Re: R&D Tax Incentive Review Report - Comment on Findings and Recommendations

Transcription

1 Ernst & Young 200 George Street Sydney NSW 2000 Australia GPO Box 2646 Sydney NSW 2001 Tel: Fax: ey.com/au The Manager Business R&D Department of Industry, Innovation and Science GPO Box 9839 CANBERRA ACT October 2016 Delivered via to: R&DTaxIncentiveReview@industry.gov.au To Whom It May Concern Re: R&D Tax Incentive Review Report - Comment on Findings and Recommendations Ernst & Young (EY) is pleased to make this submission in response to the R&D Tax Incentive Review Report (the Review Report) co-chaired by Bill Ferris, Alan Finkel and John Fraser. In the spirit of consultation, please find attached our comments and feedback on the findings and the proposed recommendations of the Review Report for your consideration. The findings and recommendations of the Review Report are rightfully seeking to ensure the sustainability and effectiveness of the R&D tax incentive program. We note that achieving these objectives involves a striking a balance between cost and ensuring the program continues to incentivise desired behaviour our primary concern is that recommendation 4 (introduction of an intensity threshold ) does not strike the appropriate balance. We suggest that a better approach is to recognise the savings that have already been made through the recent 1.5% reduction in offset rates, as well as to impose a cap on the refundable component of the program. We provide more detailed comments in our attachment to this letter. Thank you again for the opportunity to provide our comments. If you have any queries in relation to our submission, please contact me on (02) or Ezra Hefter on (08) Yours sincerely Jamie Munday Partner Enc.

2 R&D Tax Incentive Review Report Comment on Findings and Recommendations

3 Page 3 Comment on the R&D Tax Incentive Review Report Findings and Recommendations Recommendation 1. Retain the current definition of eligible activities and expenses under the law, but develop new guidance, including plain English summaries, case studies and public rulings, to give greater clarity to the scope of eligible activities and expenses. We support this recommendation: the current definition is based on the Frascati Manual definition which, in addition to being a principled (rather than prescriptive) definition, is an international standard which allows for comparison of the performance of the Australian program with international experience. Retaining the current definition of eligible activities provides certainty to taxpayers especially as the new R&D tax incentive is fairly new in its operation. It allows companies in all industries to bed down their understanding of what constitutes eligible R&D (according to the legislation) and to make compliant R&D tax claims, provided all the eligibility criteria are met. We also agree that claimants will benefit from new guidance and plain English summaries. Claimants will likely benefit from more targeted material targeted towards their specific sector and prepared in conjunction with technical experts in the field, to give a real-life sense of the activities that are claimable under the incentive program. A public rulings register comprising AusIndustry decisions on eligibility would also provide greater certainty and guidance to claimants provided the reasons for decision are clearly expressed and based on legislative grounds or by reference to technical merits (more on this below). Recommendation 2. Introduce a collaboration premium of up to 20 percent for the nonrefundable tax offset to provide additional support for the collaborative element of R&D expenditures undertaken with publicly-funded research organisations. The premium would also apply to the cost of employing new STEM PhD or equivalent graduates in their first three years of employment. If an R&D intensity threshold is introduced (see Recommendation 4), companies falling below the threshold should still be able to access both elements of the collaboration premium. We support this recommendation. We note that collaboration between publicly-funded research organisations (such as universities) and industry is comparatively low in Australia. While Australia generates world-class research, translation to commercialised products and services is low. As such, we support the proposition that where possible the R&D tax incentive could be improved to provide additional support for collaboration.

