Budget's Innovation Bet: Who Bears the Cost When Incentives Misfire?
Australia is spending $39.1 billion on innovation — but how much of it is actually changing what businesses would have done regardless?
The Albanese government's budget innovation package is, on paper, a coherent set of measures: loss carry-backs to reward risk, refundable tax offsets for start-ups, expanded venture capital incentives, and a beefed-up R&D tax incentive. Treasurer Jim Chalmers and Industry Minister Tim Ayres have framed it as the mechanism by which Australia's next generation of technology companies gets off the ground. The question worth asking is not whether the goals are admirable, but whether the design of these measures actually produces the outcomes claimed, and at what cost to everyone else.
The loss carry-back is the package's strongest instrument — and the exception, not the rule
The package's strongest element is probably the two-year loss carry-back for companies up to $1 billion in turnover. This is a reasonably well-designed instrument. A firm that takes a genuine risk, invests heavily, and loses money in the short term currently faces an asymmetric tax treatment: profits are taxed immediately, losses are only useful when offset against future profits. Loss carry-back addresses that asymmetry. There is real economic logic here. If you want companies to take risks, the tax system should not penalise the downside of risk disproportionately. This one mostly works in the direction it is supposed to.
Making the R&D incentive more generous does not solve its eligibility problem — it scales it
The R&D tax incentive is trickier. Australia has run versions of this scheme for decades, and the pattern is familiar: firms that do genuine research claim it, firms that call ordinary business improvement "research" also claim it, and the Australian Tax Office spends considerable energy trying to tell the difference. The government's announced changes, increasing offset rates by 25 to 50 per cent and raising expenditure thresholds, are generous, but generosity to a scheme with fuzzy boundaries does not solve the fuzziness.
A 2016 review by the Department of Industry found that a significant share of R&D expenditure claimed under the incentive was for activities that had low additionality, meaning the company would have done the work regardless of the tax subsidy.
A 2016 review by the Department of Industry found that a significant share of R&D expenditure claimed under the incentive was for activities that had low additionality, meaning the company would have done the work regardless of the tax subsidy. Making the incentive more generous does not correct that problem. It scales it.
The $39.1 billion headline is a portfolio, not a strategy
The $39.1 billion headline figure, spread across higher education, grants, scientific organisations, defence and agricultural research, tells you something important: this is not a targeted innovation bet. It is a very large portfolio of public spending that has been aggregated under an innovation narrative for budget announcement purposes. Some of that spending, CSIRO research, national measurement infrastructure, the Square Kilometre Array, is legitimate public investment in scientific capability where the case for public funding is well-established. Basic science has genuine public-good characteristics: the returns are diffuse, the time horizons are long, and private capital consistently underinvests. That part of the package is defensible.
Government VC incentives risk drawing in more capital, not better capital
The venture capital incentives are where the analysis gets genuinely uncomfortable. Governments expanding VC incentives face a structural problem: the skill required to identify which early-stage companies will generate returns is exactly the skill that distinguishes good private venture capitalists from the rest of the market. If the government's role is to subsidise more VC activity, the question is whether that subsidy draws in better capital or simply more capital. More capital chasing the same set of viable start-ups does not produce more innovation. It produces higher valuations and more money losing its way into the market.
Public innovation spending lacks the feedback loop that disciplines private capital
There is also an accountability gap that rarely gets examined directly. Private venture capital operates on a feedback loop with real teeth. A fund that backs too many losers does not raise a second fund. The incentive to be right is direct and financial. Public innovation spending operates through a much longer chain: a programme funds a company, the company underperforms, an audit is eventually conducted, the findings land in a minister's inbox three years after the funding decision, and by that point the political and institutional incentives all favour explaining why the decision was reasonable rather than learning from why it failed. This is not a criticism unique to this government. It is a structural feature of how public capital allocation works, and it does not disappear because the intent behind the spending is genuine.
