The Definition of Industry Matters
The term ‘industry’ is used as a cornerstone of Australian innovation policy. It appears in every collaborative funding program, every roundtable, even the name of the Department. In policy settings industry engagement acts as a proxy for demand, pseudo evidence of research having relevance, and the outcomes will deliver meaningful impact. The term presumes everyone has the same interpretation of what industry is, and that mistake creates a fundamental problem.
If we use the banking industry as an example, most recognise there is more to the sector than the big four banks, though far fewer could name the companies that make the sector function: the core banking software providers, the payment processing platforms, the ATM network operators, regulators, the fraud detection systems, the data centre operators running the underlying infrastructure. While very few of these companies hold money, combined they are what makes the banking industry work. Banking is a relatively simple example. Many important industries in Australia are far more complex, with supply chains, technology layers and service ecosystems that make the visible operator at the top a small fraction of the whole.
The term industry can legitimately mean many different things depending on individual perspective, and most sectors in Australia are no different.
What “industry” ends up meaning
Settling on a single definition of industry is neither practical nor useful. Presuming it has shared meaning is a different problem entirely.
When policymakers import German or UK “industry-led” research collaboration models requiring an “industry partner,” they are referencing economies with the absorptive capacity to translate research into product. Manufacturing represents 17.8 per cent of Germany’s economy and 8.0 per cent of the UK’s. In Australia it sits at 5.4 per cent, below half the world average (World Bank Development Indicators). The entity answering to the word “industry” in an Australian funding room is far more likely to be a primary industry operator or financial service provider, with a completely different relationship to research. Australian industry policy does not lack impact because the model is wrong, but importing international models without considering our market structure will not deliver the same results in an economy like ours.
As we see when a collaborative funding program opens, the participants who turn up are generally researchers bringing along stakeholders within end-market operators. The recognisable names are generally large enough to co-fund research directly as a more cost-effective option than procuring solutions from the market. What is overlooked is the layer beneath; the suppliers, technology vendors, solution providers and system integrators. These are the Australian companies that could turn a research outcome into something a market can buy (see The Missing Bridge Builders). A generic use of the term industry quietly excludes an entire layer of our economy without anyone realising it.
The design flaw that follows
In sophisticated economies with serious research capability, collaborative funding programs require multiple industry partners, including naming partners with demonstrated commercialisation competency. The mandate is simple: show evidence of industry demand and name a local partner capable of producing the result. That binary requirement forces funding to reach the layer where commercialisation actually happens. You cannot access collaborative industry funding without naming the partner that will guide the innovation to market.
This is where Australian industry policy diverges. Our collaborative funding requires a research partner and industry partner. No local supplier needs to be named, and no commercialisation pathway needs to be evidenced. While historically this has not been a major factor, most of our large-scale technology companies have been acquired and taken offshore. The smaller number of MSEs remaining lack the incentives and patient capital to commit to collaborative research timelines.
The outcome is a funding model where money flows to research, but local companies who could commercialise it are bypassed entirely. This is not a trivial administrative issue. The structural gap has been inadvertently embedded in the design of innovation funding, and it follows directly from never asking who “industry” is meant to include.
Where the value goes
The consequences are far greater than a missed opportunity. The value does not simply fail to materialise; it leaves the country.
Primary industry contributions to research are critically important, but they do not retain the outcomes. Once a new method, technique or process is defined, they incorporate the requirement into their engineering and procurement contracts. This is a subtle but important nuance. They do not commercialise innovation, they procure it. And by doing so they do not encourage startups, they standardise suppliers. They seek vendors with the scale to respond and integrate new technologies, which tends towards large international technology firms who invest in the underlying intellectual property.
Australian publicly funded research becomes a source of competitive advantage owned offshore, sold back to our local industry (The Economic Stakes Are Too High). This outflow is not bad luck or poor execution, but an entirely predictable outcome of a funding model with no mechanism to support local commercialisation.
Why the pattern holds
Once the problem becomes apparent, the obvious next question is how it has survived for so long. The answer is structural and worth stating carefully, because it is more about incentives than intent.
Each review of Australia’s R&D system (including SERD) has examined programs, funding models and governance structures. No one has considered the assumption sitting underneath what industry means for collaborative research.
Research organisations depend on industry partners who can co-invest. End operators have far deeper pockets than the small and medium companies that make up most of the supplier ecosystem. End markets bring scale, certainty and large cheques, but none of the commercial pressure a vendor lives with. Their survival does not depend on building something a market will buy, they are simply the easiest counterpart.
Nobody is acting in bad faith; incentives have biased towards the structure of our economy, which contains the inherent gap. Everyone is responding rationally within the structure they operate. This is not something that will correct itself through goodwill alone.
So what is the point?
What each stakeholder means by industry is not a question of semantics, it is foundational to the design of effective innovation policy. To presume common understanding when designing policy sets the system up to fail in a very predictable way.
When industry inadvertently means end markets, programs are blind to the layer where translation happens. ACIL Allen highlighted what the alternative looks like in their evaluation of the METS Industry Growth Centre. In their specific collaborative project funds, innovators were required to bring matched funding from commercial partners (see Middle Australia). This simple fix connects the translation layer by design, where currently it never shows up.
We must consider more carefully what each stakeholder means by industry. Otherwise, we will keep funding research but lose the value, making the same decisions that continue to deliver the same results.