Issue 09 | Against the Romance of Community

Hi friend,

Over the years I have facilitated a good number of intergenerational consultations, many of them with Global Majority community groups, and I have learnt that sharing a background with the people in the room does not mean they share a view. In one recent conversation, which I am keeping deliberately vague because the work is still ongoing, people who shared a postcode and, broadly, an ethnic background disagreed deeply about the place of faith in their lives, and as they spoke about their own experiences of structural racism, some of them reached for sweeping statements about other minoritised groups. That scepticism had a long history behind it, but it still landed in the room, and I had to decide in that moment how to hold it.

A funding proposal would have called that room "the community", and I keep coming back to how little that word would have told anyone about what was actually happening inside it.

I have started to think of community, as we use it in NGO land, as a fundable fiction, a unit coherent enough to sit in a budget line and vague enough that no one has to account for the disagreements, hierarchies and harms inside it. Miranda Joseph called this the romance of community, and I recognise it in more than one proposal I have written that promised work "co-designed with communities" without ever saying who, or how, or what it cost.

We can see it most clearly in our indicators. "Communities feel more empowered" sounds like an outcome, yet it never tells us who answered, who was never asked, or whether the person who was harmed in the focus group is counted in the figure. When we let one marker, whether a postcode, a diagnosis or an ethnicity, stand in for a whole group of people, we flatten exactly the differences that decide who benefits from a programme and who is quietly left out of it.

And now that an AI tool can write "co-designed with communities" into a proposal without anyone ever being asked anything, I would argue that the slow, uncomfortable and properly resourced work of actually bringing people together matters more than it ever has, because it is the one thing the language can no longer prove on its own.

Try this week

Twenty minutes, on paper, with one proposal, report or project page you have written this year.

The word. Find every place you wrote "community" or "communities". For each one, write down exactly who you meant, and who sits inside that word but was never in the room.

The differences. Which tensions within that group, across faith, age, gender, migration or class, does the document leave unnamed, and what would change in the design if you named them?

The harm. If someone in your process said something that wounded another participant, what would happen next? Who would follow up, and is that anywhere in the budget?

The method. Was a focus group or a single consultation your default, and was it actually the right way to bring these particular people together?

The indicator. Take one outcome that begins "communities feel" and rewrite it so that it names who you mean, and how you would know if the answer differed from one person to the next.

The point is not to stop using the word, which still names something real about how people find one another and act together. It is to stop letting it do our looking for us.

Until next time.

Jess

P.S. The long version is on Substack. It goes into how differently people living with HIV are represented even within peer-led work, Bernice Johnson Reagon on why coalition work was never meant to feel like home, and what a duty of care looks like when the risk sits inside the group rather than outside it. Read it here

P.P.S. Design for Social Impact Lab will be running learning experiences again in 2027, for non-profit practitioners, learning and evaluation and philanthrophic funds: eight weeks, a global cohort, with parity pricing and payment plans. It is where this kind of honest interrogation of our own practice gets done properly. Find out more and join the waitlist