Submission to the Treasury: Better targeting the Research and Development Tax Incentive - Exposure Draft Legislation

Table item 1A only applies where the R&D entity’s aggregated turnover is less than $50 million. Because the aggregated turnover test already consolidates the turnover of connected and affiliated entities, an entity to which item 1A can apply is, by definition, already a member of a small group. We acknowledge that the use of the “connected with” and affiliate test for these purposes is to prevent related parties potentially overcoming the 10-year or 15-year rule, however as noted below there are several critical, unintended consequences. We have provided recommendations which may better serve the intended policy objectives.

The most concerning aspect of the Exposure Draft and which needs urgent attention is that subsection 355-100(6) extends the start-day test in item 1A to any entity connected with the R&D entity, or of which the R&D entity is an affiliate, or where the other entity or entities is an affiliate of the R&D entity. This is coupled with the fact that the start-day test is the earlier of when the R&D entity (or connected/affiliate entity) first started carrying on an enterprise.

This imports the commercial history of investors, founders and other group members (none of whom need ever have conducted, funded, or held any interest in R&D activity, and in many cases would not even know the R&D entity) into the determination of a single R&D entity’s access to the refundable offset, without any corresponding integrity benefit.

The following examples (not exhaustive) illustrate how the proposed legislation might practically apply:

• Person A, whilst at University, carries on an enterprise of selling eBay products. 15 years later Person A is a founder at an AI start-up business. The AI start-up business would be precluded as Person A, as a connected with entity, first carried on an enterprise 15 years earlier.
• Investor A (either an individual or a Venture Capital firm), who holds 45% voting rights in Company A, invests in a new AI start-up business (company B). Company A has been operating for 10 years and currently has annual turnover of $10m. Company A had claimed R&D for 3 of those years. The operation of the “connected with” test which treat Company A and Company B as connected. Company B would never be able to access the refundable R&D tax offset.
• Person A is an affiliate of his wife (Wife A) as he respects her business acumen and will listen to her directions and wishes in relation to his own business. Wife A has operated a florist business for 20 years (non-R&D). Person A founds Company A. The operation of the “connected with” test, being affiliate inclusive, will connect the florist business with Company A. Any business which Person A is connected to will never be able to access the refundable R&D tax offset.

We would also observe that this mechanism works against other measures announced on the same day. The expanded venture capital tax incentives and the Investment Boost/IBCC are both directed at attracting institutional capital into early-stage, high-growth companies. Because subsection 355-100(6) reaches any entity connected with, or an affiliate of, the R&D entity, the arrival of an institutional investor (who becomes “connected with” the R&D entity through its equity stake) can itself import that investor’s own enterprise history into the start-day calculation, potentially extinguishing the R&D entity’s access to the refundable offset at precisely the point in the company’s life that the Government’s other measures are designed to encourage institutional capital to arrive.

Finally, we note that having previously carried on an enterprise is, if anything, a positive indicator of the likelihood of successful commercialisation, rather than a marker properly associated with reduced eligibility for incentives. A mechanism that reduces refundable-offset access by reference to a founder’s or investor’s prior, unrelated commercial experience does not appear calibrated to the stated objective of directing support to high potential young companies.

Independently of the subsection 355-100(6) issue, we consider “carrying on an enterprise” in paragraph (a) of item 1A to perhaps not be the best statutory test for fixing a start date for accessing refundable R&D tax offsets. 
Reasons are as follows:

• The concept extends to the commencement of an enterprise, meaning preparatory or commencement activity is sufficient to fix the start day. The start day can therefore precede the existence of any actual business, and indeed the incorporation of the R&D entity itself.
• Because commencement activities count, the start day is not a single ascertainable date fixed by an objective record (such as an ASIC registration), rather it is a question of fact as to when preparatory steps commenced. Arguably a matter inherently more difficult, and more contestable, for both taxpayers and the Commissioner to determine with confidence than a registrable event.
• The concept has no Australian nexus. Offshore activity conducted by an offshore connected or affiliated entity can fix the start day, notwithstanding that the activity has no connection to Australia or to the R&D entity’s own operations.

