Strong on Research, Weak on Innovation: The SERD Report and the Boundary Between the Research System and the Innovation System

John H. Howard, 2 April, 2026

An abridged version of this Insight was published in InnovationAus on 1 April 2026.

This Innovation Insight follows on from The Strategic Examination of R&D: Can Australia’s Innovation System Reform Itself? published on 24 March, A Proposed National Innovation Council: Is Australia Chasing a Mirage? published on 27 March, and Lord of the Flywheels—SERD's Denholm Review Arrival. published on 31 March 2026.

Those earlier Insights examined the political economy of the SERD report’s prospects for implementation, including the structure of the proposed National Innovation Council, fiscal constraints, the complementarity problem embedded in its recommendations, and the “management chasm” between diagnosis and execution. This Insight turns to a different question: whether the report’s analytical framework is adequate to the task it sets itself.

The Conceptual Conflation

The Terms of Reference for the Strategic Examination of R&D (SERD) sought an examination of Australia’s “research and development system.” The report, however, presents itself as a reform of Australia’s “research, development and innovation (RD&I) system.” It proposes a National Innovation Council, organises its analysis around an “RD&I flywheel,” and frames its 20 recommendations as a package for national innovation reform.

This shift creates an obligation that the report does not fully meet.

In innovation studies, the science and research system is a component of the innovation system, albeit an important one, but the innovation system is a much broader concept. Research is one driver of innovation, but not the only one. In many sectors, it is not the primary one.

Innovation arises from multiple sources, including entrepreneurship, captured in the widely used term Innovation and Entrepreneurship, as well as breakthrough ideas in the practice of management, design and architecture, market formation, user engagement, socio-cultural and environmental contexts, and more broadly the capacity of firms and institutions to adopt and apply new knowledge - some of which may come from scientific research.

A report that proposes to reform the innovation system but analyses primarily the research system leaves a significant gap between its framing and its substance.

The report’s central organising device, the RD&I flywheel, illustrates the issue. It suggests a system in which inputs (funding, researchers, infrastructure) generate proportional outputs (knowledge, commercialisation, growth).

This framing essentially retains the essential logic of the linear model of innovation, widely attributed to Vannevar Bush’s 1945 report to President Roosevelt,  Science: The Endless Frontier, and criticised in the innovation literature for decades.

The report itself acknowledges that “breakthrough innovations rarely follow a linear and fast trajectory from discovery to commercialisation.” A stakeholder submission cited in the report observes that innovation “thrives in spaces that foster cross-sector collaboration” and requires “embracing non-linear innovation processes.” These insights, however, do not reshape the analytical framework.

The consequence is a report that is strong on the research system and its funding architecture, but markedly weaker on the broader innovation system that determines whether research investments translate into economic and social outcomes.

What the Report Does Well

Where the report addresses research system elements, it does so with considerable depth, albeit with some gaps.

Research funding receives detailed treatment: competitive grants, indirect cost reform benchmarked against the UK’s TRAC model, and indexation provisions that address long-standing concerns about the erosion of the funding base.

Research governance is redesigned through the proposed National Innovation Council, six sectoral pillars, National Strategy Advisory Councils, and National Strategic Initiatives. Research infrastructure is addressed through four sub-recommendations covering ongoing funding, coordination, rapid prototyping, and high-performance computing capability. The research workforce is examined thoroughly, from PhD reform and stipends through to migration settings and workforce planning.

The capital market reforms are also well developed: angel investment incentives, venture capital scheme expansion, superannuation settings, fund-of-funds mechanisms, and exit pathway reform. The RDTI reform package, differentiated across startup, SME, and corporate segments, addresses the tax incentive architecture in a level of detail that earlier reviews have not matched.

These are genuine strengths. The difficulty is that they are primarily strengths of a research and capital formation strategy. They address the supply of knowledge and funding. They are less attentive to the conditions under which knowledge and funding produce innovation.

What the Report Misses or Underweights

Management Capability and the Demand Side

The report does not substantively address management capability as a determinant of whether R&D investments translate into firm-level productivity. This is what I have described elsewhere as the “management chasm” (Howard, 2025b): the persistent gap between Australia’s research quality and its commercialisation performance, which has less to do with a shortage of ideas or funding and more to do with weaknesses in the capabilities needed to bring innovations to adoption and use.

The ARC’s submission to the SERD, cited in the report, makes the point directly: “no amount of supply side (government and university) ‘push’ can drive increased collaboration without corresponding engagement and capacity ‘pull through’ on the demand side.” This insight is noted but does not develop into an analysis of what builds demand-side capacity at the firm level.

