality: User-centred design and usability
9. Concessional lending carries a cost the headline figures do not show
Reporting
Because the loans are interest-free for 25 years, the audit estimates that 'the cost to government of HAFF concessional loans will be around $221 million per annum' (para 2.44) in foregone interest on $4.491 billion of budgeted loans - a cost the audit itself had to calculate, not one the headline figures surface. A Finance cost concealed by the funding instrument that creates it, visible only once Reporting is reframed in net-present-value terms.
- Function: Finance
- Value: Finance
- Quality: Cost control and value for money
Control focus
| ICS phase | Control function | Cases |
|---|---|---|
| Functions applied to all stages | Documentation | 1. No minutes and no decisions register leave governance action untraceable |
| Coordination | 1. No minutes and no decisions register leave governance action untraceable<br/>4. Risks shared with the delivery agency are named but not managed | |
| Reporting | 7. Public reporting still does not show whether the program is working<br/>9. Concessional lending carries a cost the headline figures do not show | |
| Initial phase | Responsibility | 2. The risk management plan was approved a year after the program began<br/>4. Risks shared with the delivery agency are named but not managed |
| Goal-setting | 3. The risk register looks five years ahead at a twenty-five-year commitment | |
| Guidance | 6. The review meant to catch problems early was cancelled | |
| Design | 8. The people the fund exists to house were never consulted | |
| Work processes | Monitoring | 2. The risk management plan was approved a year after the program began<br/>3. The risk register looks five years ahead at a twenty-five-year commitment<br/>5. No measure exists for the efficiency the program promises |
| Completion of process | Results review | 6. The review meant to catch problems early was cancelled |