Evaluation becomes a tool for social justice when it stops extracting and starts learning – and lets that learning transform practice, including that of philanthropy.
How can we better understand the impact of funding without overburdening community partners? And how can such an assessment avoid devaluing community knowledge, treating the views of funders and outside evaluators as “objective” and the practitioners’ as merely “subjective”? Evaluating social transformation takes many kinds of knowledge but cannot be done without that of those doing the work on the ground. Taking practitioners’ knowledge seriously does more than sharpen our understanding of change. Our argument goes further: evaluation becomes a tool for social justice when it stops extracting and starts learning – and lets that learning transform practice.
Many funders are experimenting with ways to move away from traditional reporting and the power imbalance it reinforces, a shift often gathered under the banner of trust-based philanthropy. Some let grantees define how they evaluate change, dismantling the false dichotomy between trust and results; others take decolonial approaches that foster peer learning. This fall, Philanthropic Foundations Canada chose reciprocity as the theme of its national conference, a sign of how central the willingness to rethink power and relationships is to the sector. Funders are increasingly willing to share power and trust community organizations to act, but are they ready to trust their knowledge of what has changed?
Really listening . . . rests on a conviction that community, practitioner, funder, and research knowledge are all necessary.
We want to build on this work and go one step further, showing not just that community knowledge matters, but how it can be built into evaluation and travel back to reshape how funders act. The learnings we describe here emerged through dialogue between community partners, funders, and researchers, each an active player in social change, with a different perspective to offer. Really listening, in our view, is not a check box. It rests on a conviction that community, practitioner, funder, and research knowledge are all necessary. The point is not only to value and exchange that knowledge, but to let it transform practice, including philanthropy’s own.
Evaluation as dialogue
What might this look like in practice? Let us look at one case we have been learning from firsthand: the Collective Impact Project (CIP) in Montreal. One of us works as an action researcher inside the project, leading its evaluation; the other is an academic who sits on the CIP’s evaluation and learning committee. For more than a decade, the CIP has pooled philanthropic funding for collective action in Montreal neighbourhoods; it now brings together 12 foundations across 31 of them. In its current phase, we have tried to design its evaluation as an act of dialogue, moving as far as possible away from extraction.
This begins with trust, and in the CIP that trust is built into the funding itself. Flexible and multi-year, it is not tied to specific targets a neighbourhood must hit to be refunded, but rather a willingness to strengthen collaboration, experimentation, and learning. That matters, because partners will not share what is not working if their funding is at risk. On this basis, our learning approach follows three orientations.
Gather learning in existing spaces
The first is to gather learning in existing spaces. Instead of adding extra meetings for the purpose of evaluation, our team works within the spaces the CIP already uses to reflect together: biannual meetings with partners, learning events, and governance bodies bringing community, institutional, and philanthropic partners together. These spaces are designed to help learnings emerge while also strengthening relationships. Apart from brief financial reports, written reports were dropped entirely in the current phase. What stakeholders share about outcomes, barriers, and good practice is treated as data and analysis to be taken seriously, not as anecdote to be verified against “harder” evidence. Under the leadership of an in-house action researcher, the funders’ project team gathers these learnings, then shares back summaries that partners are invited to correct and adjust. These summaries are the core source for the evaluation of the whole program, which is itself designed in a participatory way, through a learning and evaluation committee where people from the community, philanthropy, institutions, and academia work together. In short, the material our evaluation draws on comes from spaces that would have existed anyway, so the documentation burden sits with our team rather than with community partners.
Reciprocity
The second is reciprocity, which goes beyond sharing power in decision-making or fostering mutually beneficial relationships: it is also about learning. It is one thing not to overburden community partners with evaluation requirements; it is another to put the funders’ own learnings on the table. For now, this is more an aspiration than a settled practice. The intention is for exchange to move in several directions, which means that not only community partners share what they are learning, but funders and researchers as well. Some partners share openly. Many funders still find it uncomfortable, which shows how deeply ingrained roles are, and how hard they are to shift.
Let knowledge travel
The third is to let knowledge travel. What is learned is not meant to sit in a report or remain organizational or internal knowledge but to circulate among community partners, funders, and outward to other actors so that it can shape practice well beyond neighbourhood initiatives or even the CIP as a whole. This happens through accessible channels: learning-exchange events, publications, and a public “neighbourhood profile” for each of the 31 neighbourhoods, with learning briefs now in preparation. This kind of sharing was thin at the outset and has grown significantly over the last few years.
