Examining the proposed revamp of Australia’s RDTI regime

Ambitious Australia and reception of proposed RDTI reforms

Recommendations made in the Ambitious Australia report have resulted in several Federal Budget 2026-27 proposals to make structural reforms to the Research and Development Tax Incentive (RDTI) regime, to take effect from 1 July 2028.

Many of the threshold-related changes are uncontroversial, if not welcomed, and broadly align with some of the report’s recommendations, albeit not in quantum.

However, two proposed structural reforms are proving to be (justifiably) controversial. 

These are:

• A 10-year time limit based on company age on the refundability of the increased SME R&D tax offset. 
• The permanent removal of eligibility for expenditure incurred on supporting R&D activities, whether in Australia or overseas (with an Advance finding). 

Together, these reforms represent the most significant narrowing of the RDTI program since its inception. 

Uncontroversial RDTI measures

For completeness, the relatively uncontroversial RDTI measures include:

• Lifting the refundable R&D tax offset threshold from $20m to $50m of aggregated turnover, to encompass both small and medium sized entities (SMEs).
• A 4.5 percentage point increase in all R&D tax offset rates.
• Reducing the non-refundable intensity threshold from 2% of total expenses to 1.5%, pushing more eligible expenditures into the higher rate. That said, retaining the tiered rates disregards the recommendation to remove this complexity and maintains bias towards industry sectors with low non-R&D expenses.  
• Raising the minimum R&D spend from $20,000 to $50,000 for all claimants, with lower spends still eligible only if undertaken with registered service providers (RSPs) or CRCs. This is fortunately less than the recommended $150,000 threshold.
• Increasing the R&D expenditure cap from $150m to $200m, a move that will benefit only the largest R&D claimants but could restore some of Australia’s international competitiveness. Notably, this falls short of the Ambitious Australia recommendation to remove the cap outright.

Numerical examples of these changes are set out in Appendix II. 

From an administrative perspective, the ATO will also receive increased fraud-prevention funding. RDTI claims were highlighted as a key focus of additional compliance activities over the two years from 2026-27. The ATO will receive a further $2.8m over three years from 2027–28 to support the implementation of the new measures.

Leaving aside these uncontroversial measures, we can now examine the anticipated impact of the two key areas of concern more closely. We will also discuss RDTI interactions with the proposed start-up tax loss rules, the now-permanent $20,000 instant asset write-off and tax loss carry-back measures.

Removal of eligibility for supporting R&D activity expenditures

Basis for proposed reform

The first key controversial proposal is that in return for increased R&D tax offset rates on core R&D activities (and in the alleged name of simplicity), the category of supporting R&D activities will be permanently eliminated from 1 July 2028. 

In practice, only core experimental activities – those directly generating new knowledge of technical uncertainties through hypothesis-driven experimentation, where the outcomes cannot be known – will remain eligible for the RDTI.

This proposal aligns with the Ambitious Australia report recommendation to introduce “a deemed rate for supporting activities – a fixed offset benefit relative to the amount of core R&D expenditure claimed.” The 4.5 percentage points increase for core R&D expenditure does represent such a fixed offset benefit, relative to the core expenditure incurred. 

At first blush, this proposal could be seen as significantly reducing the compliance efforts surrounding R&D activity registration and the costing of supporting R&D activities, both in Australia and overseas (where an Advance overseas finding exists). 

However, this reform is arguably predicated on fiscal rather than innovation policy-oriented goals. Supporting activities appear to have increased substantially over the last decade. And in the government’s view, unlike core experimental activities, they generate no additional R&D per dollar of offset. 

RDTI costs reached approximately $4.5bn in 2023–24, and supporting activities are understood to be a major contributor to this growth. These ‘reforms’ are arguably seeking to push back on broader costly claims viewed as diluting the program’s purpose.

The forward estimates included in Budget Paper 2 support this interpretation. The government expects significant savings from the RDTI program over the forward estimates, with most of the savings reflected in the 2029-30 year. Enacting the new measures without change is expected to materially reduce the quantum of future RDTI claims. It indicates a government more concerned with fiscal issues than addressing Australia’s faltering innovation environment.

Real world impacts 

A deeper analysis indicates that the 4.5 percentage points increase in R&D tax offset rates applicable to core R&D expenditures will rarely, if ever, compensate for the elimination of supporting R&D expenditure. See Appendix II for a numerical example.

Supporting activity types can represent between 40% and 70% of total R&D efforts in software, engineering, biotech, medtech and advanced manufacturing. Their wholesale removal could materially reduce the quantum of eligible expenditure and deny refund or offset amounts for many companies, even those whose R&D is only moderately made up of supporting activities. That said, the binary distinction between core and supporting activities will now become crucial. 

From an international perspective, the proposal also departs from the Frascati Manual’s treatment of R&D inputs, which explicitly recognises supporting, ancillary and indirect activities as part of R&D when they are undertaken solely for the performance of R&D. As such, many OECD jurisdictions (e.g., Canada, UK, Singapore) recognise ancillary or supporting R&D costs, although treatment varies across incentive regimes, meaning Australia would become an outlier. 

