Australia's economy is failing its tech startups and the data proves it
Australia produces world-class technology companies at extraordinary capital efficiency, then systematically drives them offshore through policy failure. The nation collects more tax from beer drinkers than from multinational gas corporations extracting irreplaceable resources, charges students more than it charges fossil fuel giants, and runs a competition regulator that has lost every contested merger case in recent memory — all while a $4.2 trillion superannuation pool sits largely uninvested in the domestic startup ecosystem. This report assembles the empirical evidence across five critical dimensions of Australian economic policy failure.
1. The brain drain is real, and the numbers are staggering
Australia's startup ecosystem is worth US$360 billion in 2025 — the second-fastest growing venture ecosystem globally, up 6.5x since 2018. The country has produced six decacorns (companies valued above US$10 billion) and generates 1.5 unicorns per US$1 billion of VC invested, making it the most capital-efficient startup ecosystem in the world, ahead of both the US and Israel at 1.1 unicorns per billion. Yet the pattern is unmistakable: once Australian startups reach global scale, they leave.
Atlassian, founded in Sydney in 2002, listed on NASDAQ in December 2015 at a US$4.4 billion valuation and redomiciled its parent entity from the UK to Delaware in 2022. Its current market capitalisation fluctuates between US$17–22 billion (peaking above US$80 billion in 2021), and it generated US$5 billion in revenue and US$1.4 billion in free cash flow in FY25. It remains operationally headquartered in Sydney but is legally a US company. Canva, valued at US$42 billion, redomiciled to Delaware in early 2025 ahead of an anticipated NASDAQ IPO in H2 2026. The ATO treated the redomiciling as a vesting event, triggering tax bills for Australian employees. Afterpay was acquired by US-based Block Inc. in January 2022 in an all-stock deal originally valued at A$39 billion (US$29 billion), delisting from the ASX permanently.
Airwallex, Melbourne-founded and now valued at US$12 billion after a US$498 million Series G in 2025, has all but confirmed a NASDAQ listing. Rokt, founded in Sydney in 2012 by former Jetstar CEO Bruce Buchanan, relocated its headquarters to New York and trades at secondary valuations of US$4.1–7.2 billion with revenue forecast to exceed US$1 billion by August 2026. AirTrunk, Australia's largest data centre company, was acquired by Blackstone for approximately A$24 billion in 2024 — now entirely US-owned. Zoox, co-founded by Australian Tim Kentley-Klay, was acquired by Amazon for US$1.2 billion. BigCommerce went public on NASDAQ in 2020, hitting a US$5 billion valuation in its first trading week. Deputy moved its headquarters to San Francisco. Synchron, the brain-implant startup spun out of the University of Melbourne, is now headquartered in New York.
A conservative tally of Australian-founded companies now primarily domiciled, listed, or acquired overseas exceeds US$120 billion in peak combined value. Australia generated US$63 billion in VC-backed exit value since 2020, ranking 8th worldwide, but the vast majority of that value accrued to foreign shareholders, exchanges, and tax jurisdictions.
The venture capital gap explains the exodus
Australia invests just 0.18% of GDP in venture capital, compared to 0.55% for the United States, 0.35–0.64% for the United Kingdom, and 0.77% for Israel. In absolute terms, Australian VC investment totalled approximately A$4.1–4.2 billion in 2024–25, while Silicon Valley and San Francisco alone attracted US$69 billion. Only 61% of early-stage funding in Australia comes from local sources, compared to 73% in Europe and 80% in the US. This dependency on foreign capital creates a gravitational pull toward overseas markets.
Australia's gross expenditure on R&D stands at just 1.7% of GDP, well below the 2.7% OECD average and less than half of Israel's extraordinary 6.35%. Business R&D spending is 0.89% of GDP versus the 1.9% OECD average. The R&D Tax Incentive program cost the government approximately A$4.6 billion per year, yet large business R&D investment has declined 24% ($2.9 billion) over the past decade. R&D costs in Australia run 12% above the OECD average, and the rate of industry-university R&D collaboration is the lowest in the OECD — just 9.5% of registered projects reported any collaboration. The BCA and Mandala Partners estimated that six targeted R&D reforms could unlock $7.72 billion in annual economic output at a cost of $1.41 billion per year, generating $5 of value for every $1 spent.