The SME voucher scheme (Recommendation 6) connects firms with research institutions. It does not address the capabilities required for those connections to produce lasting change. The broader literature on absorptive capacity, beginning with Cohen & Levinthal (1990), suggests that firms require internal technical competence, management systems, and organisational routines to identify, assimilate, and apply external knowledge. Without these demand-side capabilities, even well-funded research systems produce knowledge that remains unused.

This is not a theoretical concern. Australia’s broad innovation adoption rates are well below OECD averages. The ABS Innovation Survey consistently shows that the majority of Australian businesses do not engage in innovation activity. A reform package directed primarily at the supply of knowledge addresses perhaps half of the problem.

Startups and the Business of Knowledge Transfer

The report’s treatment of commercialisation reinforces this supply-side orientation through a heavy emphasis on startups. In the SERD Final Report, startups receive dedicated analytical sections, including a proposed premium RDTI stream, angel investment incentives, and a central role within the National Strategic Initiatives. The startup ecosystem is treated, in effect, as the primary channel through which publicly funded research reaches the economy.

Startups are one pathway to market, and an important one. They are not the only pathway, and in many fields, they are not the dominant one. Research also reaches application and use through direct sales and licensing of intellectual property, options and assignments to established firms, contract and commissioned research undertaken in partnership with industry, and through the movement of knowledgeable graduates into employment in industry and government.

These channels, sometimes described collectively as the “business of knowledge transfer,” account for a substantial share of the economic return on public research investment. A framework that places startups at the centre of its commercialisation narrative underweights the broader institutional infrastructure through which knowledge is actually adopted, absorbed and applied.

Industry Structure and the Roots of Low BERD

The report documents the decline in business expenditure on R&D thoroughly: from 2.24% of GDP in 2008–09 to 1.69% in 2023–24, with BERD intensity falling 31% since 2009. It compares Australia unfavourably with OECD peers.

What the report does not do is investigate the underlying causes. It assumes the problem is primarily one of incentives and adjusts incentives accordingly. But the possible explanations for low BERD extend well beyond incentive settings. Australia’s industry structure is dominated by mining, finance, and property services, sectors with inherently lower R&D intensity than the advanced manufacturing and technology sectors that drive BERD in comparator economies.

Market scale also matters. Australia’s domestic market constrains the returns to R&D investment for firms that cannot readily export their innovations. Supply chain relationships may discourage local R&D where Australian firms are embedded in global supply chains managed from headquarters abroad. The privatisation of formerly research-intensive government business enterprises, including Telstra and Australian Defence Industries, removed significant performers from the BERD figures without creating private sector replacements of equivalent research intensity.

Each explanation implies a different policy response. If the binding constraint is industry structure, then R&D incentive reform will produce limited results without accompanying structural change in the economy and development and growth of new knowledge-intensive industries. A more searching diagnosis might have asked whether Australia’s industry mix is capable of absorbing significantly higher R&D investment, or whether industrial transformation is a complementary precondition for the R&D uplift the report seeks.

Arts, Design and Creative Practice

The report frames innovation almost exclusively through the lens of science and technology. The arts, design, and creative practice receive no substantive analytical treatment. This is a notable omission. International evidence demonstrates that design and aesthetic appeal, creative problem-solving, and artistic practice are integral to innovation, particularly in sectors where user experience, communication, and cultural context shape market adoption (Howard, 2008).

Countries such as Denmark, the United Kingdom, and South Korea have embedded design and creative capabilities within their national innovation strategies, recognising that technology alone is insufficient for products and services to meet human needs or gain market acceptance. The European Commission’s approach to innovation ecosystems explicitly integrates arts and humanities perspectives alongside STEM disciplines. Australia’s creative industries contribute meaningfully to GDP and employment, and function as a bridge between technical capability and market relevance.

The absence reinforces a narrow conception of innovation that privileges laboratory output over the broader process of bringing new value to market, and reflects the disciplinary narrowing that follows from treating the innovation system as synonymous with the research system.

AI as a Transformative Force in Research

The report mentions AI as an enabling technology alongside quantum and robotics, and as a sector warranting specific RDTI startup timeframes. These references are welcome, but they do not engage with AI as a general-purpose technology that may reshape the practice and economics of scientific discovery within the timeframe of the proposed reforms.

AI is already accelerating research in structural biology, materials science, drug discovery, and climate modelling. It is beginning to enable semi-autonomous laboratories in which AI systems design experiments, execute them via robotic platforms, and iterate on results with reduced human intervention. These developments raise issues about researcher training, workforce planning, research productivity, and the governance of AI-generated findings.

A strategic examination of the RD&I system might have been expected to assess how AI could alter the system it proposes to reform. Ambitious Australia lists AI among the technologies the system should support. It does not consider how AI might change the system itself. This is an omission that may prove consequential, given that the proposed reforms are set for a 10-year implementation horizon during which AI capabilities are widely expected to advance at a breakneck pace.

