Government, universities and infrastructure: the institutions around Australian startups

Public intervention was a design question, not a binary choice

Kate Jones rejects the idea that government should simply stand back. Her qualification is the important part: an active role still has to be designed well. The Queensland Government's 2022–2032 innovation roadmap records that Advance Queensland launched in 2015 with stated goals around economic diversification, regional growth and jobs. That establishes chronology and intent, not success. Together, the sources frame the public role as a question of what government should do, through which institutions, and with what discipline.

“Government has a role. I don't think it should be hands-off. I think it should be hands-on, but it has to be hands-on in the right way.”

Programs could assemble support — or multiply fragmentation

Nicole O'Brien describes government as one actor in a wider system of investors, incubators, accelerators, talent and funding. Advance Queensland's current Ignite program makes that mixed offer concrete: it describes three stages from early concepts through market entry and growth, combining funding with mentoring, networking and professional development. Aaron Birkby supplies the counterargument. In his view, new funding buckets could generate new programs without resolving collaboration. These are not mutually exclusive findings: a program can assemble useful support while a larger field of programs remains fragmented.

“So, you know, the role of government is really critical in fostering that and creating the ecosystem, really contributing to the ecosystem, because it takes a whole community of ecosystem players to really raise a, you know, a successful startup.”

“And anytime there's a new funding bucket, we all spin up a new program to do it. And we lack the collaboration piece.”

Universities were asked to become more entrepreneurial

The Office of the Chief Scientist released Boosting High-Impact Entrepreneurship in Australia on 30 October 2015 and identifies Colin Kinner as its author. The official record also preserves the then Chief Scientist's view that universities should be central to building an entrepreneurial culture. Kinner remembers the benchmarking behind the report more sharply: some Australian universities were doing useful work, but he believed the sector remained well behind overseas examples. The date matters. This is evidence of the problem as framed in 2015, not a current ranking of universities.

“So that report, which was released in late 2015, looked at the data and looked at the international benchmarks around the world and came to the conclusion that although we had some universities doing some good stuff, we were actually a very long way behind.”

Research funding and commercialisation solved different problems

Malcolm Turnbull separates the production of knowledge from the financing of companies. His preference is for government to put more into pure and deep research, particularly through universities and CSIRO. Commercialisation then raises a different institutional problem: how businesses and universities work together around research expertise, equipment and intellectual property. The Australian Government's voluntary Higher Education Research Commercialisation IP Framework says that work can be complex and supplies step-by-step guidance and standardised agreements. The framework records an attempt to reduce process friction; it does not establish adoption or better commercial outcomes.

“Where the government should be putting more money is in pure research. You know, I do think we need to spend more on pure science, deep science, if you like. And that's obviously mostly done by universities, CSIRO, and so forth.”

Precincts joined physical space to relationship infrastructure

Yolanda Redrup describes hubs and specialised precincts as part of the architecture around startup activity. Official records show what governments put inside that idea. Victoria's 2020 Cremorne plan proposed a hub built around shared space, equipment, skills and collaboration. The current Tech Central Innovation Hub describes flexible workspace, events, mentors, programs and ecosystem connections. These records illustrate a model in which infrastructure is both physical and relational. They do not show how widely the facilities were used, who could afford access or whether the precincts caused investment and growth.

“Then you needed to have the sort of architecture around it, I guess. So having the little hubs, you've got a thriving community around Cremorne and Richmond in Victoria, and you've got a health precinct, which is well known around Parkville. And then obviously in Sydney, great software and fintech hub up there.”

Government also shaped the market through rules and demand

The interview evidence in this section is stronger on policy continuity than on contracting outcomes. James Alexander argues that a startup agenda needs a durable institutional home capable of carrying it across governments. The current Commonwealth Procurement Rules show another part of the state's role: they govern how entities buy goods and services, require value for money, transparency and accountability, and prioritise Australian businesses and small and medium enterprises for some procurements. Those rules are not startup-specific and do not prove that a young company can win a contract or that the wider policy agenda will remain consistent.

“And so where I think it could do a much better job is supporting a consistent agenda over multiple different governments. And that's really difficult because it needs to have an agency. And that's what we lack in Australia.”

Regional policy tried to create more local entry points

James Fielding remembers startup opportunity once being concentrated in a small number of recognised centres, then seeing local and state-backed technology parks create more visible avenues elsewhere. Advance Queensland's current Regional Enablers Program offers a contemporary example of that institutional logic: it funds established organisations outside Brisbane to expand innovation services, strengthen local ecosystems and improve collaborative networks. The official page establishes purpose and eligibility, not whether regional access is sufficient, representative or equal.

“And if you were anywhere else, then you weren't really on the map. And what's happened over the last 5 or 6 years is these little pockets have opened up, these technology parks. There's been local government ones, state government ones, lots of people trying to create an avenue for people to bring their ideas to life.”

Institutions became less useful when their roles blurred

Alex Carpenter's criticism is about role confusion. In her view, governments, universities, funds and schools should operate at their strengths and partner rather than duplicating one another. A 2018 departmental evaluation of the Incubator Support initiative offers a different but compatible lens: it found early implementation had progressed well while identifying room for improvement, and deliberately limited its scope to program need, design and initial delivery rather than long-term impact. Carpenter supplies a design principle; the evaluation shows why implementation and measurement still need scrutiny. Neither source establishes one universal division of institutional work.

“I think I'm a huge fan of everyone doing their thing, and it infuriates me so much when, when, like, the government tries to do something that, you know, the university should be doing, and the university tries to do something that the fund should be doing, and the schools try to do something that the government should be doing. And it's like, guys, like, we need everyone playing at their absolute optimal point in order to make this really move. And so we've got to be more collaborative and do more partnerships.”