In this post, it gives me great pleasure to introduce a colleague whose work I’m very excited about. Heidi Peterson has been working in the Value for Investment space for some time now, including innovative work on the UK’s Global Challenges Research Fund (Peterson, 2022) that explored how to make value for money assessments more meaningful in complex, equity-focused research partnerships. Her new article on democratic deliberation as a “north star” for assessing the value for money of systems-change efforts (Peterson, 2025), and her PhD research, takes that work deeper into the systems transformation space.
Heidi is pushing VfI to do more of what it was intended for in complex settings: make evaluative reasoning explicit, bring mixed methods together, and involve stakeholders in judging what “good” looks like. In the rest of this post, she explores four things:
Why democratic deliberation is her north star for this work
How rubrics can scaffold that deliberation without collapsing complexity
How we can expand the types of value we can bring into a VfI assessment of systems-change work
What this means for the evaluator’s role, given real-world constraints on time, people and power.
Over to you, Heidi - starting with a brief introduction to your background and how you came into this work.
A quick introduction
I first came across VfM when I moved to the UK to work on some of the country’s largest research-for-development funds – £2.2 billion of public investment that was under a lot of scrutiny for not having an approach to VfM. This was also how I first met Julian, when I desperately emailed him for help on how to assess the VfM of such large funds. This challenge set me on a path to grappling with how we can assess value meaningfully in complex initiatives. I started a PhD (with Julian as one of my supervisors) while still in the UK to give myself the time and space to really wrestle with it, which I have just finished!
Over the course of my PhD, the pressure on governments to demonstrate good and efficient use of public funds has continued to increase. But at the same time, there has been a growing recognition of the complex and wicked problems facing society, leading to more and more systems-change initiatives that seek to change the conditions holding these problems in place. Evaluators asked to assess the VfM of such initiatives can feel caught between a rock and a hard place of efficiency and complexity (certainly I have!). The usual methods for VfM struggle with the unpredictable, emergent and non-linear outcomes of messy systems-change work, where attribution is impossible and time horizons stretch long.
However, the VfI approach, grounded in the principles of deliberative democratic evaluation, can provide a way forward – so let’s jump into the rest of the post.
Why democratic deliberation is my north star
When we evaluate systems-change work, we step into contested terrain. Beneath disagreements about the VfM of an initiative lie deeper debates: what are the problems in the first place? What types of systems change should we be striving for? And whose perspectives – of the system, and of what counts as valuable – actually matter?
Participatory evaluation is often put forwards, and rightly so, as a way to work through the different perspectives that underpin systems-change work. In this context, however, it is not enough to say that our VfM evaluation is participatory because we interviewed stakeholders or ran a workshop. Participation can still leave the underlying valuing work untouched.
Democratic deliberation, on the other hand, asks more of us. In deliberative democratic evaluation (House and Howe, 2003), three principles are central:
Inclusion: all relevant perspectives, especially those most affected, should have a meaningful opportunity to shape how value is understood and judged.
Dialogue: stakeholders should have space to articulate their views, hear others, and interrogate assumptions.
Deliberation: the group works towards shared judgements about merit, worth or value through reason-giving and negotiation, not by aggregating individual scores.
For systems-change VfM, these principles are my north star – the principles I strive for. That doesn’t necessarily mean that I put them into practice in every project – time, budget and politics almost always get in the way – but they orient the decisions I make throughout an evaluation. They push me to ask, for example:
Whose perspectives must be in the room if this judgement is to be legitimate?
What power dynamics might get in the way of meaningful involvement in criteria-setting and judgement-making?
What kinds of evidence and lived experience will we treat as credible, and why?
How will we handle situations where commissioners, implementers and communities hold genuinely different ideas of value?
Crucially, democratic deliberation changes how we think about “who decides”. Instead of an evaluator or a small group of technical experts determining whether an initiative offers VfM, judgement making becomes something we do alongside those who have a stake in the answer.
In systems-change work, where no single actor can see or understand the messiness and complexity of the whole system, I don’t believe there is any better way to proceed.