Given the global mobility of capital, a move to a strict core-only model is likely to further reduce Australia’s attractiveness for global R&D investment and multinational project allocation. Existing investors may seek to offshore new multi-disciplinary or capital‑intensive R&D functions. The proposal also risks substantive misalignment with global R&D measurement standards. 

 Other real-world impacts may include: 

• Entangled project activities. 
  Wider R&D projects comprise an entangled mix of both experimental and non‑experimental tasks – data preparation, prototype fabrication, integration, test environment setup, validation work. Requiring a potentially complex binary distinction ignores how innovation occurs in practice and may penalise complex or regulated industries.
• Infrastructure-heavy projects. 
  Depending on the ultimate interpretation applied by the regulatory bodies, projects requiring substantial infrastructure, integration or iterative prototyping will become less viable if only narrow experimental steps are eligible. This may discourage investment in hardware‑heavy, regulated or deep‑tech sectors.
• Start-up funding gaps. 
  Start-ups also often rely on refunds derived from supporting R&D activities such as engineering time, testing and data work. If interpreted narrowly, removing this eligibility may reduce refund amounts at the exact valley of death stage when capital is scarce. Such material falls in financial support across Australian business could result in lower R&D intensity and re-prioritisation of projects toward short-cycle or software-only R&D. 

Heightened focus on R&D categorisation 

The distinction between core and supporting R&D activities has always been contentious but has been relatively unimportant while the financial impact was neutral. Moving to a single eligible core category will heighten uncertainty around accurately classifying activities and will merely refocus ATO disputes to whether an activity is “directly related” to the core experimentation.

Although the subject of some satire online, all activities previously framed as supporting will in practice be rescrutinised for whether they directly contribute to the experiment, and there will be an increased need for granular documentation to justify the boundaries of core and supporting classifications.

The following are examples of expenditures that are likely to have been classed as supporting but could plausibly be considered as directly producing or directly contributing to the experimental hypothesis testing work:

• Experimental test environment setups, such as configuring test rigs, preparing datasets, calibrating instruments, or building controlled test conditions.
• Prototype build components including costs of constructing prototypes used directly in hypothesis testing (e.g., hardware assemblies, firmware integration, pilot plant or lab scale models).
• Algorithm training cycles such as data preparation and model training runs if they are integral to testing whether the experimental approach works.
• Integration work required to run the experiment such as software or hardware integration needed to execute the experimental step rather than for production deployment.
• Specialised tooling or instrumentation created solely to perform the experiment (e.g., custom jigs, test harnesses, measurement devices) where they are inseparable from the experimental activity.
• Regulatory or compliance testing tied to the hypothesis where the regulatory test is the only way to validate the experimental outcome (common in medtech, biotech, aerospace).

Proposed 10-year refundability limitation 

A second key proposal is to restrict the refundable component of the higher-rate SME R&D tax offset (with a new threshold of $50m) to companies that also meet an age test, those incorporated for less than 10 years.

Companies older than 10 years will continue to access the quantum of the higher refundable offset rate but, where excess, would not be able to access a cash refund. This effectively creates a second category of non-refundable R&D offsets.

The potentially transformative development of raising the threshold to match the base rate entity threshold is dramatically tempered by limiting refundability to companies under 10 years old. In many companies undertaking R&D, whether a start-up or mature company, a cash refund may be the difference between continuing or abandoning their R&D. 

As already apparent in business social media feeds, the cliff-edge 10-year age limit will cut off cash support as many companies begin to scale. Many high impact industries, including biotech, medtech, clean tech, and advanced manufacturing, routinely require 12–20 years to reach commercialisation. These firms may still be pre profit at year 11, meaning that the policy would remove cash-flow support before commercial viability is achieved, and may unintentionally penalise the very companies it aims to support.

In practice, refundability often funds ongoing payroll, trials, prototyping, and regulatory work. The loss of refunds at year 11 could create project deferrals, force capital raising earlier than planned or reduce R&D intensity. This cliff effect will be particularly acute for companies with multi year experimental programs.

The reform also creates further structural inequity across industries. Sectors with short development cycles (e.g., SaaS) will be largely unaffected, while long cycle sectors may be disproportionately penalised. This may distort investment patterns away from areas with the highest national spillovers, by reducing investor appetite for deep tech, increasing the cost of capital and pushing founders to jurisdictions with more predictable long term innovation incentives. These concerns are amplified when combined with the unpopular CGT reforms. 

Defining company age will also give rise to administrative and business planning challenges for founders, CFOs, and investors. Integrity measures can be anticipated to prevent contrived scenarios designed to retain refundability. 
The key questions will hinge on how ‘age’ is determined for groups, restructures, mergers, demergers or spin outs and whether reincorporation will reset the clock.

Policy considerations that stakeholders may raise during consultation include extending the age threshold (e.g., to between 12 and 15 years), introducing sector specific exceptions for long-cycle industries or implementing a graduated phase-out of refundability rather than a hard cut-off.