Place-Based Innovation and Regional Ecosystems

The report contains references to regions, precincts, and place-based initiatives. The proposed NSIs are expected to operate a “hub-and-spoke model” with centres of excellence across cities and regions, and state and territory governments are expected to contribute local assets. Place-based initiatives are mentioned as schemes that government partners might propose within the NSI framework.

This treatment is modest compared with the international evidence and practice. In Germany, the United Kingdom, and the United States, place-based innovation policy is a subject of strategic analysis in its own right. Innovation ecosystems are built at the regional, district, and precinct level, where research institutions, industry, and government interact in proximity, commonly referred to as the "triple helix effect". The dynamics of co-location, knowledge spillovers, and trust-based collaboration that drive innovation at these scales are well documented (Howard, 2025a).

Australia’s regions and districts currently face intertwined challenges of energy transition and economic diversification that call for precisely this kind of place-based innovation approach. The Terms of Reference explicitly required the panel to examine how R&D benefits could be equitably distributed across regions and communities. Place appears in the report as operational detail within the NSI framework rather than as a subject of strategic analysis. The regional equity dimension of the ToR is largely unaddressed.

Policy Instruments Beyond Funding

The recommendations rely heavily on funding, incentives, and program expansion. The procurement recommendation (Recommendation 15) is substantive, proposing an “if not, why not” principle for Australian RD&I procurement, challenge-based innovation procurement, and training for procurement professionals. This goes beyond the usual passing reference to government purchasing.

Beyond procurement, however, the broader regulatory toolkit receives limited systematic analysis. Regulation, legislation, licensing, place-based statutory planning, development controls, and conditional approvals can shape innovation behaviour as powerfully as funding does. Competition policy, intellectual property regimes (beyond university technology transfer), and environmental standards influence market formation and firm behaviour.

The recommendation to “drive cultural change” (Recommendation 20) through a “national narrative” is an aspiration without a mechanism. Culture in organisations, regions, and economies seldom shifts through exhortation. It evolves through processes of engagement, leadership, and change management that are well examined in the social sciences, including organisational development, institutional sociology, and political economy. These disciplines receive little attention in the report.

The Broader Pattern

The report’s analytical framework treats the innovation system as though it were coterminous with the research system. It proposes innovation system governance, through the National Innovation Council, on a research system diagnosis.

The six pillars are organised around sectors (health, agriculture, defence, environment and energy, resources, technology) rather than around the outcomes or capabilities that an innovation system is intended to produce. The flywheel metaphor, for all its visual appeal, suggests a mechanical relationship between inputs and outputs that the innovation literature has extensively critiqued.

The result is a set of research system recommendations that are thoroughly worked through and, in many cases, overdue, paired with an innovation system framing that the analysis does not adequately support.

Summary

None of this diminishes the value of the SERD report’s research system recommendations. The governance architecture, the funding reforms, the capital market package, and the workforce provisions each address real deficiencies in Australia’s R&D settings. Policymakers should engage with these recommendations on their merits.

The caution is that implementing these recommendations, even in full, would reform the research system but would not, by itself, reform the innovation system. The determinants of innovation that the report underweights, including management capability, absorptive capacity, demand-side dynamics, industry structure, creative and design capability, place-based ecosystems, and the broader regulatory toolkit, will need to be addressed through separate but complementary policy work.

The SERD report provides the research architecture. The innovation architecture remains to be built. As I have argued elsewhere (Howard, 2025b), structure is an enabler of strategy, not a substitute for it. The question is whether the governance architecture proposed is connected to a strategy that comprehends the full scope of what an innovation system requires.

References

Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128–152.

Denholm, R., Chubb, I., Wood, F., & Cornick, K. (2025). Ambitious Australia: Strategic Examination of R&D Final Report. Department of Industry, Science and Resources, Australian Government.

Howard, J. H. (2008). Between a hard rock and a soft space: Design, creative practice and innovation. Council for the Humanities, Arts and Social Sciences (CHASS), Occasional Paper No. 5.

Howard, J. H. (2025a). The handbook of innovation ecosystems: Placemaking, economics, business, and governance. Acton Institute Publishing.

Howard, J. H. (2025b). Thinking in public: Australia’s missing innovation policy. Will it ever be found? Acton Institute Publishing.

Howard, J. H. (2026). The Strategic Examination of R&D: Can Australia’s innovation system reform itself? Innovation Insight, 24 March 2026. Acton Institute for Policy Research and Innovation.

Organisation for Economic Co-operation and Development. (2015). The innovation imperative: Contributing to productivity, growth and well-being. OECD Publishing.

Universities Accord Panel. (2024). Australian Universities Accord Final Report. Department of Education, Australian Government.