Submission to the Treasury consultation on the Better Targeting the Research and Development Tax Incentive exposure draft (c2026-792804)
Introduction
Intellect Labs is an Australian specialist advisory firm working exclusively in the Research and Development Tax Incentive and other state and Federal grant programs. Our practice has prepared and defended R&D tax claims across biotechnology, agtech, aquaculture, software, clean technologies, critical minerals and manufacturing, and we work directly with entities at each of the lifecycle stages the exposure draft is designed to affect.
We support the government's objective of directing the refundable offset toward high-potential firms with genuine cashflow need, and of simplifying the activity eligibility framework. This submission does not oppose either objective. Rather, we seek to:
raise implementation issues with the current exposure draft that could produce outcomes inconsistent with the stated policy intent, and
propose refinements consistent with the exposure draft's existing architecture and the Government’s fiscal cost containment objective.
Summary of our recommendations
In summary, we recommend the Government:
Replace the refundable offset age-gate cutoff with a cumulative, non-concurrent entitlement model, so refundable access tracks genuine periods of R&D investment rather than elapsed time since enterprise commencement.
If a cumulative entitlement model is not adopted, at minimum define the age-gate's start day by reference to first R&D registration alone, removing the “earlier of” enterprise-commencement limb.
Confine connected entity/affiliate tests to entities carrying on a similar business to the claimant (the similar business test), so a genuinely new business is not penalised for an informal connection to an unrelated enterprise.
If that is not adopted, at minimum confine connected entity/affiliate tests to entities with their own R&D Tax Incentive claim history.
Adopt a clearer test for when work forms part of an R&D activity and when it does not. A necessity test (whether the core activity could be completed without it) and/or a purpose test (whether it has any significant purpose other than to enable the core activity) would be appropriate.
Confirm in the legislation or explanatory materials that an eligible R&D activity is not excluded by subsection 355-25(2A) merely because it also relates to, or incidentally supports, a different eligible activity.
Overview of proposed changes
The exposure draft makes seven principal changes. We have chosen to comment substantively on the two that raise the most significant implementation concerns. We do not raise concerns with the remaining five, for the reasons noted against each.
| # | Change | Our position |
|---|---|---|
| 1 | Minimum expenditure threshold raised from $20,000 to $50,000 | It is reasonable to index the minimum spend required to access the program at the same time as the turnover ceiling for refundable access is raised. |
| 2 | Maximum expenditure eligible for the premium offset rate raised from $150 million to $200 million | This is consistent with the stated objective of backing large-scale R&D and does not alter the eligibility framework in a way that creates uncertainty. |
| 3 | Turnover threshold for refundable offset eligibility raised from $20 million to $50 million | We support this change. It extends cash-refundable access to a larger population of early-revenue R&D-intensive companies, consistent with the policy objective. |
| 4 | All offset rates increased by 4.5 percentage points |
We support this change in principle but note this increase should not be treated nor characterised as compensating for the loss of eligibility of supporting R&D activities.
A higher rate applied to a narrower base of eligible expenditure is not a net benefit for claimants. |
| 5 | Intensity premium threshold lowered from 2% to 1.5% of total expenses | We support this change. This makes the premium rate accessible to a broader range of claimants. |
| 6 |
Refundable offset age-gate limiting eligibility to:
|
We raise a concern with the design of this measure. See section 1 below. |
| 7 | Removal of the supporting R&D activities category | We raise a concern with the implementation of this measure. See section 2 below. |
Implementation Challenges
1. The refundable offset age-gate should measure R&D engagement, not enterprise age
The problem
New table item 1A ties refundable offset eligibility to the tenth anniversary of the earlier of (a) the day an entity first started carrying on an enterprise, or (b) the day it first registered for R&D activities. In practice, (a) will predate (b) for most entities, meaning the ten-year clock most commonly runs from general enterprise commencement rather than from any point connected to R&D activity.
This creates a mismatch between the eligibility trigger and the policy rationale. Two entities with identical R&D investment histories can receive materially different treatment depending on how long they operated before beginning R&D:
An entity incorporated in year one that begins R&D in year eight has only two years of refundable access remaining, despite being at the earliest stage of its R&D program.
An entity that begins R&D activity concurrently with incorporation receives the full ten years, even where its underlying commercial operations are otherwise mature.
The design also compounds unfavourably with the $50 million turnover threshold. A young entity that becomes ineligible on turnover grounds during a strong revenue year, then falls back under the threshold once its enterprise-age clock has already expired, permanently loses refundable access despite never having had a genuine period of continuous eligibility.
Our recommendation
We recommend the government consider a cumulative, non-concurrent entitlement model in place of a fixed calendar cutoff.
Under such a model, entities draw down a bank of refundable-eligible income years (calibrated to the same 10 or 15 year total) at their own discretion across their lifecycle, rather than forfeiting access purely based on elapsed time since enterprise commencement.
This preserves the fiscal cost containment objective, since the total number of refundable years available to any entity is unchanged, while ensuring the measure targets periods of actual R&D investment rather than an enterprise-age proxy that frequently diverges from it.