2. Australia has every ingredient except the right policies
Silicon Valley did not emerge spontaneously. It was engineered through deliberate government investment, university-industry coupling, immigration openness, and risk capital formation. Stanford's Frederick Terman placed students in startup pathways from the 1930s, and the Stanford Industrial Park (1951) incubated what became the world's densest innovation cluster. DARPA funded 70% of all US computer science research in the early 1960s, created ARPANET (which became the internet), and seeded technologies from GPS to voice recognition. NASA purchased 100,000 integrated circuits from Fairchild Semiconductor in 1964 alone, providing the demand signal that built the semiconductor industry. Every key component in the iPhone — touchscreen, GPS, Siri, internet, lithium-ion batteries — traces to US government-funded research.
The immigration factor is decisive: two-thirds (67%) of Silicon Valley tech workers are foreign-born, and 55% of America's billion-dollar unicorns have at least one immigrant founder. Australia has six universities in the global top 50 — an extraordinary density for a nation of 26 million — yet its innovation metrics are declining. Australian patent filings per billion dollars of GDP have fallen from 1.97 in 1990 to 1.35 in 2021, while the US figure grew from 3.14 to 3.7 over the same period.
Capital gains tax: the single biggest structural disadvantage
The policy gap that matters most is capital gains tax treatment of startup equity. Under the US Qualified Small Business Stock (Section 1202) provision, recently expanded by the One Big Beautiful Bill Act to a $15 million cap (inflation-indexed from 2027), startup founders can achieve a 100% exclusion of federal capital gains tax. A founder realising a $20 million gain pays zero federal CGT. Australia's best case is a 50% CGT discount applied to the top marginal rate of 47%, yielding an effective rate of approximately 23.5%. There is no Australian equivalent to QSBS. The UK's Business Asset Disposal Relief offers a 10% rate (rising to 18% by 2026/27) up to a £1 million lifetime limit, while its Enterprise Management Incentive scheme provides favourable treatment for employee options.
Employee share scheme taxation compounds the problem. Australia's default position taxes employees upfront on the discount to market value, even when shares cannot yet be sold. The 2015 startup concession improved matters — deferring tax for eligible companies under 10 years old with turnover below $50 million — but uptake remains minimal: only 200,000–300,000 ESS tax exemptions are sought annually, compared to the US option pool estimated at approximately A$1.9 trillion. The US combination of Incentive Stock Options, the 83(b) election (which allows founders to pay tax at nominal par value, converting all future appreciation to capital gains), and QSBS can result in an effective tax rate of zero percent on startup equity gains. The UK's EMI scheme achieves a 14% CGT rate with no income tax or National Insurance on exercise.
The $4.2 trillion superannuation failure
Australia's superannuation pool reached A$4.2 trillion by March 2025 — 150% of GDP, the 4th largest pension pool globally, and exceeding the entire ASX market capitalisation of approximately A$3.3 trillion. It is projected to reach A$8.1 trillion by 2035. Yet only 1.4% of super fund assets are allocated to private equity, with far less in venture capital specifically. Just 4.9% of total super assets go to private equity, venture capital, and private credit combined. The Australian Investment Council found this under-allocation is costing workers up to $20,000 each and the economy up to $54 billion.
The Your Future, Your Super reforms (July 2021) made this worse. The annual performance test uses an 8-year rolling benchmark, and funds underperforming by more than 50 basis points for two consecutive years are prohibited from accepting new members. This creates powerful incentives to herd around prescribed benchmarks and avoid VC's pronounced J-curve effect. Mandala Partners found these regulations are direct barriers contributing to under-allocation, despite private equity and venture capital achieving annualised 10-year returns 10.8 percentage points higher than Australian listed equity. The Future Fund, unconstrained by these regulations, allocated over 4% to venture capital and generated 23.3% returns on its VC/growth equity component.
The 2026 Denholm Review (Strategic Examination of R&D) delivered 20 recommendations including a 5% higher R&D tax incentive rate for qualifying startups, quarterly advance payments, and expansion of eligible expenses to include deployment and early commercialisation. The Productivity Commission's 2018 superannuation inquiry found a positive relationship between super fund returns and investment in alternatives. The Grattan Institute's "Back in Black?" report argued the CGT discount is "expensive, regressive and ineffective" for entrepreneurship — the top 10% of households receive nearly 75% of CGT discount benefits, and most entrepreneurs who claim it didn't know about it when starting their company.
