Australia’s Research and Development Tax Incentive is one of the government’s largest innovation programs, providing billions in tax offsets to companies conducting R&D activities. In the 2023-24 financial year, the program delivered approximately $3.5 billion in benefits to more than 13,000 companies.

The policy logic is straightforward: R&D generates positive spillovers that benefit the broader economy, but individual companies under-invest in it because they can’t capture all the benefits. Government support helps correct this market failure, encouraging companies to invest in innovation they otherwise wouldn’t pursue.

But after years of operation and several rounds of reform, serious questions persist about whether the R&D Tax Incentive is actually achieving its goals or simply subsidizing activities companies would have done anyway.

What the Program Is Meant to Do

The R&D Tax Incentive allows companies to claim tax offsets for eligible R&D expenditure. For companies with turnover under $20 million, the offset is refundable—meaning they get cash back even if they don’t owe tax. For larger companies, it’s a non-refundable offset that reduces their tax liability.

The program is meant to support experimental activities that advance scientific or technological knowledge, not routine development work. Companies should be taking technical risks, attempting to solve problems where the solution isn’t obvious, and generating new knowledge.

In theory, this encourages companies to invest in riskier, more ambitious projects than they would under pure commercial logic. It supports early-stage technology development before there’s a clear path to revenue.

What’s Actually Being Claimed

The gap between the program’s intent and how it’s actually used has been a persistent concern. A significant portion of R&D Tax Incentive claims come from companies doing what looks a lot like normal product development, not fundamental research.

Software companies claim tax offsets for building new features in their products. Manufacturers claim for designing new product variations. Mining companies claim for operational improvements that reduce costs. These activities might technically meet the eligibility criteria, but they’re often things companies would have done anyway to remain competitive.

The Australian National Audit Office reviewed the program in 2022 and found significant issues with compliance and oversight. Many claims that were approved shouldn’t have been, while some legitimate R&D was rejected. The eligibility criteria are complex and subject to interpretation, creating inconsistent outcomes.

Industry insiders acknowledge that the program has become something of an open secret—companies know that with the right documentation and phrasing, they can classify substantial portions of their normal work as R&D. Specialist consultants help companies structure their operations and documentation to maximize R&D claims, taking a percentage of the benefit as their fee.

The Additionality Question

The critical economic question is “additionality”—would companies have done this R&D without government support, or did the tax incentive genuinely cause additional innovation to occur?

If companies are just claiming offsets for work they’d do anyway, the program is transferring taxpayer money to businesses without generating additional economic benefit. That’s expensive corporate welfare, not effective innovation policy.

Research on additionality in R&D tax incentives internationally shows mixed results. Some studies find substantial additionality—every dollar of tax support generates more than a dollar of additional R&D. Others find minimal additionality, with companies primarily relabeling existing activities as R&D to claim the benefit.

Australian-specific research is limited, but what exists suggests the additionality effect is modest. A 2016 study by the Australian Bureau of Statistics found that for every dollar of R&D tax support, companies increased their R&D spending by about $1.30—meaning the program is generating some additional investment, but not dramatically transforming company behavior.

For smaller companies, particularly startups, the additionality appears stronger. The refundable tax offset provides much-needed cash flow for companies burning through venture funding, allowing them to extend their runway and pursue more ambitious technical development.

For large, profitable corporations, the additionality is more questionable. These companies have R&D budgets that are determined by strategic necessity and competitive dynamics, not tax incentives. The R&D offset reduces their costs but probably doesn’t meaningfully change how much R&D they conduct.

The Compliance Burden and Gaming

The R&D Tax Incentive has created a substantial compliance industry. Companies hire consultants to prepare claims, document their R&D activities, and defend against audits by AusIndustry (which administers the program) and the ATO.

This compliance cost eats into the program’s effectiveness. If companies are spending 10-15% of their R&D benefit on consultants to claim it, that’s deadweight loss that doesn’t support innovation.

There’s also concerning evidence of gaming. Some companies structure their operations specifically to maximize R&D claims rather than to maximize actual innovation. Work that could be done more efficiently gets broken into smaller experiments to generate more R&D-eligible activities.

The software industry is particularly prone to this because the line between R&D and normal development is ambiguous. Is building a new AI feature in your product R&D? It involves solving technical problems without obvious solutions, which meets the eligibility criteria. But it’s also core product development that any competitive company would do regardless of tax incentives.

The Innovation Ecosystem Impact

Beyond direct additionality, there’s a question about whether the R&D Tax Incentive contributes to a stronger innovation ecosystem. Does it help build capabilities, train researchers, or enable collaborations that have spillover benefits?

