A supermarket chain has submitted a new application to build on a riverside meadow at the eastern edge of Bath – a site bordered by a busy road, a floodplain and a growing list of ecological concerns. On the surface, it looks like any other planning submission: a set of documents, a three-week consultation window and the usual assurances that everything has been carefully considered. But digging into the application reveals more than just development intentions. It offers a glimpse into how local democracy functions – or fails to – when filtered through the machinery of planning.
There are nearly a hundred documents. Some are drawings, some technical reports, others impact assessments or consultation responses. Each serves a specific statutory purpose and collectively they form the official narrative: this project is necessary, manageable and ecologically benign. The language is polished and formulaic. Negative impacts are always “low,” “negligible,” or “temporary.” Habitats aren’t lost – they’re “offset.” Inconveniences aren’t created – they’re “mitigated.” It’s not spin, exactly. It’s a genre.
Accessing the material is an exercise in endurance. The council’s planning portal does not support batch downloads. Navigation is slow, documents open in new tabs and filenames are cryptic. There’s no executive summary – just a hope that residents will piece things together themselves if they care enough. This isn’t conspiracy or cover-up. It’s just what happens when civic infrastructure evolves to meet legal obligations, not human needs.
The ecological assessments are a case in point. Surveys are often conducted at times of year when species are least visible. Lighting impact studies use thresholds that don’t reflect real-world conditions. Biodiversity Net Gain calculations appear rigorous but rest on assumptions about what counts, what doesn’t and how loss can be measured in units. There’s no malice in this – only compliance. The developer has done what the system asks of them. The system, in turn, has been designed not to ask too much.
Then there’s traffic. The site sits on a known bottleneck – one of the main routes into the city. Adding a large retail unit with parking and delivery access will clearly change how vehicles move through the area, yet the modelling is based on best-case assumptions and offset by the promise of a few bike racks. Air quality around nearby public spaces, including a children’s park, is barely discussed.
Underlying all this is a deeper question: what kind of city are we trying to become? There’s no shortage of shops in Bath. There’s no shortage of traffic. There is, however, a shortage of undeveloped land that still functions as habitat, floodplain and visual break – not because it’s protected but because no one has got to it yet.
Planning applications like this aren’t exceptional. They’re routine. That’s part of the problem. Every few years, a new version appears, dressed in new language, updated with fresh graphics and a slightly different access route. It becomes harder to track changes or remember what was promised last time. Public memory is fragmented. Objections are reset. The game restarts.
To object is not to idealise the land or demonise development. It’s simply to say: this site matters. Not in some abstract heritage sense but in practical, long-term ways that current models don’t capture well. It buffers the river. It softens the urban edge. It holds space for things we can’t easily measure or price.
And once it’s gone, it won’t come back.