Other corporate tax interactions

From a wider corporate tax perspective, there are several interactions with the RDTI that are worth understanding. These include:

• the proposed refund of tax losses for start-ups in their first two years of operation 
• the now-permanent tax loss carry back regime for two income years
• the now-permanent $20,000 instant asset write off in Subdiv 328-D.

Interaction with new start-up loss measures 

Under the newly announced measure, for income years starting on or after 1 July 2028, new companies with an aggregated turnover of less than $10m will be able to claim a refund of tax losses in their first two years of operation, capped at the amount of PAYG withholding and FBT payments made in the income year the loss is generated.

This measure will clearly benefit companies that employ staff in their first two years of operation, in contrast to, for example, life sciences companies that contract CROs to conduct clinical trials. 

Any interaction with the RDTI will depend on how the company’s capped eligible tax loss is calculated and how much of that loss might instead be absorbed by the RDTI claim (if any). A detailed worked example is set out in Appendix I. 

Notably, the RDTI regime currently takes structural precedence over other deductions and tax offsets under s 355-715 where there is an entitlement to the R&D tax offset. This entitlement relies on the expenditure being incurred on validly registered R&D activities. 

If the legislation for the new start-up offset does not explicitly remove or reverse this ordering, the precedence of RDTI claims may constrain the amounts that may be refunded via the start-up loss refundable tax offset route. For example, where most R&D expenditure is made up of salary costs, and the company has minimal other costs to contribute to tax losses, the need for the tax losses to absorb the Label D R&D addback will reduce the use of the start-up loss route.  

Simply stated, where prior-year and current-year tax losses available are less than the combined total of the start-up loss cap and notional R&D deductions, the start-up loss usage must be constrained or capped to preserve sufficient tax losses to absorb the Label D addback.

However, as the example in Appendix I illustrates, the same total cash refund will be received regardless. As such, there is no need to restrict an RDTI claim to access the start-up loss route. The key differential will be whether, once enacted, the start-up loss refundable offset also gives rise to deferred franking debits (as is the case with a refundable R&D tax offset). 

It is to be hoped that the draft legislation will explicitly address these potential interactions and ensure identical franking outcomes for both refundable offsets. If so, this will necessarily reduce the attactiveness of the new measure as it will reduce the future franking credits that can accumulate in affected start-ups. 

Tax loss carry back regime 

The permanent reintroduction of a tax loss carry back regime for entities with aggregated turnover of less than $1bn from income years beginning on or after 1 July 2026 once again treats tax losses more favourably than excess non-refundable R&D tax offsets for larger R&D entities. 

As with previous incarnations of the carry back measure, refunds are capped by the franking account balance and can only be carried back for up to two prior income years. In a similar way to the start-up loss refunds, the need for tax losses to absorb the R&D Label D addback must also be considered. Where non-refundable R&D tax offsets become excess because of the use of tax losses, an R&D uplift has been gained but the resulting tax attributes can only be carried forwards, not carried back and cashed out.  

The cash flow benefit of carrying back the tax losses may exceed the future benefit of the R&D uplift. This may lead corporate taxpayers to forego the R&D tax uplift in exchange for the immediate cashing out of tax losses, effectively disincentivising RDTI claims in certain circumstances.  

Interaction of RDTI and Subdiv 328-D IAWO 

The permanent increase in the subdivision 328-D small business entity instant asset write off (IAWO) to $20,000 will intensify the consequences of its interaction with the RDTI regime. 

This legislative interaction is complex and little understood. Where a tangible depreciating asset is immediately used for R&D purposes, the subdivision 328-D IAWO is not available. Instead, the usual Division 40 decline in value must be included in eligible R&D expenditure, with future declines in value dependent on the use of the asset. 

In contrast, where a tangible asset is immediately used for a taxable purpose, and later for an R&D purpose, no R&D entitlement exists and the subdivision 328-D IAWO is available. Care should be taken to ensure that IAWO amounts are not included in R&D claims and that R&D assets are properly reflected in tax fixed asset registers. A numerical example is set out in Appendix II.

Conclusions 

These reforms will undoubtedly materially reduce RDTI benefits. Combined with increased ATO funding, all RDTI claimants will continue to face relentless scrutiny from the ATO, regardless of any statutory changes enacted. This compliance burden may of itself offset some of the positive economic impacts anticipated from these announced Budget changes. 

The interactions between the RDTI regime and mainstream corporate tax provisions are also becoming increasingly convoluted. Corporate tax and R&D advisers will need to understand these issues thoroughly to advise on the consequences of the differing company income tax regimes.

As we transition to the new RDTI rules, detailed guidance from the regulatory bodies will be necessary to understand the scope of expenditures that will be accepted as incurred on the core R&D activities when previously framed as supporting, for example, experimental test set-up, integration work, specialised tooling, regulatory testing.

Finally, should the changes proceed, it would be tempting to re-categorise the future available RDTI offsets as ‘SME refundable/non-refundable offsets’, and ‘large business non-refundable offsets’. 

FOR MORE INFORMATION

For more information about the Research and Development Tax Incentive (RDTI) regime please get in touch below.