5 - Policy report: Integrity within the R&D tax credit
|Date||13 March 2018|
|Title||Integrity within the R&D tax credit|
|Date:||13 March 2018||Priority:||Medium|
|Security Level:||In Confidence||Report No:||IR2018/132
|Minister of Research, Science and Innovation||Agree to inclusion of proposals in the R&D Tax Credit Discussion Document||22 March 2018|
|Minister of Revenue||Agree to inclusion of proposals in the R&D Tax Credit Discussion Document||22 March 2018|
Contact for telephone discussion (if required)
|Keith Taylor||Policy Manager||[Withheld under section 9(2)(a) of the Official Information Act 1982]|
|Richard Walley||Manager, Innovation Policy|
|Becci Whitton||Manager Stakeholder and Government Engagement|
9 March 2018
Minister of Research, Science and Innovation
Minister of Revenue
Integrity within the R&D Tax Credit
1. This report seeks your approval to proposals to be included in the Discussion Document - R&D Tax Incentive – building a diverse, knowledge-intensive economy through Research and Development.
2. It responds to a request from Ministers for further advice around measures to ensure the integrity of the tax credit. It proposes certain adjustments to the definition of eligible R&D. It responds to a request from Ministers at DEV for advice about the penalty rules that would apply to inappropriate claims under the tax credit.
3. Integrity will be an important feature for the R&D tax credit. Perceptions that the scheme is rewarding claims that do not represent genuine R&D or that tax advisors are claiming too great a share of the expenditure will undermine the credibility of the scheme.
4. A review of claims from the 2008 tax credit highlights circumstances in which firms were able to receive the tax credit for business as usual expenditure. Consequently, officials propose tightening the rules in two areas – where there is dual purpose expenditure and where firms are being paid by a third party to undertake work.
5. Software is an important area of R&D expenditure. It is also an area where officials consider tightening of the rules that applied in 2008 could be warranted. We are working on developing a robust set of proposals for defining eligibility of software expenditure. We expect this to be complete by May 2018 and anticipate following this by targeted consultation with relevant stakeholders.
6. Officials consider the shortfall penalties framework that applies to all tax returns is a suitable basis for applying penalties to incorrect R&D tax credit claims. However, we also consider the risks around R&D tax credits may be greater when advisors are paid on a contingency basis. Therefore an extension to the framework is proposed where the offense is non-trivial and the advisor shares in the value of the claim.
We recommend that you
Note that a review of claims from the 2008 R&D tax credit indicated circumstances where applicants were able to receive the credit for business as usual expenditure
Agree that the Discussion Document include the proposals:
- that to be an eligible activity for the R&D tax credit, the activity must be for the sole purpose of R&D
- that eligible expenditure does not include expenditure when the R&D performing entity or its associate had received or could reasonably be expected to receive consideration
Agreed / Not agreed Agreed / Not agreed
Note that limits or exclusions may be appropriate with respect to internal software development and certain types of activities within the software development process
Agree that once officials have developed proposals with respect to eligibility of software, they undertake consultation with targeted stakeholders
Agreed / Not agreed Agreed / Not agreed
Note the shortfall penalties framework that applies to all tax returns will apply to the R&D tax credit
Note the risk of inappropriate claims may be greater where tax advisors are paid on a contingency basis
Agree that the Discussion Document include a proposal that the penalties framework be extended to apply to an advisor who receives a fee contingent on the R&D tax credit and the tax credit application demonstrates gross carelessness or a more serious offense.
Agreed / Not agreed Agreed / Not agreed
[Withheld under section 9(2)(a) of the Official Information Act 1982]
Manager, Innovation Policy
Hon Dr Megan Woods
Minister of Research, Science and Innovation
Hon Stuart Nash
Minister of Revenue
7. Cabinet has agreed, subject to Budget decisions, to the introduction of a Research and Development (R&D) tax credit from 1 April 2019. (CAB-18-MIN-0056 refers) Officials have prepared a discussion document (2216 17-18, IR2018/133 refers) based on decisions by Ministers on the design features for the tax credit (1714 17-18, IR2018/083 and 1892 17-18, IR2018/084 refer)
8. As with any scheme involving expenditure of public money, there is a need to ensure the integrity of that expenditure. There are risks of inappropriate claims with an R&D tax credit.
9. This report responds to a request from Ministers for further advice around measures to ensure the integrity of the tax credit. It proposes certain adjustments to the definition of eligible R&D and seeks Ministers’ approval for their inclusion in the Discussion Document. It also responds to a request from Ministers at DEV for advice about the penalty rules that would apply to inappropriate claims under the tax credit.
10. This report has three sections. The first outlines the framework for the R&D tax credit and how this supports its integrity. The second proposes some particular refinements to the definition of R&D that was described in the technical design features report in order to support appropriate expenditure. The third outlines possible new penalties for the R&D tax credit to complement existing penalty rules.