4 Page 4 However, we would highlight that there are potential administrative overheads in applying the premium to the cost of employing new STEM PhD or equivalent graduates in their first three years of employment. That is, large organisations may not track this information (i.e. maintain records on all of their employee s qualifications or length of service). Maintaining this additional information may represent an added administrative burden. In addition, not all STEM PhD or equivalent graduates go on to perform research and development work. Whilst the policy intent is well intended, Government should consider whether any new policy will result in additional, unintended compliance burden and what can be done to minimise these impacts. Additional Recommendations / Alternatives for consideration The collaboration premium could be extended to expenditure with Research Service Providers (RSPs). RSPs already exist as natural partners for undertaking R&D activities and a collaboration premium will further promote and encourage this behaviour by industry. The R&D Tax Incentive could be tailored to make collaborating with RSPs more attractive. In addition, collaboration premium expenditure should be exempt from the expenditure cap, currently at $100 million, to encourage greater collaboration with the research community this may encourage large companies to choose to conduct more R&D with publicly funded research agencies. The collaboration premium should also apply to SME companies accessing the refundable R&D offset. For example, there is no reason that SME s and start-ups employing STEM PhDs are less deserving of this support. Mirroring the premium across both the refundable and non-refundable aspects of the offset will also simplify reporting requirements for companies which switch between the refundable to nonrefundable offset (or vice versa). Finally, the intent of this recommendation is to promote greater industry collaboration with publicly funded research organisations by providing support under the R&D Tax Incentive. Problematically, however, the current definition excludes non-experimental research activity from consideration as Core R&D (including many areas of theoretical scientific and mathematical research). The current eligibility criteria means that much activity undertaken with universities does not attract incentivised treatment, as the research fails to satisfy the experimental requirement. As such, Government should consider assessing eligibility to the 20 percent collaboration premium by reference to the status of the recipient, or qualification of the graduate (rather than adopting the current definition of eligible R&D activity).

5 Page 5 Innovation Vouchers A complementary program that should be considered is the use of innovation vouchers, as has been used in some states / overseas jurisdictions. Innovation voucher schemes aim to promote collaboration between publicly-funded research organisations and industry, and are being increasingly used to support innovation activities of SMEs. Vouchers support SMEs in seeking research expertise to solve a specific innovation-related problem, for example: a technical question, a small scale prototype, laboratory verification or field testing. This improves collaboration by direct knowledge transfer as well as acting as a catalyst for longer, more comprehensive collaboration. 1 The issuing of innovation vouchers (which under some program designs, require the recipient firm to match the voucher value with private funds) overcomes major barriers to engagement between SMEs and knowledge agencies. That is, the voucher empowers the SME to approach knowledge providers with their innovation problems; and secondly, the voucher provides an incentive for the knowledge agency to work with the SME (rather than working only with larger firms, or with no industry engagement). 2 Please let us know if you would like further input on the potential design of an innovation voucher system. Recommendation 3. Introduce a cap in the order of $2M on the annual cash refund payable under the R&D Tax Incentive, with remaining offsets to be treated as a non-refundable tax offset carried forward for use against future taxable income We understand the premise of a cap is to ensure the future sustainability of the program but would suggest a cap of $4 million on the annual cash refund payable under the R&D Tax Incentive as more appropriate. The proposition of a cap The Review Report is concerned with sustainability of the R&D Tax Incentive, and notes that the cost of the refundable component of the program has been increasing since its introduction in 2011 (a trend that is forecast to continue). The trend of increasing program cost is, in the Review Panel s opinion, posing a threat to future program sustainability. 1 Innovation Policy Report, October 2012 Australian Government, Department of Industry Innovation, Science, Research and Tertiary Education 2 Innovation vouchers, The Organisation for Economic Co-operation and Development (OECD)

6 Page 6 As a preliminary point, we note that the true cost of the program is not reflected in the cash amount of the refundable R&D Tax offset, as claimant companies forgo future tax deductions (it is acknowledged that some companies will not become tax payable). Determining the true impact of the proposed changes (and whether the improvement to sustainability outweighs the distortionary effect of the change) requires modelling to be undertaken taking into account the future deductions forgone. In addition, when the report was released, the refundable offset rate reduction to 43.5% had not passed into law (which has therefore already reduced the cost of the program). That is, there has already been significant savings achieved in the forward estimates by the reduction in the offset rates available. This should be borne in mind when considering the suitability of measures to improve the overall program sustainability. Having noted that there have already been savings made through reduction in the offset rate, we recommend that the Government undertake further modelling (taking into account the future deductions forgone of the 1.5% rate cut). Depending on the modelling, if Government still feels that it is appropriate to impose an expenditure cap, we would suggest that a $4m expenditure cap be considered in the interest of maintaining the future sustainability of the program. What is the appropriate cap amount? Any cap is, by definition, an arbitrary limit; there is no reason in principle why R&D expenditure above a certain amount should not be incentivised. In addition, the limit is unlikely to be adjusted for inflation reducing the R&D benefit over time. Care should be had to ensure that the benefit is not eroded to the extent that the incentivising impact is lost. The Review Report suggests a refundable cap of $2 million on the annual cash refund payable under the R&D Tax Incentive (with the remainder carried forward as a non-refundable tax offset). Selecting $2 million as a cap has the potential to introduce an innovation support cliff just at the valley of death for innovative small businesses. That is, new innovative businesses that are incurring a high level of initial R&D costs while not earning significant income, could be harshly penalised by the introduction of a $2 million cap just at the time when the R&D tax support is needed most. We are of the opinion that introducing a $4 million cap on the annual cash refund is more appropriate than a $2 million cap due to the funding challenges experienced when small companies require greater staff and facilities to enable development and commercialisation of their innovations. SME s often comment on the struggle of obtaining funding to enable them to traverse the valley of death to achieve a scale that enables sustainability. Innovative small companies require support to enable them to transitioning from owner operators or employing relatively few staff to having the size necessary to support staff with dedicated R&D roles. By introducing a higher cap at $4 million, the R&D Tax Incentive will continue to allow innovative companies to achieve the scale necessary to create jobs during this

