Australia stands to lose proportionally more than the United States if it replicates America's restrictive stance on scientific cooperation with China, according to an analysis published by Crikey.

The argument lands at a moment when research-security policy is moving from discussion paper to legislation in both countries. For R&D managers at Australian universities and publicly funded institutes, the stakes are concrete: China is one of Australia's largest research partners by co-authorship, and restrictions on that collaboration would touch grant pipelines, student recruitment and ongoing lab-to-lab projects.

The asymmetry at the heart of the analysis is structural. The United States, with a research system roughly an order of magnitude larger by expenditure, can absorb the loss of Chinese collaboration across a broad portfolio of partners. Australia's system is smaller and more concentrated. A comparatively small number of institutions — the Group of Eight universities chief among them — carry a disproportionate share of international co-publication with Chinese partners. When a large share of your collaborative output runs through one partner, severing that connection cuts deeper than it would for a system with many equivalent partners.

The comparison matters because Australian policy has tended to track Washington's lead on research security. US measures have expanded visa scrutiny for Chinese researchers, restricted technology transfer and pushed universities to disclose foreign funding arrangements. Canberra has followed with its own foreign-interference frameworks and defence-export control reviews. The Crikey analysis argues that importing the American posture without adjusting for Australia's different weight class would impose costs the US simply does not face.

For portfolio planning, the exposure runs through several channels. Co-authored publications with Chinese institutions form a significant share of output in fields where Australia has concentrated strengths — materials science, engineering and agricultural science among them. Joint doctoral supervision and the flow of Chinese doctoral students into Australian laboratories support research capacity that domestic funding alone does not cover. Equipment, reagents and sample flows between labs in both countries would also need review under stricter regimes.

The analysis treats claims from all sides with appropriate scepticism. Proponents of decoupling argue that security risks in dual-use research outweigh collaboration benefits; critics counter that the quantified damage from espionage cases is small relative to the measured output of joint work. Neither side has published a full cost-benefit accounting, and the Crikey piece positions itself against adopting restrictions on the strength of alliance solidarity alone.

What the analysis does not do is dismiss security concerns outright. The case it makes is comparative: whatever the merits of restricting Chinese collaboration, Australia's cost-benefit calculus differs from America's because its research system has fewer alternative partners, less redundancy and a higher dependence on international students and co-funding. A policy calibrated for the US ecosystem will not be calibrated for Australia's.

The immediate audience for this argument is the government's research-security review process, which must decide how far Australian law follows US export controls and disclosure regimes. University research offices will be watching for whether new rules mandate disclosure only, or reach into approval requirements for individual collaborations.

Expect the debate to sharpen as Canberra finalises its next round of research-security guidance.