11. As noted in a previous report (1714 17-18, IR2018/083 refers), the objectives for designing an R&D tax credit are to provide easily accessible support to a broad range of business R&D, in a fiscally responsible way, while maintaining trust and confidence in the tax system. Each element of these objectives has integrity implications.
12. Providing easily accessible support means that the compliance burden associated with applying for the tax credit should be kept to a minimum. This requires the information requirements be consistent with normal business practices.
13. It also means there is as little ambiguity as possible as to what constitutes eligible R&D expenditure. This can be challenging as research and development, as it is used colloquially, doesn’t have a precise definition. Firms may consider they are doing R&D because a project is innovative and challenging but not comply with the definition within the legislation. Consequently, an important aspect of the definition of R&D is to favour clarity as to what qualifies and what doesn’t, even if – at the margin – this might disqualify some expenditure that is worthy of support.
14. It is also the case that government will learn with experience. There needs to be a flexible regulatory environment both to keep up with how R&D is changing for firms and to address what appears to be abuse of the scheme. Though stability of the scheme is an objective because this will be conducive to firms planning to undertake R&D, it has to be tempered by a commitment to adjust the legislation when warranted.
15. Applying the concept of fiscal responsibility to integrity is also challenging. On the one hand, increased expenditure as a result of firms undertaking more R&D is exactly what the scheme is aiming for. On the other hand, rapid increases in expenditure under the scheme may signal a proliferation of low value claims.
16. Perceptions that the scheme is rewarding claims that do not represent genuine R&D or that tax advisors are claiming too great a share of the expenditure will undermine the credibility of the scheme.
17. As stated in the Discussion Document, the Government is committed to monitoring the scheme and will have the ability to speedily identify and remedy issues that could compromise the integrity of the scheme. This sends a clear signal that the Government will maintain a focus on the scheme meeting its objectives and delivering value for money.
18. Being part of the tax system means that the tax credit will influence perceptions of that system. The tax system relies on voluntary compliance which in turn hinges on public confidence that the system is fair. An erosion of trust in one part of the system can undermine compliance in other parts of the tax system.
19. Inland Revenue will audit a selection of claims to test their validity and enforce compliance with the legislation. But this will not be a comprehensive safeguard so the transparency measures (discussed in 1892 17-18, IR2018/084) and integrity measures discussed in this report will be necessary adjuncts to support the integrity of the tax credit.
20. The technical design features briefing (1892 17-18, IR2018/084 refers) outlined the definition of R&D, including eligible and ineligible activities and expenditure categories. Since finalising that report, officials have completed a review of claims from the 2008 R&D tax credit. This has led us to propose certain refinements to the R&D definition. These have been incorporated into the draft Discussion Document that has been presented to Ministers as an attachment to briefing 2216 17-18, IR2018/133.
21. A common theme from the 2008 claims is that firms undertook R&D as part of a broader project. The claim was for all the project; business as usual expenses were included as R&D. The result was claims many times greater than what the IR investigators considered was genuine R&D. However, applicants successfully argued that their claims were permissible according to the letter of the law.
22. For the 2019 tax credit, officials therefore propose the following adjustments to the previous definition in order to provide greater clarity and to better target the credit to genuine R&D.
Issue: Dual purpose R&D
23. [Withheld under section 81 of the Tax Administration Act 1994]
24. Officials consider R&D can legitimately occur with respect to the operation of a manufacturing plant. However, the appropriate amount of eligible R&D expenditure would be the extra expenditure incurred over what it would have cost to manufacture the firm’s output without any R&D. This would isolate the actual cost of the R&D from the business as usual costs of the entity.
25. One way to address this would be to introduce a “to the extent” test. This would mean that where an activity has an R&D purpose and another purpose (in this case manufacturing output), the applicant would be required to apportion expenditure to R&D and the other purpose.
26. The problem with this test is that its application is fact specific. This will potentially lead to a high administration and compliance costs associated with establishing how much of an activity is for an R&D purpose and how much for another purpose. There is also a risk that it would not be as robust as expected and still allow recharacterisation.
27. Officials therefore propose a clearer rule that to be eligible, an activity must be for the sole purpose of R&D. This would mean that dual purpose activities would not be eligible R&D expenditure. In the example above, this rule would have excluded the costs of the manufacturing operation as eligible R&D expenditure because there would have been an additional purpose of producing output for sale.
28. This rule would not exclude genuine pre-production trials and manufacturing innovation. For example, if a production process is run solely to trial an innovative new method, this could be eligible.
29. Nonetheless, there is a risk of over-reach with this exclusion – that is valid R&D is excluded or firms incur additional costs in order to separate R&D from other activities. However, officials note the approach is consistent with the brightline (a clear boundary) test used elsewhere in the tax system. It is also similar to that adopted in other jurisdictions.
30. On balance, officials consider the benefits arising from clarity are likely to outweigh the possible costs. The proposal will be highlighted in the Discussion Document so will be tested through consultation.