A rubric as the scaffolding for democratic deliberation
Rubrics are sometimes dismissed as glorified rating scales. I see them differently. For me, a rubric is the scaffolding that allows democratic deliberation about value to happen in an explicit and transparent way.
Others have written elsewhere about what rubrics are and how we build them, so I won’t go into definitions here. But I want to highlight three things that rubrics enable when we are trying to reason together about VfM in complex systems.
First, rubrics make our thinking visible. When we articulate criteria (what matters) and standards (what “good” looks like), we’re putting our ideas about value on the table where others can see, question, and refine them. That’s a precondition for any meaningful deliberation and allows the rubric to become the shared artefact that anchors the group’s evaluative reasoning.
Second, rubrics give structure to the work of reaching judgement. Rather than an unbounded discussion about whether an initiative is “good value”, the group moves criterion by criterion, standard by standard, considering the evidence and reasoning in a disciplined way towards a final judgement.
Third, rubrics create a shared language for disagreement. In a deliberative panel, it is much easier – and safer – to say “I’m not convinced we’ve met this standard under the criterion about contribution to systems change” than to say “I just don’t think this program is very good”. The rubric focuses contention on the criteria and standards, not on the people in the room.
How this plays out in practice: WorkWell
The WorkWell program in Victoria, Australia, provides a concrete example of this approach in action. WorkWell was a large, multi-year systems-change initiative aimed at improving workplace mental health, with a focus on shifting employers from individual resilience strategies towards primary prevention.
In the final evaluation, we used a VfI-inspired, rubric-based process anchored in five types of value (more on that below!) and supported by a deliberative panel of stakeholders. Because there is a journal article that goes into WorkWell in detail, I won’t unpack it here, except to use it to illustrate the value types below. If you are interested in how the framework and deliberative process played out in practice, you can read that case study in full here.
Five types of value: reframing what counts as “good use of resources”
One of the biggest challenges in assessing the VfM of systems-change initiatives is that much of the value they create isn’t visible in typical funding cycles. If we only look for short-term, attributable outcomes, we risk seriously underestimating – or even mischaracterising – what an initiative has actually achieved.
To address this, I developed a schema that combines the Water of Systems Change framework (Kania et al. 2018) with the Cycles of Value Creation from Wenger and colleagues. The Water of Systems Change argues that systems change occurs when the conditions holding a problem in place shift – moving from explicit conditions like policies and resource flows, through relational conditions like relationships and power dynamics, down to the most implicit and transformative: mental models.
The Cycles of Value Creation was originally developed for social learning networks – not VfM at all – but when I came across it, I realised it could be overlaid with the systems-change framework to give us something genuinely useful. Together, they produce five types of value for systems-change work:
Inherent value – value in the activities themselves.
This is the immediate value that comes from doing the work – evident even before anything else shifts. For WorkWell, it included the community and connection created for the 30,000 workplace leaders who engaged with the program, and measurable improvements to worker mental wellbeing in several funded projects.Potential value – enablers for future systems change.
This is the type I find most critical – and most often overlooked. It’s what happens when you change relationships, connections, and power dynamics: you’re not yet seeing outcomes, but you’re creating the conditions for them. In systems change, this loading of potential can be a long, slow process until a tipping point is reached. For WorkWell, it looked like bringing together stakeholders across Victoria who had never worked together on prevention before – and those partnerships became increasingly mature and collaborative over the course of the program.Applied value – the system starts to shift.
This is when potential turns into action: changes in practice, policy, and resource flows. For WorkWell, this included changes to employer practices and policies, including some industry-wide ones, and the Victorian Government’s decision to continue funding WorkWell as a business-as-usual function of WorkSafe.Realised value – outcomes for people and systems.
This is the type funders most often go straight to, and that traditional VfM methods were built to measure. For WorkWell, this would have meant a reduction in mental injury claims – and we knew, going in, that this hadn’t happened yet. This is exactly why the broader framework was needed.Transformative value – systems transformed.
Transformative value is about shifts in mental models and underlying purposes: when deeply held beliefs about “how things are done around here” move. For WorkWell, this was arguably its most significant contribution: a paradigm shift in how employers, regulators and industry understood their role in preventing mental injury.
