Treasurer Scott Morrison has given the R&D Tax Incentive a $2.4 billion haircut and promised to crackdown on what government says has been a misuse of the scheme, strengthening anti-avoidance laws and giving the ATO additional resources to hunt down cheats.
The changes drive three primary policy objectives: to improve the integrity of the system, to continue supporting the startup end of the R&D spectrum and to refocus support for larger companies with greater rewards for companies undertaking higher-intensity R&D.
As expected, the R&D Tax changes seek to reign in the ballooning cost of the incentive. When it was introduced in 2011-12 the scheme cost $1.8 billion annually. By 2016-17 that annual cost had climbed to nearly $3 billion.
The reforms map to government’s desire to reduce its fast-growing indirect support to industry – such as tax incentives – and to increase direct support through grants and other means.
The changes take $2.4 billion out of the system over the next four years, undoubtedly creating winners and losers, although precisely where will only become be known through the detail of the legislation.
“To support companies genuinely investing in R&D we are refocusing the R&D tax incentive to give more support to companies that invest a higher proportion of what they spend in R&D, over and above what others would do anyway,” Scott Morrison said in his Budget address.
The concern for industry is the continued confusion among startup companies in particular about what qualifies as a legitimate activity, coupled with plans for both a tightening of the anti-avoidance rules and for significantly better resourcing of enforcement within the Australian Taxation Office to deny ineligible claims.
The budget papers also allude to the strengthening of anti-avoidance rules through changed legislation, and has given additional resources to Innovation and Science Australia to produce more accurate and plainly-worded guidance about what constitutes eligible R&D – if only to make sure companies don’t inadvertently put themselves out of business by unintentionally misinterpreting the meaning of the law.
The savings of $2.4 billion across the forward estimates is effectively a redistribution of support, with large increases in science and technology support infrastructure, including significant new spending on supercomputing, the national GPS infrastructure, new artificial intelligence and machine learning capability, and a national space agency.
For companies with an annual turnover of $20 million or less, a $4 million cap on annual cash refunds has been introduced, although amounts that are in excess of the cap can become a non-refundable tax offset that can be carried forward into future income years.
The refundable tax offset has been increased to a premium of 13.5 per cent above the company’s tax rate for that year.
The changes also include a carve-out of sorts for clinical trials in the development of pharmaceuticals and medical devices, where R&D tax offsets for clinical trials are excluded from the $4 million cap.
Startups will be disappointed that government has still not moved to allow the tax refunds to be claimed quarterly.
As expected, the “refocusing” of the scheme seeks to reward larger companies with an increasing scale of support based on the intensity of the research effort (with ‘intensity’ defined as spending on R&D as a percentage of a company’s total annual expenditure). This is for companies with an annual turnover of $20 million or more.
The non-refundable R&D Tax offset is set at the claimant’s tax rate plus 4 per cent for R&D expenditure between 0 per cent and 2 per cent R&D intensity; the tax rate plus 6.5 per cent for R&D intensity between 2 per cent and 5 per cent; up to a top incentive of the tax rate plus 12.5 per cent for R&D expenditure above 10 per cent intensity.
The tax changes are a response to the so-called 3F review of the R&D Tax Incentive scheme, which was handed to government more than two years and four Industry ministers ago.
The review, by Innovation and Science Australia chief Bill Ferris, the Australian Government Chief Scientist Alan Finkel and Secretary to the Treasury James Fraser, found among other things that with larger companies, those with higher R&D intensity provided the greatest benefit to the Australian economy.
The new R&D ‘premium’ for larger companies simply encourages them to lift the intensity of their R&D spending above and beyond the ‘ordinary’.
The changes also increase the R&D spending threshold from $100 million to $150 million for larger companies to enable to continue to access the premium scheme as they undertake additional research as they grow.
The government has also flagged plans to change legislation to enable the ATO to publicly disclose for the first time how much a company has claimed and details of their R&D expenditure under the scheme.
Tech startup advocacy group StartupAus’ chief executive Alex McCauley welcomed the long-anticipated response to the 2016 Ferris, Finkel, Fraser review of the R&D Tax Incentive.
“On the whole the changes were fairly neutral for startups, with the government having listened to the sector’s advocacy in a number of key areas,” Mr McCauley said.
“But the Treasurer also noted the ATO would be ‘cracking down’ on the scheme, making a clarification of the central role of digital startups in the scheme more critical than ever.”
Mr McCauley noted that while some of the changes recommended in the 2016 review had been introduced, the impact on startups would be limited.
“Among other things, the FFF review had recommended keeping a lid on costs by introducing a $2 million cap on refunds under the scheme. Startups around the country said that would hurt them, and thankfully the Government listened, raising the cap to $4 million. We see this as a good compromise position,” he said.
As part of its Australian Technology and Science Growth Plan the government has earmarked $29.9 million over four years to strengthen the nation’s capability in Artificial Intelligence and machine learning – the first time these rapidly emerging platforms have attracted specific budget attention.
The measures are aimed at supporting economic growth and the productivity of Australian businesses, with the lion’s share of the allocation to be disbursed through Cooperative Research Centre-funded projects.
The CSIRO’s Data61 unit will fund AI and machine learning PhD scholarships and will develop a Technology Roadmap, Standards Framework and a national AI Ethics Framework to identify global opportunities and guide future investments.
The funding includes a modest sum for the Department of Education to develop curriculum and school-related learning tools to help address the skills gap.