31. Proposal: To be an eligible activity for the R&D tax credit, the activity must be for the sole purpose of R&D.
Issue: Eligibility where there is commercial consideration
32. [Withheld under section 81 of the Tax Administration Act 1994]
- The 2008 scheme had an “at risk” rule - firms claiming the tax credit had to bear the risk of the R&D expenditure. The intention was that the party commissioning the R&D would claim for the R&D tax credit.
- Often the investigators considered the genuine R&D was a small part of the project but the applicant successfully claimed for most of the project cost. The rules applying in 2008 did not enable claims to be confined to the genuine R&D portion.
33. Officials consider these claims are not consistent with the intention of the scheme and therefore propose to strengthen the “at risk rule” by incorporating a restriction that is included in the Australian R&D tax credit. This specifies that eligible expenditure does not include expenditure when the R&D performing entity or its associate had received or could reasonably be expected to receive consideration.
34. This rule, in conjunction with the restriction on dual purpose R&D, would mean that where a commissioned project included R&D, the commissioning party could claim the credit provided the R&D-related portion was identified as a separate project. If the R&D performing entity undertook extra R&D separate from what is was commissioned to undertake, it could claim a credit for that extra R&D.
35. Proposal: Eligible expenditure does not include expenditure when the R&D performing entity or its associate had received or could reasonably be expected to receive consideration.
36. Software R&D will be an area of significant expenditure within the R&D tax credit. It accounted for approximately 40% of all tax credit claims under the 2008 credit and Australian officials indicated it makes up a higher proportion of expenditure under its R&D tax credit.
37. Officials consider that software development activities should qualify for the R&D tax credit where the activity meets the core or supporting R&D activity definitions. However, across OECD countries, mapping the definition of R&D onto software is an acknowledged difficulty. Inland Revenue investigators reported that when they investigated claims it was difficult to distinguish between R&D and standard software development. In itself, this means that re-examination of the 2008 rules is warranted. Also, technological change means that rolling over the 2008 rules relating to eligible software expenditure may not be appropriate.
38. The only limit placed on software under the 2008 credit was a $3 million cap for eligible expenditure on internal software development. Internal software development was an area of particular concern for investigators of the 2008 claims. Officials are therefore considering whether internal software development should be excluded from eligible expenditure altogether, as is the approach in Australia.
39. Another approach adopted by other countries is to indicate, via guidelines or within legislation, the type of activities that would be eligible or ineligible for the tax credit. An example of the first could be developing new operating systems and of the second could be security testing. Officials consider that having a schedule of the software activities that would not qualify for the tax credit, with the ability to update that schedule more quickly than by primary legislation, would make the tax credit more responsive to changes in the way software is developed and used.
40. Officials will continue to work on these issues. Having a robust approach to software will be important for the integrity of the R&D tax credit. It is anticipated that this work will be completed by May 2018.
41. Within the draft Discussion Document, this is flagged as an area where more work is being undertaken. Once more robust proposals have been developed, officials propose undertaking targeted consultation with relevant stakeholders.
42. Voluntary compliance is the fundamental basis of the tax system. Administration of the R&D tax credit will also rely on voluntary compliance by taxpayers. Guidance and education will be provided to assist taxpayers understand what is eligible and make correct claims.
43. However, there is a risk with the R&D tax credit of applicants submitting incorrect claims. Penalties are one factor that will deter this.
44. Within the tax system, the shortfall penalties framework applies to all tax returns. This establishes a hierarchy of offenses from lack of reasonable care at the lowest level to tax evasion involving fraud at the highest. Appropriate penalties are associated with each level of offense, with civil penalties applying to the lower level offenses and criminal penalties applying to the higher level offenses.
45. Officials consider the shortfall penalties framework provides a suitable basis for applying penalties to incorrect R&D tax credit claims. However, we also consider the risks around R&D tax credits may be greater when advisors are paid on a contingency basis as this means they gain an incentive to inflate the claim. For this reason it is proposed that:
i. where it is found that the R&D tax credit application demonstrates gross carelessness or a more serious offense:
ii. if an external advisor has received or would have received a direct financial benefit from the claim (in the form of a fee contingent on the R&D tax credit),
the advisor will be joint and severally liable (with the taxpayer) for the appropriate penalty (including repayment of tax shortfall and interest).
46. This proposal has been included in the draft Discussion Document.
47. There are two further elements which officials are investigating as possible extensions to the penalties regime. These are not covered in the draft Discussion Document.
48. A risk is that advisors might promote template schemes for claims under the R&D tax credit. Officials are doing further work to assess whether the current promoter penalties regime is adequate to cover this situation or should be revised.
49. Officials are also investigating whether there should be an over-ride to the standard secrecy provisions applying to tax records so that Inland Revenue could report a tax advisor associated with problematic claims to the appropriate professional body.
50. The Treasury was consulted in the preparation of this report.
51. The Department of Prime Minister and Cabinet was informed.
 The US excludes research that is conducted after the beginning of commercial production; Ireland requires eligible expenditure to be wholly and exclusively for the carrying on of the research; Australia requires a production activity to be for the dominant purpose of supporting R&D.