If the policy intent is that refundable-offset access should be confined to young companies, we consider the date of the R&D entity’s incorporation to be a substantially more direct and readily ascertainable proxy. We note there is already precedent for using date of incorporation as a bright-line test of this kind elsewhere in ITAA 1997, see subsection 83A-33(3).

Subsection 355-100(6) provides for connection to be tested only as at the relevant time, not by reference to any point since the R&D entity's incorporation. This produces two related but distinct problems.

First, the provision is over-inclusive: an entity can be disqualified by the unrelated commercial history of a founder or investor who happens to be connected with it, with no bearing on the R&D entity's own age or genuine circumstances. If a founder incorporated a company fifteen years ago and has long since ceased trading, the eligibility of a newly incorporated R&D entity for the refundable offset would, on the current drafting, appear to depend on whether that founder has taken the administrative step of deregistering the earlier company with ASIC (for example, by lodging a Form 6010).

Second, the provision is under-inclusive against the restructuring it is presumably designed to prevent. Because connection is tested only at the relevant time, an R&D entity's refundability can change from year to year based solely on whether a disqualifying connection happens to still exist, without any change in the entity's own activities. An entity connected to an entity past its own tenth anniversary could be refundable one year, lose that status the next while still connected, and regain it in a subsequent year simply by severing the connection before the relevant income year, the opposite of the considered, one-off youth test the policy appears to intend. We do not consider this an appropriate basis on which the availability of the refundable tax offset should turn.

Paragraph 1.15 and Table 1.1 of the explanatory materials (EM) describe the test as being satisfied where a day in the income year falls before, or on, the tenth anniversary of the start day. Item 1A as drafted instead disqualifies the entity where any day in the income year falls on or after that anniversary. These two formulations produce different results in the anniversary year itself. The worked example at paragraph 1.33 of the EM follows the drafted legislation rather than the test as described in the EM's own narrative, which in our view confirms the inconsistency lies in the drafting of the EM rather than in the legislation, however, this should be corrected before introduction, given the EM will be relied upon as an aid to interpretation.

Separately, item 1A is not expressed as being subject to item 2 of the same table. As drafted, an R&D entity with aggregated turnover under $50 million that is past its tenth anniversary, and that also falls within item 2, does not appear to be tie-broken between the two items.

The start date of the R&D entity, whether for “carrying on a business” or “enterprise” should not impact when the
R&D activities should receive the incentive. Instead, we would recommend:

• Utilising an Export Market Development Grant style limit on the number of refundable R&D tax claims, rather than number of years; or
• In line with the SERD Report, provide for a preliminary number of years for access to the refundable R&D tax offset, followed by the ability to apply for ongoing eligibility depending on KPIs such as average annual revenue growth rate (as stated in the SERD Report) or other measurable metrics.

Address entity churning through a targeted successor rule (suggested drafting below), rather than through the broad “connected with” entity / affiliate mechanism in subsection 355-100(6).

Suggested drafting for consideration:

Replace subsection 355-100(6) with the following successor rule:

‘(6) For the purposes of table item 1A in subsection (1), if: (a) another entity (the predecessor) carried on activities; and (b) all or a substantial part of the assets, employees or research and development activities of the predecessor were transferred (directly or indirectly) to the R&D entity; and (c) at the time of the transfer, the predecessor was connected with the R&D entity or was an affiliate of the R&D entity, or the R&D entity was an affiliate of the predecessor; the start day of the R&D entity is taken to be the earlier of the R&D entity’s start day and the day on which the predecessor was incorporated or otherwise came into existence.’

Clarify the operation of paragraph (b) of item 1A and, if retained, replace the reference to the day of first registration with a reference to the first income year for which the entity is or was registered under section 27A of the Industry Research and Development Act 1986 (Cth) (IRDA).

Reconcile the explanatory materials with the legislation on the treatment of the anniversary year, and test the anniversary at a single point in time rather than by reference to ‘any day during the income year’.

Express item 1A to apply only if item 2 of the table does not apply.