7 Page 7 difficult transition phase and allow companies to build a stronger foundation of R&D within their organisations. Additional Recommendations / Alternatives for consideration An additional recommendation to the introduction of a cap is to also provide a discretionary path for companies to request for R&D expenditure that exceeds the expenditure cap to be accessed at as a refundable (rather than non-refundable) offset. This will provide policy containment on costs, but allows genuine innovation intensive companies to request for the refundable tax credits for amounts exceeding the cap. In these instances, assessment could be made on the likelihood of companies to become taxpayable, knowledge spill-over, job creation, additionality and national interest of R&D to be undertaken. In addition, this pathway could be linked to the test of an innovative company for the purposes of the Early Stage Innovation Company (ESIC) tax incentives. Recommendation 4. Introduce an intensity threshold in the order of 1 to 2 percent for recipients of the non-refundable component of the R&D Tax Incentive, such that only R&D expenditure in excess of the threshold attracts a benefit. We do not support this recommendation as introducing an intensity threshold would significantly reduce the incentivising impact of the R&D Tax Incentive program due to increased complexity and uncertainty it introduces into the program. In addition, it will also likely disproportionately impact lower margin industries. Increased complexity and compliance burden The complexity associated with predicting a company s total spend, as well as R&D spend for an income year means that some companies may be practically unable to determine whether they are eligible for the R&D incentive (or what proportion of their expenditure may be eligible for incentivised treatment) prior to the end of the financial year. This uncertainty can arise due to the inherently volatile nature of markets and foreign exchange rates. The presence of so many moving variables can make predicting intensity unfeasible. The presence of this uncertainty reduces the incentivising effect of the program, as companies on the threshold will be unsure if they are able to claim R&D expenditure at the time when they are making R&D investment decisions. Alternatively, it will provide a windfall gain with no incentivising effect for companies on the threshold which end up exceeding the threshold in a given income year.

8 Page 8 Implementing this change will also impose an additional compliance burden on companies (increasing program cost and reducing the incentive to participate in the program). In addition, imposing this requirement may increase business planning complexity companies will have to make a technical determination in advance on what might or might not qualify (a technical function), as well as make an assessment of likely expenditure over an entire organisation (a finance function). This is an undesirable distortion as it increases the administrative tasks companies have to undertake, taking away resources available to core business functions. Sectorial bias As acknowledged by the report itself, an intensity test will lead to a bias against certain industries / sectors which feature lower levels of intensity (for example, because they are a lower margin industry) there is no principled reason for this and this may have a distorting effect. This would mean for example - some innovative companies may be less able to access the R&D incentive compared to other companies simply by reason of the industries in which they happen to exist. This sectorial bias will have a real effect on reducing innovation across a number of sectors where margins are low. For example, it is difficult for many companies in the manufacturing sector to spend more than 1-2% of their expenditure on R&D. This will be catastrophic for many manufacturers who are seeking to invest in innovation but will receive little to no support from the R&D Tax Incentive. This goes against the innovation policy platform and contrary to the words of Greg Hunt that innovation is not about setting up a binary struggle between what some people call the old economy versus the new economy. 3 This test in effect disproportionally act against the companies in the old economy such as manufacturing. Distortion of company behaviour For very large companies, 1-2% of total business expenses requires a significant investment of expenditure which may not be feasible or practical in the current economic environment. Such a threshold may discourage innovation that these companies have currently planned to undertake. It may also cause a distortion of R&D behaviour; companies may delay projects in order to meet the intensity threshold in a particular income year. As companies may not be certain regarding their eligibility at the beginning of an income year, employees engaged in R&D may choose not to track their R&D activity as this may be viewed as an 3 Greg Hunt, Speech At The Public Sector Innovation Awards, Wednesday, 27 July 2016 accessible from: Sector-Innovation-Awards.aspx