And yet this approach, based on dialogue and trust, still produced a clear reading of change. In most of the 31 neighbourhoods, we were able to identify outcomes at all three levels: citizen, organizational, and systemic – 15 in total, eight of them shared across more than half of the neighbourhoods. At the citizen level, they range from better access to services through better outreach, new offers, and clearer communication, to a stronger sense of agency among residents and a stronger social fabric. At the organizational level, many adjusted their practices through experimentation and learning. One food bank has looked closely at what was keeping people away and completely adjusted how it operated: its opening hours, the way food was distributed, the way staff connected with people. Eventually it made access entirely free. At the systemic level, community advocacy secured a commitment to inclusive development within a major urban development project. The methods we used to understand these outcomes might be specific to the CIP, but the principles aren’t.
New paths of collaboration began to emerge. As relationships deepened, community partners began to ask for support beyond funding, in particular for advocacy and influence, to help carry an issue beyond the neighbourhood scale or mobilize networks and connections. That these requests emerged at all is itself an outcome of the trust the approach is built on.
The approach has also shown clear limits. Most often, it does not measure change in numbers, and it reaches citizens only indirectly, as the spaces we work in bring together partners, rarely residents themselves. We learn about residents’ experience through those partners. They decide who to invite to our meetings, who they listen to in their work, which voices to bring forward, and what to share with us. Our focus here is the knowledge of those who do this work, not the experiential knowledge of residents themselves, which raises its own questions.
Toward knowledge democracy in evaluation
Why does this matter? Because how we evaluate is never just a technical choice; it is a question of whose knowledge counts. The reflex in much of the sector is to treat distance as objectivity and proximity as bias: the funder’s or the external expert’s reading is taken as rigorous, the practitioner’s as merely anecdotal. But distance is not neutral, either. The external evaluator or researcher comes with their own background and experience and, usually, with a framework built elsewhere. Their view is positioned, too; it is simply rarely recognized as such. All knowledge is situated, including ours. Once we recognize this, the usual assumption doesn’t hold. Practitioners and community actors face the issues daily; they hold a fine-grained understanding of what is shifting and why, which no external observer can reproduce. Their proximity is an asset, and leaving their knowledge out of an evaluation is what makes it incomplete. Rigour comes from mechanisms built into the process: summaries that partners correct and validate, findings discussed with a committee where three kinds of knowledge interact, and patterns tested across 31 neighbourhoods. Recognizing a plurality of knowledge is not a concession to be made to community partners; it is a matter of rigour, and of justice.
This is what Budd Hall and Rajesh Tandon, the UNESCO co-chairs in Community-Based Research, call knowledge democracy: recognizing many ways of knowing and making sure they are accessible, valued, and shared. What people learn from experience, from practice, from living in a community counts and needs to be heard alongside academic and professional expertise. None of it is sufficient on its own or replaces academic research. Rather than simply gathering these knowledges and setting them side by side, they have to meet and interact, so that their dialogue can change how each of us acts – community organizations, institutions, researchers, and funders alike. Understood this way, evaluation is no longer a verdict on other people’s work. It becomes a shared process of learning that can reshape practice, including in philanthropy.
The question is no longer whether community knowledge deserves recognition, but why evaluation so often still does without it.
This loop is not hypothetical; the Collective Impact Project has been reshaped by it. Community partners found the evaluation of the first phase of the project too heavy, so the approach and its reporting were redesigned around lighter, dialogue-based methods, informed by their input. Training offered to initiatives has been adjusted the same way. Most tellingly, as the present phase draws to a close in 2027, the design of the next phase was co-constructed with these partners, who were by far the most present voices at the table. The knowledge gathered through evaluation did not simply describe the work; it changed the program itself.
None of this is easy, and what stands in the way is less a matter of method than of mindset. Equating objectivity with distance runs deep. Trust, too, takes time to build; the reciprocity we described remains uneven precisely because roles and power dynamics that have long been fixed are slow to shift. Listening and building a transformative dialogue is not a method you adopt overnight. It takes willingness and practice, something funders and community organizations grow into and have to learn. You don’t need a project of the size of the CIP to start. Any funder can begin by asking themselves, Whose reading of change do we treat as rigorous, and whose do we quietly dismiss? Does that choice shift power or reproduce it? And what do we fail to learn about the work we fund when making these choices?
The question is no longer whether community knowledge deserves recognition, but why evaluation so often still does without it. This brings us back to where we started: how far are we willing to push reciprocity? Can it go beyond mutually beneficial relationships, as far as recognizing and sharing each other’s knowledge? The systemic change and inclusion that funders pursue in society must extend to the way knowledge itself is produced and shared, because a fairer distribution of resources and a more equitable recognition of whose knowledge counts are, in the end, two sides of the same pursuit of social justice.