The exposure draft removes the existing category of supporting R&D activities so that only activities satisfying the core R&D requirements will remain eligible. The announced 4.5 percentage point increase in the offset rate for core R&D expenditure does not resolve the threshold issue: expenditure on an activity that is necessary to enable, conduct or evaluate an experiment may become wholly ineligible merely because that activity does not, viewed in isolation, satisfy the core activity definition.

In practice, eligible experimentation is rarely undertaken as an isolated laboratory or technical step. It commonly relies on preparatory work, experimental design, the creation and maintenance of test environments, specialist data generation and collection, prototype or test article preparation, trial coordination, quality controls, safety monitoring, analysis and validation. These activities may be inseparable from the systematic progression of work even where they are not themselves the step through which new knowledge is directly generated.

A complete removal may therefore create a false distinction between the experiment and the work necessary to make that experiment technically valid, safe and capable of producing interpretable results. It is also likely to shift, rather than remove, compliance complexity. Claimants and regulators will need to determine whether personnel, consumables, facilities, equipment, contractors and other costs are incurred “on” the core activity or relate to a separate ineligible activity. This creates new boundary and apportionment disputes and may encourage inconsistent characterisation of the same real-world work.

The economic effect of the reform will depend on the proportion of a claimant’s existing expenditure that remains attributable to core R&D activities. The higher rate may compensate claimants whose supporting activity expenditure is small, but it may materially reduce the effective benefit for claimants whose experimental programs require substantial enabling work. The reform should therefore be assessed by reference to the effective incentive across representative project profiles, not solely by reference to the headline increase in the core offset rate.

A reduction in the effective incentive is particularly significant where the excluded work is not ordinary commercialisation or routine administration, but expenditure without which the eligible experiment could not be properly designed, conducted, monitored or evaluated. In those circumstances, removal may reduce additionality and influence whether, when or where the wider R&D program proceeds.

The proposed change will not affect all industries equally. Sectors characterised by long development cycles, complex testing environments, significant regulatory requirements or integrated development pathways are likely to be disproportionately affected. This includes biotechnology, pharmaceuticals, medical devices, advanced manufacturing, defence, clean technology, mining technology and agritech (non-exhaustive). In these industries, experimentation commonly relies on a broader ecosystem of enabling activities that may not themselves satisfy the core activity definition but are nevertheless necessary to design, conduct, monitor, validate and interpret experimental outcomes. 

These activities are often not peripheral to experimentation. They may be necessary to protect participants, maintain the integrity of the test article and experimental environment, satisfy scientific and regulatory standards, and generate data capable of supporting valid conclusions. Treating all such activities as ineligible may disproportionately reduce support for biotechnology, pharmaceuticals, medical devices, defence and advanced manufacturing, notwithstanding these are stated areas of national priority.

The exposure draft also recognises that qualifying therapeutic-goods innovators may require a longer refundable period by providing a potential extension from ten to fifteen years. That sector-specific recognition reflects the longer development and regulatory pathways faced by these innovators.

More broadly, it demonstrates an acknowledgement that certain activities and industries may require tailored treatment where innovation occurs through long, complex and interdependent development pathways. In our view, the same policy rationale supports calibrated treatment of supporting expenditure that is necessary and integral to those pathways.

RSM recognises that retaining the supporting activity rules entirely unchanged may not align with the Government’s announced policy. We nevertheless consider that complete removal is unnecessarily blunt.

Treasury should adopt a proportionate model that excludes remote, routine and commercially oriented activity while preserving a pathway for expenditure and activities with a direct and demonstrable experimental nexus. We provide further details below on the potential impact of the current proposal. In terms of RSM Australia’s suggested solutions, our preferred suggestions (all easily implemented) are as follows:

Recommendation 1: Constrain the supporting activity limb by strengthening the dominant purpose test(similar to New Zealand)

Retain a supporting activity category, but require the activity to have the dominant purpose of supporting an
eligible core R&D activity (i.e. use the wording already in s355-30 and simply removing the “directly related”
test). This would target the expenditure identified by Treasury while avoiding the administrative complexity and
litigation risk associated with attempting to distinguish experimental and enabling activities solely through the
definition of core R&D.
 