In practice, this typology gives stakeholders a wider vocabulary for talking about, evidencing, and valuing what systems-change work creates. We can even build rubrics around these types of value! It allows a deliberative group to ask not “have headline outcomes changed yet, yes or no?” but “given the nature of this work and its timeframe, has enough value of the right kinds been created to justify the investment?” That’s a more honest and generative question.
The evaluator as a trusted process-holder
All of this has significant implications for how we understand the evaluator’s role in systems-change VfM assessments – and it’s something I’ve spent a lot of time thinking about, both in practice and in my PhD.
Traditionally, evaluators have played the role of judge in VfM work. Maybe not overtly with a wig and a hammer, but by conducting highly technical analyses before presenting numbers that carry considerable weight – and can’t easily be interrogated by stakeholders – the evaluator is seen as an independent, neutral, objective arbiter.
I see a real opportunity – particularly in systems-change contexts – to rethink that role. When you work with stakeholders to surface what they value, and when you support a group of people to interrogate the evidence and reach a shared judgement, you’re shifting from judge to trusted process-holder. And I find that genuinely exciting! Our job becomes ensuring the right people can participate meaningfully, finding creative ways to present evidence so that diverse people can understand and interrogate it, and facilitating dialogue across different perspectives and power levels. Yes, it’s a different skillset, and yes, that might feel scary to some evaluators – but it’s a skillset I think we’re better placed to develop than we might expect.
This requires what colleagues have called “holding expertise in equal measure” (Bell and Anderson, 2023) – designing processes where technical knowledge, programmatic knowledge, and lived experience each have a legitimate place. It requires awareness of power, privilege, and positionality. And it calls for genuine humility and curiosity: stepping away from the role of “expert” in favour of facilitator, convenor, collaborator.
But this is also delicate work. We are often working within tight budgets and timelines, with groups that cannot possibly include everyone affected by the system. Some stakeholders may be unfamiliar with rubrics or sceptical about deliberative methods. Others may be accustomed to evaluations that deliver “the answer” without asking them to do much valuing work themselves. The evaluator is in the middle, shaping processes that are rigorous and trusted enough to support defensible judgements, but flexible enough to allow meaningful participation from diverse stakeholders.
A few practical notes for evaluators working in this space:
Start with a modest but real deliberative space. You may not be able to convene the perfect group. Start by bringing together a small but diverse group who can credibly speak to different parts of the system – and design a process where their reasoning genuinely matters for the judgement.
Make valuing schemes explicit. Use the rubric development process to surface whose values are shaping the criteria and standards. Ask openly: what kinds of value – and whose values – are we privileging?
Treat the process itself as a learning intervention. The work of articulating criteria, agreeing standards, and working through evidence can itself shift how people think about the system and their role in it. As Gates and Schwandt argue, we’re not just determining value – we’re developing it. Make space for that.
Keeping the north star in view under real-world constraints
Every evaluation is a compromise. We make choices about who to involve, how much time we can reasonably ask for, what kinds of evidence we can collect, and how far we can push against institutional norms and expectations. There will always be limits on how democratic and inclusive our processes can be.
This can feel really hard for evaluators trying to conduct an evaluation with integrity. For me, this tension is exactly why having a north star matters. Democratic deliberation, anchored in transparent rubrics and a richer vocabulary of value, gives us an aspiration rather than a perfect recipe.
By treating value as something co-constructed, making evaluative reasoning visible, and situating our judgements within democratic deliberation rather than behind closed doors, we can make VfM assessments more trusted, more transparent, and more worthy of the complex work they’re asked to judge.
At the end of the day, we may still need to deliver a judgement: was this a good use of resources? But the way we answer it – who gets to be part of that process, how we reason together, and what we’re willing to count as valuable – can itself be part of the systems change we’re trying to create.
Final words from Julian
Heidi makes a point that I think many of us in evaluation feel in our bones: it is not enough to say an evaluation is participatory because we interviewed stakeholders or ran a workshop. We can run a “participatory” evaluation and still leave the valuing work untouched. If we accept that eVALUation is about determining the value of things, and that things do not have value, people place value on things, then meaningfully sharing power with stakeholders and rights-holders when systematically determining value is not just an intriguing methodological option - it becomes necessary for validity, credibility, ethics, and use.