3. Beer drinkers pay more tax than gas giants
In 2023–24, Australia's beer excise raised approximately $2.7 billion while the Petroleum Resource Rent Tax collected just $1.48 billion. Treasury Deputy Secretary Shane Johnson confirmed these figures to Senator David Pocock at Senate Estimates in February 2026. Senator Pocock asked: "How do we live in a country — one of the biggest gas exporters in the world — and we're getting more tax from beer than PRRT?"
The comparison has been consistent for years. In 2019–20, beer excise exceeded PRRT by approximately $1.7 billion. Even in 2021–22, when energy price spikes from Russia's invasion of Ukraine pushed PRRT to its recent peak of approximately $2 billion, beer excise still exceeded it by roughly $600 million. By 2027–28, beer excise is projected to be nearly three times PRRT revenue as the PRRT is forecast to decline to approximately $1 billion while beer excise keeps rising with biannual CPI indexation — a mechanism that has produced 75 automatic tax increases since Paul Keating introduced it in 1983. Australia's spirits excise rate alone stands at $107.99 per litre of alcohol, among the highest in the world. Total alcohol excise exceeds $7 billion annually — roughly five times total PRRT revenue. X
The PRRT is structurally broken
The PRRT is levied at 40% on "super-profits" from offshore oil and gas projects, but its deductible expenditure carry-forward provisions have rendered it largely ineffective. Undeducted capital and operating costs compound at the long-term bond rate plus 5% (previously plus 15% for exploration), effectively doubling deductions every four years. As oil industry analyst Juan Carlos Boué stated: "The uplift rate applicable to the unrecovered expenditure is so great that, with the miracle of compound interest, there is just no way for project net income to catch up, ever." By 2019–20, carried-forward PRRT expenditure credits had reached $282 billion.
The budget papers themselves state: "To date, not a single LNG project has paid any PRRT and many are not expected to pay significant amounts of PRRT until the 2030s." Chevron's Gorgon (US$54 billion cost) and Wheatstone projects paid zero PRRT until August 2025, when the new 90% deduction cap forced first payments. INPEX's Ichthys (US$47.7 billion) has paid zero PRRT ever and doesn't expect to pay until at least 2030. Shell's Prelude FLNG is projected to never pay PRRT — despite Shell originally claiming it would generate $12 billion in taxes over the project's life.
The company-level data is extraordinary. INPEX exported $21 billion in gas between 2015–2025 and paid zero PRRT, zero royalties, and less than $500 million in company tax across 11 years on $36 billion in total revenue — an effective rate of approximately 0.3%. In 2022–23, INPEX's two main entities recorded $20.4 billion in income with zero corporate tax and zero PRRT. ExxonMobil reported $15.5 billion in revenue in 2019–20 with zero taxable income. Collectively, Woodside, Exxon, Shell, Chevron, INPEX, and APLNG earned $34 billion in income in 2020–21 and paid $0 in corporate tax. In 2021–22, their combined income rose to $56.3 billion, yielding just $454 million in company tax — a 0.8% effective rate. The ATO has labelled the oil and gas industry a "systemic non-payer" of tax.
Australia's LNG exports were valued at $92.2 billion in 2022–23 (a record), totalling $265 billion over four years — of which $149 billion was royalty-free because over 56% of gas production capacity pays no royalties at all. Australia collects approximately 10–11.5% of petroleum resource value as government revenue, compared to Norway's approximately 66% and Qatar's approximately 66%. Norway's sovereign wealth fund stands at A$1.9 trillion (approximately $350,000 per citizen). Australia's Future Fund holds approximately A$250 billion. Qatar exports similar gas volumes to Australia but generates five to six times more government revenue from it.
4. Students repay 4.6 times more than gas companies pay in PRRT
In 2023–24, Australians repaid $5.1 billion in HECS-HELP debt through the tax system while gas companies paid just $1.1 billion in PRRT — meaning students paid 4.6 times more. In 2022–23, the ratio was 2.1 times ($4.9 billion vs $2.3 billion). Over the seven years to 2022–23, the government collected a total of $14,962 million more in HECS-HELP repayments than in PRRT — 168% more from students than from the resource rent tax. The Australia Institute also found that people with HECS debts paid at minimum $14.87 billion more in income tax than the gas industry in 2022–23.
