Most planning workflows are drawn as pipelines: a needs assessment feeds a strategy, the strategy feeds a design, the design gets built, and the diagram ends. The TU Delft Planning Cycle is drawn as a loop instead, because the tenth step, Upscale, does not conclude the process. It feeds back into the first: what was learned scaling one intervention becomes part of how the next community’s needs are read. The Cycle assumes that no plan is ever really finished, only current.

Where power is fragmented

The Cycle is built for contexts where power is distributed unevenly across many actors rather than concentrated in one authority: municipal departments, private developers, civil society organisations, informal residents’ associations, all with partial and sometimes conflicting mandates. In these contexts a single masterplan handed down from above tends to be resisted, ignored or hollowed out at implementation. The Cycle’s participatory steps exist because collaboration is not an ethical add-on to planning in fragmented settings; it is the only route through which a plan can actually hold.

No universal recipe

The ten steps are not a checklist to be applied identically everywhere. Different cities and communities carry unique social, economic and cultural dynamics that shape which participatory methods are realistic, which stakeholders hold real influence, and how long each step can reasonably take. A neighbourhood association with decades of organising capacity can run a co-design step very differently from a community still building the relationships that make co-design possible. Adapting the Cycle to local conditions is not a deviation from the method; it is the method.

The Cycle is process; the Spatial Justice Conceptual Model is the substantive framework that process is meant to serve. Every step touches more than one dimension in practice, but each has a centre of gravity: a dimension of justice it is chiefly built to advance. The table below marks that centre of gravity for each of the ten steps, so the two resources in the Package can be read against each other rather than in isolation.

Read across a row and the pattern is not evenly spread, and it should not be. Diagnosis work leans recognitional, because the first task is to see who and what has gone unseen. The design-heavy middle of the Cycle leans procedural and distributive, because that is where decisions about voice and allocation actually get made. This is a starting point for discussion, not a fixed rule: a facilitator running Step 2 with a community that has long been excluded from consultation may find it doing as much recognitional work as procedural.