Recommendation 2: Introduce a statutory exclusions list

Retain the concept of a supporting activity however, legislate a closed list of activities that can never qualify as
supporting activities (for example, routine administration, sales, marketing, commercialisation, market testing
and ordinary production). This approach is already adopted for core R&D Activities under s355-25 in the
existing legislation and works well.

Recommendation 3: Introduce a capped supporting expenditure ratio or deemed rate (as proposed in the SERD Report)

Permit supporting activity expenditure up to a prescribed percentage of core R&D expenditure, or allow a fixed
uplift on eligible core expenditure in lieu of specific supporting activity substantiation. If Treasury's concern is
excessive claims for supporting activities, a cap directly targets the quantum of expenditure rather than
removing eligibility entirely and will provide simplicity of administration.

We also recommend providing a higher deemed percentage, carve-out or prescribed-activity pathway for
qualifying therapeutic goods, biotechnology, medical technology and other priority-sector activities with
demonstrably long, regulated and interdependent development pathways.

Finally, we propose that any changes be applied only to R&D projects commenced after the commencement
date of the legislation, rather than to existing programs already underway. This would preserve commercial
certainty for investment decisions made under the current regime while still allowing Treasury to achieve its
policy objectives on a prospective basis.

At the outset, it is noted that complete removal of supporting R&D activities appears to be misunderstanding the outcome of the SERD Report, which recognised the importance of supporting R&D activities. The SERD Report proposed that the process of accessing the RDTI should be simplified for supporting R&D activities and greater support should be provided to certain high growth businesses or those with higher probability of return on investment. The removal of supporting R&D activities achieves none of these objectives.

The proposed amendments risk creating an artificial separation between activities that generate new knowledge and activities that are necessary to support, enable, evaluate and translate that knowledge into practical outcomes. Innovation rarely occurs in discrete stages. Rather, the generation of new knowledge, development of prototypes, manufacture of trial materials, validation activities and related technical work often form part of an integrated innovation process.

While supporting activities may not themselves constitute experimental activities, they frequently provide the technical foundation upon which experimental activities depend. Activities such as background research, prototype preparation, trial manufacture, validation testing and related technical work are often undertaken not because outcomes are known, but because they are necessary to enable, inform or evaluate the corresponding experimental activities.

This is particularly relevant in sectors such as biotechnology, medical technology, advanced manufacturing and defence, where experimental activities cannot readily be separated from validation, trial production, process optimisation and other activities required to determine whether an innovation can be successfully applied in practice. In many cases, these activities are undertaken because further uncertainty remains regarding whether the outcomes of the experimental activities can be successfully realised at scale.

The proposed exclusions therefore risk reducing the economic return generated from Government support for R&D by creating a sharper distinction between the generation of new knowledge and the activities necessary to apply that knowledge in practice. This concern is particularly relevant in light of broader Australian innovation policy objectives, which increasingly emphasise the importance of research translation, commercialisation, sovereign capability, productivity growth and the development of new industries. 

The SERD Report repeatedly highlights the importance of strengthening the connections between knowledge creation, innovation, scale-up and commercial outcomes, and notes that Australia's longstanding challenge is not simply generating research, but converting that research into economic and societal value. The purpose of the RDTI is not merely to encourage experimentation in isolation. Rather, it is intended to support innovation activities that generate broader economic benefits for Australia, including investment, skilled employment, capability development, sovereign capability and commercial outcomes. Where activities are integral to enabling, supporting or evaluating experimental activities, their exclusion may diminish the effectiveness of public investment in achieving these broader policy objectives.

Internationally, several innovation systems have recognised that innovation extends beyond activities satisfying a strict R&D definition and have adopted complementary mechanisms to support innovation, translation and technology deployment activities. While approaches differ between jurisdictions, the common objective is not merely to increase R&D expenditure, but to maximise the economic and societal value created from that investment. Australia should carefully consider whether the proposed amendments inadvertently reduce support for activities that are necessary to achieve those same outcomes.