The evidence here is more positive. Companies that claim R&D tax offsets do hire more technical staff, publish more patents, and collaborate more frequently with universities. The program does appear to support research capability building, particularly in smaller firms that are growing their technical teams.

However, the program’s design doesn’t specifically encourage the types of R&D most likely to generate broad economic benefits. A company doing proprietary product development that won’t share any learnings externally gets the same support as a company conducting research that advances the broader field.

Some other countries address this by providing higher tax offsets for collaborative research with universities or for research in strategically important areas. Australia’s program is largely sector-agnostic and doesn’t differentiate based on potential spillovers.

The Reform Challenge

The R&D Tax Incentive has been reformed several times—in 2011, 2016, and most recently in 2021—but the fundamental tensions in the program remain. Make the eligibility criteria too strict, and legitimate R&D gets excluded while companies waste resources on compliance. Make them too loose, and the program subsidizes routine business activities.

Increasing audit intensity helps ensure claims are legitimate, but it also increases compliance burden and creates uncertainty that may deter genuine R&D investment. Reducing audit intensity lowers compliance costs but allows more inappropriate claims.

There’s probably no perfect solution to this trade-off. Any broad-based R&D tax incentive will involve some subsidy of activities that would have happened anyway and some compliance gaming. The question is whether the genuine innovation it supports justifies these inefficiencies.

Alternative Approaches

Some innovation policy experts argue that instead of a broad R&D tax incentive, Australia would be better served by more targeted programs that support specific strategic priorities or types of innovation.

Direct grants to companies working on defined challenges—like carbon reduction technology, agricultural innovation, or medical devices—could ensure government support goes to areas of national priority rather than whatever companies choose to classify as R&D.

Prize-based innovation incentives, where government offers substantial rewards for achieving specific technical milestones, could drive competition and focus effort on high-value problems.

Increased funding for university research and better mechanisms to commercialize that research might generate more spillovers than subsidizing private company R&D.

However, these targeted approaches have their own challenges. Governments aren’t necessarily good at picking winners or defining the right challenges. Directed innovation programs create opportunities for political interference and lobbying. Broad-based tax incentives at least let companies decide what R&D is worth pursuing.

The Startup Perspective

For Australian startups and early-stage tech companies, the R&D Tax Incentive is often described as critical. The refundable offset provides cash flow that allows companies to continue operations between funding rounds, hire technical talent, and pursue longer-term development.

Several Australian tech success stories—Canva, Atlassian in its early days, Safety Culture—have credited the R&D Tax Incentive with helping them survive the difficult early years and scale up operations.

From this perspective, even if the program has inefficiencies and is sometimes gamed by larger companies, its support for high-growth startups justifies the overall cost. These are companies doing genuinely risky technical development that might not happen without support.

However, there’s a counter-argument that if the R&D is genuinely valuable, the venture capital market should fund it without taxpayer subsidy. The VC response is that Australian venture capital is underdeveloped compared to the US or even Europe, partly because exits are smaller and less frequent. The R&D Tax Incentive partially compensates for this gap.

What the Data Actually Shows

Trying to definitively answer whether the R&D Tax Incentive drives meaningful innovation is difficult because we can’t observe the counterfactual—what would have happened without the program.

What we can see is that Australian business R&D spending as a percentage of GDP has remained roughly flat over the past decade despite the tax incentive. This suggests the program isn’t dramatically transforming overall innovation investment, though it might be preventing decline or supporting different types of R&D than would otherwise occur.

Patent filings by Australian companies have increased, though it’s unclear how much of this is due to the R&D Tax Incentive versus other factors like industry trends and global competitive dynamics.

Startup formation and growth in certain sectors—particularly software and biotech—does seem correlated with R&D Tax Incentive usage, though again, causation is hard to establish definitively.

The Verdict

Australia’s R&D Tax Incentive is probably doing some good—supporting startup growth, encouraging companies to hire technical staff, and enabling some risky projects that create valuable spillovers. But it’s also subsidizing substantial amounts of normal business activity, creating compliance burdens, and likely delivering less innovation per dollar than more targeted approaches could achieve.

The program’s greatest value may be for early-stage, high-growth companies where cash flow support and reduced risk genuinely enables additional innovation. Its weakest performance is probably in subsidizing routine development work by large, profitable companies that would do it anyway.

Whether that balance justifies the $3.5 billion annual cost is ultimately a political judgment, not a purely economic one. But the evidence suggests that the R&D Tax Incentive is a blunt instrument that could likely be improved through better targeting, stricter compliance, or alternative innovation policy approaches.