None of this is to say the government should not back innovation at all. Public research infrastructure, basic science, and correction of genuine market failures in early-stage capital markets are appropriate uses of public money. The design of some of these measures, particularly the loss carry-back, reflects real understanding of how investment decisions work. But at $39.1 billion, the aggregate number demands a level of rigour about what is actually driving genuine additionality and what is subsidising activity that private markets would have funded anyway.
Australians picking up the tab deserve that question answered plainly, not buried under a headline figure and a press release that runs from the CSIRO to the Australian Space Agency without pausing to account for the connection between them.
Sources
Treasury Ministers — A Budget that backs innovation and investment
Frequently Asked Questions
What is additionality and why does it matter for R&D tax incentives?
Additionality refers to whether a tax subsidy causes a company to do something it would not have done otherwise. If a firm claims the R&D tax incentive for work it was going to do regardless, the subsidy costs public revenue without producing any new research or innovation — the government is simply paying for activity that would have happened anyway.
Why is loss carry-back considered a better-designed innovation measure than an R&D tax offset?
Loss carry-back directly addresses a real distortion in the tax system: profits are taxed immediately but losses can only be used against future profits, which penalises risk-taking. The fix is structural and the eligibility is relatively objective. R&D offsets, by contrast, require the tax office to judge whether business activity counts as genuine research — a boundary that is persistently contested and gamed.
Does more government funding for venture capital produce more innovation?
Not automatically. The capacity to pick which early-stage companies will generate returns is the defining skill of a successful venture capitalist, and government subsidies tend to increase the volume of capital in the market rather than its quality. More capital competing for the same viable start-ups raises valuations without necessarily funding companies that would not otherwise have been funded.
Why is it hard for governments to learn from failed innovation spending?
Public innovation programmes typically run on long feedback cycles: a company is funded, underperforms, an audit follows years later, and findings reach ministers long after the original decision. By that point, political and institutional incentives favour defending past decisions rather than learning from them — a structural problem that applies regardless of which party is in government.
Is all of the $39.1 billion in Australia's innovation budget actually targeted at innovation?
No. The figure aggregates higher education funding, grants, scientific organisations, defence research, and agricultural research under a single innovation narrative. Some of that — basic science, national research infrastructure — represents legitimate public investment with well-established justifications. The aggregate figure obscures how much is genuinely aimed at catalysing private-sector innovation versus funding activity the government would have funded anyway.
Frequently Asked Questions
What is additionality and why does it matter for R&D tax incentives?
Additionality measures whether a tax subsidy actually causes a company to do something it would not have done otherwise. When a firm claims the R&D tax incentive for work it planned to do regardless, the subsidy drains public revenue without producing any new research — the government is paying for activity that was going to happen anyway.
Why is loss carry-back considered better policy than an R&D tax offset?
Loss carry-back corrects a concrete distortion: the tax system taxes profits immediately but only allows losses to be used against future profits, which penalises risk-taking in a direct and measurable way. R&D offsets require the tax office to judge whether a given business activity qualifies as genuine research — a boundary that is persistently disputed and exploited.
Does more government funding for venture capital produce more innovation?
Not automatically. The ability to identify which early-stage companies will generate returns is precisely what distinguishes good private venture capitalists, and government subsidies tend to increase the volume of capital chasing deals rather than its quality. More capital competing for the same viable start-ups drives up valuations without necessarily funding companies that would otherwise have gone unfunded.
Why do governments struggle to learn from failed innovation programmes?
Public innovation spending runs on long feedback cycles: a company is funded, underperforms, an audit follows years later, and findings reach ministers long after the original decision was made. By that point, political and institutional incentives favour defending the original call rather than drawing honest lessons from failure — a structural problem that has nothing to do with the intentions of any particular government.
Is all of Australia's $39.1 billion innovation budget actually aimed at spurring private innovation?
No. The figure bundles higher education, scientific organisations, defence research, and agricultural research under a single innovation label. Basic science and national research infrastructure within that total have well-established public-good justifications, but the aggregate number obscures how much is genuinely designed to catalyse private-sector activity versus funding what government would have spent anyway.