A discretionary drawdown model also better serves the program's underlying integrity objective. Under a fixed calendar cutoff, an entity has an incentive to lodge a claim in every eligible year regardless of the strength or maturity of that year's R&D, simply because the clock is running down irrespective of use. A limited bank of drawable years instead gives entities a direct incentive to reserve their refundable access for the years in which R&D investment and cashflow need are genuine, deferring or foregoing a claim in years where it is not, since the years not drawn down remain available later in the entity's lifecycle. This is a more targeted outcome than the current design, which rewards claiming for the sake of the clock rather than for the sake of the underlying R&D.
A narrower alternative
Should a discretionary drawdown model not be adopted, we recommend at minimum that the start day be defined by reference to first R&D registration alone, removing the "earlier of" enterprise-commencement limb. This is a narrower change that would still go some way to addressing the mismatch described above.
Connected entity overreach
The potential inconsistency with policy intent is compounded by the age-gate's connected entity mechanics. New subsection 355-100(6) extends the age-gate to entities connected with, or affiliated with, another entity that has itself been carrying on an enterprise for more than the relevant period, regardless of whether that connected entity has ever claimed the R&D Tax Incentive. An entity's own refundable access can therefore turn on the unrelated commercial history of another business it shares an investor or controlling individual with, rather than on anything connected to its own R&D activity.
This is the same design flaw described above. Just as the age-gate measures enterprise age rather than R&D engagement for the claimant itself, the connected entity rules import an unrelated enterprise's age into that assessment, without any inquiry into whether that connected entity has ever engaged in R&D at all, or carries on any business related to the claimant's.
Our recommendation
We recommend the connected entity and affiliate tests for this measure be confined to entities carrying on a similar business to the claimant, using the similar business test already applied elsewhere in the tax law to assess business continuity (Division 165 of the Income Tax Assessment Act 1997). This directly targets the risk where an entity resets its own age-gate clock by continuing the same business through a new corporate vehicle, while ensuring that a genuinely new business is not penalised merely because it shares an investor, director or other connection with an unrelated enterprise.
Should this not be adopted, we recommend at minimum that the connected entity and affiliate tests be confined to entities that have themselves claimed the R&D Tax Incentive.
2. Removing supporting R&D activities requires clarity on the core activity boundary
The problem
The exposure draft renames core R&D activities to R&D activities and removes the supporting activities category. New subsection 355-25(2A) provides that an activity is not an R&D activity merely because it is directly related to, or undertaken for the dominant purpose of supporting, another eligible activity.
This change does not resolve the harder question that the removal of supporting activities makes materially more consequential. That is, where the boundary of a registered core activity sits.
Under current law, work that is directly related to another core R&D activity can be registered and claimed as a supporting activity (subject to certain exclusions, which are assessed against a higher dominant purpose threshold). Because that pathway exists, it is often unnecessary to ask whether the same work would also qualify as a core R&D activity, or as part of the core activity it relates to. From 1 July 2028, it cannot be claimed unless it is properly characterised as part of the core activity.
As a result, the exposure draft increases the legal significance of this boundary question while providing no corresponding clarity on how to distinguish work that is properly part of a core R&D activity. In sectors where manufacture, scale-up, validation and data-collection activities have historically used the supporting activities gateway (including most notably manufacturing process development and certain clinical and field-trial support functions), entities and their advisors must now reassess on which side of the line these activities will fall. This will inevitably lead to inconsistent, activity-by-activity outcomes settled only through ATO review, objection and, in many cases, litigation before the Administrative Review Tribunal or courts, years after the event.
Our recommendation
We recommend that, rather than leaving this boundary to be settled activity-by-activity through individual findings or litigation, the Government adopt a clearer statutory test for when work forms part of a registered core activity.
In our experience advising entities across this boundary, a test that travels best would ask two things:
whether the core R&D activity could be completed without the work in question, and
whether the work has any significant purpose other than to enable the core R&D activity (as distinct from some other purpose, such as commercial production or supply, running alongside it).
An activity satisfying both limbs should be treated as forming part of the registered core R&D activity, and not excluded by subsection 355-25(2A). This ties eligibility directly to necessity and purpose.
How the test applies
This test travels well across the sectors most exposed by the removal of supporting activities, while still tightening the definition to achieve fiscal savings. To use the clinical manufacturing example, testing a hypothesis experimentally, in a clinical trial context, requires a dose to test. Where a trial cannot proceed without doses manufactured for that purpose, and that manufacture has no purpose other than to enable the trial, both limbs of the test are met.
Conversely, the test correctly excludes activity that is genuinely commercial in character. Work conducted under otherwise-standard commercial conditions, with data collection added on top, would not meet either limb even though it might satisfy the current looser “directly related” standard.
Guidance on incidental support
We further recommend the government confirm, either in the Bill or Explanatory Memorandum, that an activity is not excluded by subsection 355-25(2A) merely because it also happens to relate to, or incidentally support, a different eligible activity. This preserves the intent of subsection 355-25(2A) as an anti-avoidance style clarifier, while avoiding an unintended narrowing of what constitutes a core activity's proper boundary.
Conclusion
We welcome the opportunity to discuss either of these issues further, including by reference to specific client scenarios that illustrate the practical effect of the current drafting. We are available to Treasury and DISR officials for further consultation ahead of introduction of the Bill.