As of 2023–24, 2.93 million Australians carried HECS-HELP debt totalling $81.05 billion, with an average debt of $27,640 — up from $26,494 the prior year. Twenty-four percent of all debtors now owe more than $40,000. The June 2023 indexation rate of 7.1%— the highest since 1990 — added approximately $5.3 billion to total student debts in a single year. A person with the then-average debt of $26,500 saw their balance increase by $1,882 overnight. The government subsequently capped indexation at the lower of CPI or the Wage Price Index (backdated to June 2023, removing approximately $3 billion), and applied a flat 20% debt reduction from June 2025, wiping approximately $16 billion. Combined, approximately $19–20 billion in student debt was removed through these reforms.
Richard Denniss, Executive Director of the Australia Institute, summarised the comparison at the 2024 National Press Club: "In Norway, they tax the fossil fuel industry and give kids free university education; in Australia we subsidise the fossil fuel industry and charge kids a fortune to go to university." Total Australian fossil fuel subsidies reached $14.9 billion in 2024–25, with the Federal Fuel Tax Credits Scheme alone costing $9.6 billion — more than the Royal Australian Air Force budget. A proposed 25% flat tax on gas exports, advocated by the ACTU and supported by several crossbench senators, would raise more than $17 billion annually — roughly twelve times current PRRT collections, and more than enough to eliminate all outstanding HECS-HELP debt within five years.
5. The ACCC has lost every contested merger case — and market concentration keeps worsening
The Australian Competition and Consumer Commission has produced world-leading inquiry work — particularly its 2019 Digital Platforms Inquiry — but its enforcement record tells a different story. The ACCC has lost every contested merger case in recent years, allowed market concentration to deepen across supermarkets, airlines, and banking, and imposed penalties on big tech that amount to rounding errors on quarterly revenue statements.
A merger enforcement record of consistent defeat
In ACCC v Vodafone Hutchison Australia [2020] FCA 117, the ACCC opposed the TPG-Vodafone merger, arguing that TPG was the best prospect for a fourth mobile network operator after spending $1.26 billion on spectrum. Justice Middleton accepted TPG executive chairman David Teoh's testimony that TPG had no intention of building a standalone mobile network, rejecting the ACCC's expert evidence on the counterfactual. The mobile market remains a three-player oligopoly (Telstra ~40%, Optus ~25%, TPG/Vodafone ~20%).
In ACCC v Pacific National (No 2) [2019] FCA 699, upheld on appeal at [2020] FCAFC 77, the ACCC challenged Pacific National's acquisition of Aurizon's Acacia Ridge rail terminal. The Full Federal Court found the acquisition would not contravene section 50 even without a behavioural undertaking, releasing Pacific National entirely. In the ANZ/Suncorp matter, the ACCC denied authorisation for ANZ's $4.9 billion acquisition of Suncorp Bank in August 2023, but the Australian Competition Tribunal overturned the decision in February 2024, granting unconditional authorisation — the first time the Tribunal had overturned an ACCC merger authorisation under the current regime.
Former ACCC Chair Rod Sims acknowledged the pattern directly: "Australia's economy is getting more concentrated over time and, in my view, is already too concentrated." He stated the merger losses "demonstrate the real difficulty of applying the substantial lessening of competition provisions" and that "we've got a merger law at the moment that is probably damaging the productivity of the economy." Treasury data confirmed the ACCC was notified of less than one-third of all acquisitions under the old voluntary system.
Landmark cases that exposed structural weaknesses
ACCC v Pfizer Australia [2018] FCAFC 78 illustrated why the 2017 Harper reforms were necessary. Pfizer launched its own generic atorvastatin (generic Lipitor, then Australia's biggest-selling drug at ~$700 million in annual sales) and offered pharmacies bundled rebates. The Full Federal Court found Pfizer had no prohibited anti-competitive purpose, even where internal documents used language about "blocking" competitors. The ACCC failed to plead the "effects" test for exclusive dealing, limiting its own case. The decision was made under old law; the 2017 reforms replaced the purpose-only test with a "purpose, effect, or likely effect" standard.
In ACCC v CG Berbatis Holdings [2003] HCA 18, the High Court dismissed the ACCC's unconscionable conduct claim 4-1, holding that inequality of bargaining power is not itself a "special disadvantage." Professor Bigwood described the majority's handling of the facts as "perfunctory" in the Melbourne University Law Review. In ACCC v Flight Centre [2016] HCA 49, the ACCC won a landmark price-fixing finding at the High Court (4-1), but only after losing at the Full Federal Court — succeeding on its third attempt. The $12.5 million penalty was modest for conduct spanning four years across three airlines.
