Revenue divides the R&D credit into two tests and names them. The science test asks whether the activities are qualifying R&D activities; the accounting test asks whether the expenditure is qualifying expenditure. The science test is the one that decides most disputes, because it turns on facts about the technology that Revenue cannot assess without help, and the Act gives Revenue the power to bring in an independent expert. This page sets out the test as the Act and the manual frame it.
What it is
Qualifying R&D activities are systematic, investigative or experimental activities in a field of science or technology. They must be basic research, applied research or experimental development, must seek to achieve scientific or technological advancement, and must involve the resolution of scientific or technological uncertainty. Revenue's manual turns the definition into four questions, and all four must be answered yes.
| Question | What it means | Evidence |
|---|---|---|
| 1. Is the activity in a prescribed field? | Natural sciences, engineering and technology, medical sciences or agricultural sciences as listed in the 2004 Regulations; not social sciences, arts or humanities | The field named in the project note |
| 2. Is it systematic, investigative or experimental, and basic research, applied research or experimental development? | A planned investigation with hypotheses, tests and conclusions, not trial and error or routine work | Plans, experiment records, results |
| 3. Does it seek scientific or technological advancement? | An advance in the field's knowledge or capability, judged against the state of the art, not against the company's own knowledge | Literature or state-of-the-art review at the outset |
| 4. Does it involve resolving scientific or technological uncertainty? | Whether a result is achievable, or how to achieve it, was not known and could not be readily deduced by a competent professional in the field | The uncertainty statement and the record of attempts |
Who it applies to
Every project in every claim. The test is applied activity by activity, and a project can contain qualifying and non-qualifying activities side by side. Revenue's expectation is that the company identifies the qualifying activities, applies the test to each and keeps the evidence, rather than claiming the whole project and leaving Revenue to find the R&D within it.
Revenue applies the test with the help of experts where it needs to. The Act allows Revenue to consult any person with expertise in the field, after notifying the company, and the expert may examine the records and meet the company's staff. In practice experts are drawn from universities and industry under contract to Revenue, and their report is central to Revenue's decision. The company can challenge the report at the Tax Appeals Commission with its own expert evidence.
How it works
- Name the field. Place the activity within a prescribed category. Software claims should name the sub-field of computer science or software engineering.
- State the advancement sought. What would the field know or be able to do afterwards that it could not before? Describe the state of the art at the start and cite what was checked.
- State the uncertainty. What could a competent professional not know or readily work out? Distinguish technological uncertainty from commercial, financial or project uncertainty, which do not count.
- Show the systematic investigation. The hypotheses, the experiments or analysis, the results and the conclusions, in dated records.
- Check the exclusions. Confirm the activity is not one the Regulations exclude, such as routine testing, cosmetic changes or business application software.
- Set the boundaries. R&D starts when work to resolve the uncertainty begins and ends when it is resolved or abandoned. Scale-up, commercial production and routine development after that point are outside the claim.
Two features of Revenue's reading deserve emphasis. Advancement is measured against the field, so a company catching up with what competitors already know is not advancing science or technology, however new the work is to it. And uncertainty is judged by the competent professional, so the company's own inexperience does not create uncertainty.
Worked example
Illustrative example. Two hypothetical projects at Slaney Devices Limited.
| Question | Project A: low-power wireless sensor | Project B: customer portal rebuild |
|---|---|---|
| Field | Electronic engineering | Software engineering |
| Systematic investigation | Hypotheses about antenna geometry and duty cycling tested in a controlled series; results logged | Agile delivery of planned features; no hypotheses tested |
| Advancement | Published designs could not reach 5 years' battery life at the required range; the project sought to show it could be done | A better portal for the company; nothing new to the field |
| Uncertainty | Whether the target was achievable at all was unknown to competent engineers; two approaches failed before one succeeded | How long the work would take and whether users would like it: commercial, not technological |
| Exclusions | None | Business application software using known methods and tools |
| Result | Passes the science test | Fails |
Slaney claims Project A and leaves Project B out. When Revenue's expert reviews Project A, the dated test logs answer the four questions in the order the expert asks them.
Common questions
Does the work have to be novel to the world? It has to seek an advancement in the field. Work that reproduces what is already known in the field, even if not published, is not an advancement.
What if we were confident it would work? If a competent professional could have been confident on the basis of existing knowledge, there was no uncertainty and the activity does not qualify.
Can a failed project pass? Yes. The test is whether the activity sought an advancement and involved resolving an uncertainty, not whether it succeeded.
Who decides, Revenue or the expert? Revenue decides, informed by the expert's report. The Tax Appeals Commission decides on appeal.
Related pages
Official sources
Educational content only, not tax advice. See the disclaimer.