Collaboration, co-production and co-design in policy advice
Collaboration, co-production and co-design are now so common in Australian public policy that the terms are often used interchangeably. That blurring matters. It lets almost any form of 'working together' be called co-design, and it hides real differences in who holds power and whose knowledge counts.
In a chapter for the Handbook of Policy Advice (2025), Colette Einfeld and I set out to untangle these three ‘cos’. We asked a specific question: what do they offer policy advisory systems – the networks of people and institutions that supply knowledge to policymakers? Our short answer is that they open those systems to new kinds of knowledge, and in doing so they change what it means to be a policy advisor.
Collaboration: working across boundaries
Collaboration is the broadest of the three. Drawing on Janine O’Flynn’s work, we place it at the top of a continuum of ‘working together’: networking, then coordination, then cooperation, then collaboration. Each step asks for more trust, commitment and formality. Collaboration is the most demanding and the longest-lasting.
Governments collaborate because they have to. Complex, interdependent problems like climate change cross jurisdictions. Services are delivered through hybrid arrangements with non-government organisations. Publics are more diverse. Yet the evidence on the outcomes of collaboration is thin, and it is sometimes a last resort rather than a first choice.
COVID-19 showed both sides. Rapid vaccine development depended on collaboration across borders and sectors. But, as Sabelo Ndlovu-Gatsheni has argued, African nations were largely shut out, despite deep experience in managing epidemics. Collaboration reflects existing hierarchies of power and knowledge unless someone deliberately disrupts them.
Co-production: of services and of knowledge
Co-production has two dominant meanings. The first, which grew out of Elinor Ostrom's work, is about services: students co-produce their learning, carers co-produce health care. This kind of co-production can happen without anyone intending it.
The second meaning is about knowledge. From science and technology studies, Sheila Jasanoff’s insight is that knowledge and governance shape each other – science produces policy problems, and policy problems drive science. In practice, this includes researchers and practitioners deliberately co-creating knowledge, sometimes through creative methods. One example we discuss is Camilla Adelle and colleagues’ use of digital storytelling to bring marginalised voices into food governance in South Africa.
Across the many meanings of co-production, the motivation is consistent: current practice isn't working, and multiple forms of knowledge are needed.
Co-design: an intentional, design-led process
Co-design is the most tightly defined of the three, though still blurry in use. It draws on participatory design, design thinking and human-centred design, and uses creative methods and participatory principles to engage community members and professionals in addressing policy problems. It is iterative and design-led, using generative methods to surface less explicit needs and new options. It needs skilled facilitation – the right spaces, scaffolding and invitations for people to be creative and work across difference.
In one New Zealand example, two road safety agencies commissioned a co-design process with community members, professionals and ‘creative provocateurs’. Beyond the behaviour change campaign they were asked for, the group proposed systemic changes to driver licensing and education, some of which were later implemented.
Co-design treats people as experts in their own lives. Lived experience, local knowledge, and Indigenous and traditional knowledge systems carry real weight, not token input. That makes co-design explicitly normative: it is about sharing power. When professionals won’t let go of control, what results is better described as ‘faux design’, to borrow Jo Szczepanska’s term. Co-design is also about effectiveness – what Penny Hagen, Ingrid Burkett and colleagues call a shift towards ‘practice-based evidence’, built by testing and prototyping with communities in context.
Three ways to tell them apart
We compare the approaches on three dimensions.
Intentionality. Collaboration is deliberate. Co-production can simply happen, sometimes involuntarily. Co-design is always intentional: even when the solution is unknown at the outset, there is a shared purpose that brings co-designers together.
Power. All three can disrupt or entrench power hierarchies, depending on who is invited to the table and what happens there. When it comes to public policy, government usually defines the problem, scope and process, and chooses who participates. Critics of so-called co-design in Australian Indigenous policy have put it bluntly: at worst, the only trace of co-design is the Aboriginal art on a report's cover.
Actors. Collaboration can happen entirely within government. Co-production and co-design always involve people outside it. Co-design typically brings together public servants, community organisations and people with lived experience, usually led by a professional facilitator or designer.
What this means for policy advice
The traditional picture of policy advice is top-down: experts produce knowledge, advisors carry it to decision-makers. The ‘co’ approaches challenge that picture by bringing three kinds of knowledge into play:
local knowledge – the often tacit understanding that grows from people's daily lives and lived experience
technical knowledge – professional, scholarly and scientific expertise
craft knowledge – public servants’ practical know-how about government and the policy process, including stewardship, diplomacy and political nous.
The most interesting combination is craft and local knowledge. In a co-design process, public servants don’t just receive recommendations. They carry what they have learnt into other policy processes, at the times and in the forms that make it usable. Citizens and public servants both become policy advisors.
This shifts the advisor’s role from passing knowledge along to facilitating how it is generated. That is a significant change in capability, and it runs up against bureaucratic procedures and project timeframes that constrain how deeply people can participate.
Questions worth asking
If you are commissioning or leading work described as collaboration, co-production or co-design, the chapter suggests some useful tests:
Which of the three is this, really – and does the label match the practice?
Who defined the problem, and could the process reframe it?
Whose knowledge is being treated as evidence, and whose as input?
What would it take for the public servants involved to carry this knowledge into other decisions?
These approaches have real potential to disrupt existing power hierarchies – if they are thoroughly applied.
SOURCE
Einfeld, C. and Blomkamp, E. (2025) 'Co-production, collaboration and co-design as sources of policy advice', in M. Howlett and I. Mukherjee (eds), Handbook of Policy Advice. Cheltenham: Edward Elgar, pp. 638–648. https://doi.org/10.4337/9781035318087
