Beyond the Lab: Where innovation actually happens

Paul Egglestone
3 days ago
6 min read

If you listened only to the way we talk about it, you’d think most innovation comes from universities and their spin-outs. Heroic professors, plucky PhD founders, shiny new research precincts with inspirational banners about “ideas to impact”.

The problem is: the data doesn’t back that story up. And once you widen the lens beyond STEM and startups to social innovation, the ways we change how care, energy, education and local economies actually work, the picture gets even more lopsided.

This matters for FASTlab because we spend most of our time in the gaps between sectors: universities, industry, government, and communities. If we keep telling ourselves the wrong story about where innovation lives, we’ll keep investing in the wrong places and measuring the wrong things.

What the numbers actually say

Let’s start with the boring bit: R&D statistics. Because whatever we believe about innovation, public money and policy are driven by these numbers.

Across the OECD, the business sector now performs about 74% of all R&D spending; in the EU it’s about 66%. Universities, public labs and non-profits share the remainder. OECD

In the US, it’s even starker. In 2021, business accounted for 77% of national R&D, with universities playing a much smaller role overall – though they remain the largest performers of basic research. ncses.nsf.gov+1

Australia is unusual in that our universities are relatively big R&D players – but even here, the story is the same:

The business sector has the highest nominal expenditure on R&D,
followed by higher education, then government, then private non-profits. Industry, Science and Resources+1

Business R&D has roughly quadrupled in nominal terms since 2001, and still represents the biggest chunk of our innovation “inputs”. Industry, Science and Resources

So if you follow the money, you land somewhere obvious but often glossed over:

Most of the stuff we would recognise as “innovation work” is being done inside firms, not universities.

Universities are still critical, especially for basic science, talent development and some foundational IP, but on sheer volume of activity, they’re not the main event.

Patents tell the same story

If you swap R&D spending for patents (a very imperfect proxy, but the one policymakers love), you get a similar pattern.

The latest WIPO Patent Cooperation Treaty review shows that, in 2023, businesses accounted for 88.3% of international patent applications, with the university sector contributing 5.9% and government/public research organisations 1.8%. WIPO

Even in high-income countries where universities are pretty active in IP (Australia is one of them), they still represent around a tenth – not half – of patenting activity. WIPO

Again: that doesn’t make universities unimportant. But it does mean that if we equate “innovation” with “university commercialisation”, we’re staring at one small corner of the system and pretending it’s the whole room.

The invisible giant: social innovation and the voluntary sector

Now zoom out further.

Most of the innovations that actually change people’s lives – especially in health, care, education, local economies and climate adaptation – don’t show up neatly as patents or R&D line items. They show up as:

new models of care
new forms of local energy ownership
new ways of organising services
new practices in schools, councils, communities

And a huge amount of that work is led by civil society: NGOs, social enterprises, community organisations, volunteer groups.

The social innovation literature is pretty blunt about this. Social innovation is explicitly defined as creating new solutions to major social and ecological challenges, and the inclusion of civil society as an actor significantly increases the chances that these innovations are adopted and spread. SpringerLink

Work linking social innovation and social entrepreneurship makes the same point: the voluntary and community sector isn’t an afterthought – it is often the primary arena where new models are tested, adapted and made real. ScienceDirect

But because our standard indicators were built for STEM heavy industries – counting patents, R&D dollars and high-tech exports – most of this social and civic innovation is effectively invisible in the headline stats.

So we’ve ended up with a strange situation:

Measured innovation is dominated by business, with universities as important but minority players.
Unmeasured social innovation is dominated by civil society, but barely recognised in policy dashboards.
And yet an enormous amount of public debate still behaves as if innovation = “universities + startups”.

Why our story is out of sync with reality

How did we get here? A few reasons.