9 Page 9 additional compliance burden which may ultimately not be rewarded. This represents a shift away from desirable behaviour, and may result in poorer R&D planning and documentation of results. There also needs to be consideration and analysis of the potential impact of this change on innovative medium sized businesses which may become unable to access the R&D Tax Incentive as they go through a high growth / high expenditure phase. Funding is a key factor to enabling mid-market innovation and incentivising the R&D activities of these businesses is desirable. This is because innovative mid-market companies create competitive pressures within their industries, and can trigger a cycle of competitive innovations between rival companies. Implementing an intensity test can have detrimental impacts by constraining innovation investment, resulting in stagnation rather than growth. Rate arbitrary and subject to change There is artificiality in choosing a rate between 1-2%. The lack of a firm reason to choose a figure may make the intensity threshold subject to change in the future. The introduction of an intensity measure provides a lever to improve the budget bottom line a temptation future Governments may not be able to resist. This is problematic as the intensity measure may be adjusted without reference to guiding principle. Uncertainty with definition of intensity The effect of the intensity test cannot be properly considered without considering how intensity is to be measured. The Review Report considers three possible measures: R&D spend / total business expenses; R&D spend / turnover; or R&D employees / total employees. There are uncertainties associated with each suggested measure of intensity. For example, if intensity is measured as a proportion of total business expenses should capital expenses be included? Under the current rules, companies are unable to include the costs associated with constructing tangible depreciating asset as R&D (instead, these companies are able to claim an uplift on the depreciation when these assets are used for an R&D purpose). This may have perverse outcomes that companies investing in future R&D capabilities become unable to access the incentive for current R&D activities. Alternatively, if intensity is measured as R&D expenditure as a proportion of turnover should once-off sales be included? This would mean that once-off sales, such as a large asset disposal, could impact a company s ability to claim the R&D tax offset. Further, is it aggregated or gross turnover? What about tax

10 Page 10 consolidated entities? For foreign multinationals, is it measured on a global, national or individual entity basis? If intensity is a proportion of total staff engaged in R&D how will numbers of staff engaged in R&D be measured and validated? How will contractors be treated? Further, the complexity associated with an intensity measure means that the test is susceptible to companies re-structuring their R&D affairs to take advantage of the R&D based on the definition used. As policy, an intensity test on R&D based on arbitrary thresholds is an undesirable program design, which adds complexity and cost, and can cause behavioural distortion. Recommendation 5. If an R&D intensity threshold is introduced, increase the expenditure threshold to $200 million so that large R&D-intensive companies retain an incentive to increase R&D in Australia. We do not support this recommendation and suggest that - in conjunction with a rejection of the intensity threshold requirement in recommendation 4 - the expenditure threshold is maintained at $100 million. We make this suggestion because any expenditure threshold is arbitrary and imposed primarily to ensure the sustainability of the program. By rejecting the intensity threshold but maintaining the existing $100 million threshold, which now caps the permanent benefit from the non-refundable Incentive at $8.5 million, the sustainability of the program can be maintained without the distortionary impact of the intensity threshold. Additional Recommendations / Alternatives for consideration A pre-approvals process should be introduced for business to exceed the $100 million threshold looking at spill-over, additionality and national interest of R&D to be undertaken. There are only a limited number of companies who would be in a position to seek to use this mechanism, and not all of these companies will go over the expenditure cap. Discretion would lie with program administrators, and only companies able to demonstrate that their claims of the R&D Tax Incentive for amounts in excess of the cap would achieve program objectives of additionality and spillover would be entitled to the offset. This change would introduce additional program complexity but the existing pre-approvals process could be adapted. In addition, it is submitted that the cost in terms of complexity would likely be offset by the additionality of the R&D expenditure this preapprovals process would stimulate (given that an assessment of additionality would be made prior to approval being given).