That is why the shift she describes – from evaluator-as-judge to evaluator-as-trusted-process-holder – is so important. When you work with stakeholders to surface what they value, and when you support a group of people to interrogate the evidence and reach a shared judgement, you are changing both the answer and the way the answer is produced. Some contexts demand independent evaluation, while other evaluations are designed from the ground up to enable a group of stakeholders to form their own conclusions. Either way, I see this facilitative approach as an important part of working meaningfully with stakeholders to understand their values.
I also love Heidi’s line: “Rubrics create a shared language for disagreement”. You don’t have to agree with an evaluative judgement, and rubrics help us move from “who are you to have an opinion about my program?” to “there is a specific criterion, a specific standard, or a specific piece of evidence we need to debate here”. That is a massive win for clarity, and it reinforces her point about treating the process itself as a learning intervention. The rubric becomes the centrepiece for sense-making conversations, instead of being mistakenly treated as a set of “rules” for interpreting indicators.
Heidi’s Cycles of Value Creation model advances some long-term thinking I’ve been doing with colleagues across multiple projects to articulate value propositions (why might this be worth investing in?) and theories of value creation (through what magic does the investment create more value than it consumes?). Heidi’s schema gives evaluators and stakeholders a structured way to unpack those questions in systems-change work, inviting deliberative conversations to ask, not “are we there yet?” but, as Heidi put it, “given the nature of this work and its timeframe, has enough value of the right kinds been created to justify the investment?”
I’ll end on another line from Heidi that has stayed with me since I first heard it on Matt Healey and Tenille Moselen’s It Depends podcast: “I’d challenge people to not make their peace with what is keeping them up at night, and keep striving to do things better and differently.” Held with pragmatism, collegiality, and a healthy sense of humour, that kind of restlessness is exactly what our field needs. And it’s why I’m so pleased to be able to share her work with you here.
Further reading
Peterson, H. (2025). Democratic deliberation as a North Star: Showcasing a framework to assess the value for money of systems-change efforts. Evaluation, 0(0). https://doi.org/10.1177/13563890251386814
Peterson, H. (2022). Cost-Benefit Analysis (CBA) or the Highway? An alternative road to investigating the Value for Money of international development research. The European Journal of Development Research, 35, 260-280. https://doi.org/10.1057/s41287-022-00565-7
Heidi Peterson’s framing of democratic deliberation as a “north star” for value in systems change is a significant and much welcomed shift.
It recognises something fundamental:
That value isn’t inherent in programmes or investments - but it’s constructed through perspective, power and experience.
Her emphasis on inclusion, dialogue and shared judgement moves evaluation beyond technical assessment and closer to legitimacy.
In my own work on Experiential Intelligence in systems leadership, I have been exploring a closely related challenge - what happens after those perspectives are surfaced.
Because even where systems create space for participation or deliberation, there is still a persistent gap:
How does what is known through experience actually shape decisions, translate into impact, and become embedded as learning?
To address this I’ve developed a decision-to-impact traceability loop:
Experience → Insight → Decision → Impact → Learning → Integration
This is not just an evaluation cycle.
It is a relational governance mechanism - designed to ensure that what is heard is structurally connected to what changes.
From this perspective, democratic deliberation plays a critical role. It strengthens the front end of the system - how value is understood, reasoned, and negotiated.
But for systems to act with integrity, that deliberation has to remain visible as it moves forward. This goes into decision making, into how impact is understood, and it also goes into how learning is integrated over time.
Without that traceability, even the best designed deliberative processes risk becoming parallel to the system, rather than shaping it.
What I see and feel emerging across evaluation, systems thinking, and lived experience leadership, is a convergence:
We are moving closer towards systems that do not just measure value, but they are capable of learning from experience in real time, and acting on it transparently.
This is where democratic deliberation and experiential intelligence meet perfectly. Not just in how we judge value, but in how systems evolve as a result of it.