Big tech penalties that amount to nothing
The ACCC's enforcement against digital platforms has been limited to consumer protection actions with trivial penalties. Google was fined $60 million for misleading Android users about location data collection (ACCC v Google LLC (No 4) [2022] FCA 942) — against a theoretical maximum exceeding $145 billion. The ACCC lost its case against Google over DoubleClick data consent in December 2022. Meta's subsidiaries paid just $20 million for misleading conduct regarding the Onavo Protect VPN app (ACCC v Meta Platforms Inc [2023] FCA 842), while Meta itself was dismissed from proceedings. The ACCC has brought zero competition law enforcement actions against any digital platform, despite its own Ad Tech Inquiry finding Google holds 90%+ dominance across the advertising technology supply chain. The ACCC conceded its "existing investigatory and enforcement powers are ill-suited to tackling systematic issues in dynamic ad tech markets."
The sectors the ACCC has failed to fix
In supermarkets, Woolworths and Coles hold a combined ~67% market share. The ACCC's 2025 Supermarket Inquiry (437 pages, 20 recommendations) found the sector "oligopolistic" with limited price competition and rising margins, but did not recommend breaking up the duopoly or any structural remedies. Shares in both companies rose after the report was released.
In aviation, the Qantas Group and Virgin Australia together service 98–99% of domestic passengers following the collapse of both Rex Airlines (July 2024) and Bonza (April 2024). ACCC data showed domestic airfares on major city routes increased 13.3% after Rex's exit, with specific routes spiking dramatically: Adelaide-Melbourne up 95%, Canberra-Gold Coast up 171%— despite jet fuel falling 41% over the same period. The ACCC's role is limited to monitoring and reporting.
New merger laws: a step forward with limits
The Treasury Laws Amendment (Mergers and Acquisitions Reform) Act 2024, passed in November 2024 and becoming mandatory from 1 January 2026, introduces mandatory pre-merger notification — aligning Australia with the US, EU, and UK for the first time. The ACCC becomes the first-instance decision-maker (replacing the judicial model), with Tribunal merits review available. Phase I review takes up to 30 business days; Phase II up to 90 additional days. Notification thresholds include combined Australian turnover above $200 million with target turnover above $50 million, and a serial acquisitions limb capturing cumulative acquisitions exceeding $50 million over three years.
ACCC Chair Gina Cass-Gottlieb called it "the most significant change to Australia's merger regime since the Trade Practices Act was enacted 50 years ago." However, the reforms do not change the substantial lessening of competition test itself, do not address existing market concentration in sectors like aviation or supermarkets, and face resourcing questions about the ACCC's capacity to review potentially hundreds of additional notifications annually. Australia still lacks sector-specific digital platform regulation equivalent to the EU's Digital Markets Act or the UK's Digital Markets, Competition and Consumers Act 2024 — despite the ACCC's own inquiry work recommending exactly this since 2019.
Conclusion: a pattern of structural policy failure
The data across all five dimensions points to the same conclusion: Australia's policy settings systematically undervalue innovation, undertax resource extraction, overtax individuals, and under-enforce competition. The country produces tech companies at the highest capital efficiency in the world — 1.5 unicorns per billion dollars invested — then drives them to Delaware through a CGT regime that offers founders a 23.5% effective rate versus America's 0%. A $4.2 trillion superannuation pool Globalii allocates a fraction of a percent to domestic venture capital, constrained by performance benchmarks that penalise the very asset class generating the highest long-term returns. Beer drinkers subsidise a gas industry that exported $265 billion in four years while paying an effective PRRT rate so low that students carrying $81 billion in collective debt repay 4.6 times more annually. And a competition regulator that loses every contested merger case presides over a 67% supermarket duopoly, a 99% airline duopoly, and digital platform dominance it openly admits it cannot address with existing tools.
The reforms underway — new merger laws, HECS indexation caps, the Denholm Review's R&D recommendations — are incremental responses to structural problems. Norway collects 66% of petroleum resource value and funds free university education. The US offers zero capital gains tax on qualified startup equity up to $15 million and attracts two-thirds of the world's venture capital. Israel spends 6.35% of GDP on R&D and provides direct 50% grants for startup costs. Australia's choices are not inevitable — they are policy decisions that can be reversed. The economic cost of inaction, conservatively estimated at $7.72 billion annually in lost R&D output alone, compounds with every founder who incorporates in Delaware instead of Sydney.