11 Page 11 Recommendation 6. That the Government investigate options for improving the administration of the R&D Tax Incentive (e.g. adopting a single application process; developing a single programme database; reviewing the two-agency delivery model; and streamlining compliance review and findings processes) and additional resourcing that may be required to implement such enhancements. To improve transparency, the Government should also publish the names of companies claiming the R&D Tax Incentive and the amounts of R&D expenditure claimed We support the premise of the recommendation - options for improving the administration of the program should continue to be investigated, but question whether the adoption of a single agency model would necessarily improve the overall administration of the R&D tax incentive. Review of the two-agency delivery model We would be cautious about amalgamating two bodies with different functions and purposes: The ATO, whose mission includes fostering willing participating in our tax and superannuation Systems; 4 and AusIndustry, a specialist program delivery division of the Department of Industry, Innovation and Science whose vision is to enable growth and productivity for globally competitive industries. 5 The existing split provides a good check and balance between the policy directive (encouraging use of the program R&D) and protecting tax revenue. We note that there is a potential for conflict between these objectives, and it would be difficult to reconcile revenue protection and R&D promotion into one body. One objective is likely to take primacy. We also note that the expertise required to administer the two facets of the programme (the assessment of R&D activities and the assessment of expenditure) will not change, and a move to a single-agency delivery model will not reduce the number of functions to be performed. In addition - because these are quite separate functions (technical and expenditure) - to move to a single agency model may not result in efficiency gains. 4 Vision, mission and values statement, Australian Taxation Office, accessed 18 October About us, Department of Industry, Innovation and Science, accessed 18 October 2016

12 Page 12 Improvements to administrative efficiency As such, it is suggested that the better approach is to keep the existing two agency model but seek to improve the program processes between the two organisations. There should be a focus on enhancing the sharing of information and the information systems and processes between the two agencies in order to create efficiencies for industry dealings with government. One possibility that is canvassed by the Review Paper is to create one registration form which goes to both agencies, eliminating some administrative inefficiency for both the administrative bodies and for claimants. There could be improvements to the application process, including amending the AusIndustry R&D Registration form to include expenditure matters. The R&D Registration Number could then be quoted on the Income Tax Return to claim the R&D benefit. Publishing the names of companies and the amounts of R&D expenditure claimed We would caution this suggestion for companies engaging in innovation in their respective industries as publishing amounts being spent on R&D will make transparent commercially sensitive information to their competitors and give away some of their competitive advantage. This is particularly problematic for private companies which are currently under no obligation of disclosing their R&D spend to the market. Additional Recommendations / Alternatives for consideration It is suggested that the program could benefit if AusIndustry adopted the culture of legislative rigour of the ATO, and place less reliance on AusIndustry prepared guidance/interpretations during compliance reviews, as some of these views have not been tested. Other potential enhancements to the compliance review and findings processes include the provision of sanitised information regarding both positive and negative findings made under sections 27J, 28A and 28C of the Industry Research and Development Act This would be similar to the current process of the ATO publishing sanitised private binding rulings (PBR) which can greatly aid taxpayers understanding in how AusIndustry might interpret and apply the R&D legislation. This would assist businesses to understand particular areas that have previously been determined as ineligible for the program thus reducing compliance costs. The release of broad based de-identified information regarding activities deemed ineligible would assist businesses in understanding if the activities they are conducting would meet the legislative requirements of the program and reduce the overall complexity of the program. We note that the ATO currently has similar mechanisms in place for rulings.

13 Page 13 Finally, we would suggest that technical support be provided to members of AusIndustry tasked with decision making. That is, it is important to ensure that sufficient technical advisory is available to decision makers (both scientific/technical and legal/interpretive) to promote consistency in the administration of the program. This support should be in the form of experts able to provide a consulting role on decisions relevant to their field of expertise, with the final decision being made by AusIndustry decision makers. This resource could take the form of a pool/panel of experts to be drawn upon as needed (including early on during the assessment process).