Universities are visible and organised They have rankings, strategies, communications teams and big physical campuses. It’s easy to build a national innovation narrative around institutions that already know how to talk about themselves.
Metrics shape imagination We measure what’s easy: R&D spend by sector, patent counts, spin-out numbers. Once those graphs are on the minister’s slide deck, we start to believe they are innovation, not just a narrow slice of it.
The time lag problem Universities specialise in long-horizon foundational work. The real social and economic gains often show up decades later, in entirely different sectors and geographies. That makes it very tempting for governments to demand quick, visible wins in the form of spin-outs, licensing agreements and incubators – whether or not that’s the best use of the underlying research.
The voluntary sector is structurally under-resourced NGOs and community organisations are busy keeping the lights on, not hiring policy teams to tell a grand innovation story. Their work often sits in project-based funding, with little support to codify, scale and measure their innovations in the language governments recognise.
We’ve mistaken “commercialisation” for “innovation” Selling more things based on existing IP is part of innovation, but it isn’t the whole. Changing how aged care residents are treated, how energy is owned in a suburb, or how young people participate in local planning can be every bit as innovative – and arguably far more important in the long run – than the next app or platform.

What this means for Australia – and for places like FASTlab

We hear the phrase 'Evidence based' or 'data informed' decision making rolled out by policy makers almost every day. If that's ever to be anymore than an empty mantra - and seemingly smart people start to take the evidence seriously, a few implications follow.

1. Stop pretending universities are the main engines of innovation

They are essential nodes – especially for basic research and talent – but not the main volume producer of new products, services and processes. Business is doing that heavy lifting, and civil society is doing a lot of the social re-wiring.

For Australia, the data is clear: the business sector is the largest performer of R&D and has seen the strongest long-term growth. Industry, Science and Resources+1

So if our innovation policies, funding programs and public conversations centre universities and treat industry and communities as “translation partners” at the end of the pipeline, we’ve got it backwards.

2. Treat civil society as a first-class innovation partner

If civil society and the voluntary sector are critical for social innovation and for the adoption of new practices, they need to be designed into the system, not bolted on as “consultation”.

That means:

funding models that back community-led experiments, not just university-led trials
metrics that can see social innovations in care, energy, youth engagement and local economies
governance that gives communities real say over the design and ownership of solutions (think mutuals, community energy, youth assemblies – not just glossy engagement reports)

3. Build institutional capability between discovery and adoption

This is the bit FASTlab keeps running into.

Most of the real work sits in the messy middle: taking good ideas (from labs, communities or firms) and making them:

adoptable by real services
fundable in sensible ways
governable so that risk, benefit and control are shared fairly

That’s not “more founder training”. It’s building professional capability in business builders, civic entrepreneurs and place-based coalitions who can shepherd innovations from promising ideas to embedded practice.

Our work on things like.....

community energy mutuals (GEN)
Healthy@Home and Meaningful Measures for ageing and care
Education programs like Sustainnovation Challenge and Net Z[Ed}

…is basically an ongoing experiment in exactly that middle space: transdisciplinary teams, shared governance, new funding models, and creative practice as a way of making change tangible.

None of that fits neatly into a patent box. But it absolutely is innovation.

So what should we be asking for?

If I had to boil the argument down for policymakers, university leaders and industry partners, I’d put it like this:

Align the story with the evidence. Business already performs the bulk of R&D. Universities perform most basic research. Civil society drives adoption in social systems. Let’s stop pretending one sector can do it all, and design policies that reflect the real division of labour. SpringerLink+3OECD+3ncses.nsf.gov+3
Upgrade the metrics. Keep measuring R&D and patents – they matter – but add serious, resourced measures for social and organisational innovation: new care models, new energy ownership structures, new youth participation mechanisms, new partnership architectures.
Fund the middle. Invest not only in “research excellence” and “startup ecosystems”, but in the boring, difficult work of translation: cross-sector teams, mutual structures, community-led pilots, shared playbooks. That’s where good ideas go to live or die.
Treat voluntary and community organisations as co-designers, not consultees. If the social innovation literature is clear that involving civil society increases the chances of adoption, then build that into how programs are designed, governed and evaluated from day one. SpringerLink+1

At FASTlab, we’re unapologetically biased: we think some of the most important innovation of the next decade will happen in exactly these in-between spaces – where universities, businesses and communities meet to re-design energy, care, education and local economies together.