When the numbers do the refusing

There is a particular form of quiet sabotage in English planning that rarely makes the headlines. It does not involve outright refusal. It does not require a dramatic political confrontation. It works by arithmetic.

Allocate just enough homes to trigger expensive, lumpy infrastructure — a secondary school, a new river crossing — but not enough homes to make that infrastructure economically rational. The scheme then fails the viability test, the inspector raises an eyebrow, delivery stalls, and the housing never arrives. The local authority can claim it tried. The numbers did the rest.

A strategic allocation in the emerging Southshire Local Plan offers a textbook illustration.

The numbers

The allocation for the hypothetical village of Fernley has settled around 2,225–2,300 homes. Earlier drafts floated higher figures; the published version landed lower. At this scale the development generates roughly 480–500 secondary-age pupils and 680–700 primary-age pupils under the county’s own yield rates. That is enough to require a new secondary school and land for primary provision. It is also enough to demand a new crossing of the River Dene and associated link roads, because the existing narrow bridge and village centre cannot absorb the traffic.

Independent cost ranges put the education package in the £55–85 million bracket and the transport works at £20–40 million. Combined, a central estimate sits around £90–110 million. These are not speculative figures. They rest on published pupil yields, recent actual school delivery costs in the county, and the Strategic Transport Assessment’s own identification of the new crossing and links as essential mitigation.

The viability cliff

A new secondary school has a minimum efficient scale. Department for Education guidance and most county authorities treat four forms of entry (around 600 places for Years 7–11) as the practical floor. Many prefer five or six forms for curriculum breadth and financial resilience. At 2,250 homes the Fernley allocation generates only enough pupils for roughly 3–3.5 forms of entry. The school is therefore forced into an undersized, high-cost-per-place configuration. The river crossing, meanwhile, is almost pure fixed cost. Spread across 2,250 dwellings it is heavy. Spread across 4,000 it is manageable.

This is the Goldilocks zone: large enough to trigger the big-ticket items, too small to fund them efficiently. Developer contributions scale with homes; the secondary school and the bridge do not. The residual land value is squeezed. Affordable housing percentages become harder to sustain. Phasing becomes fraught because the infrastructure is needed early while the cash arrives late. Viability consultants then report a gap. The plan is said to be “at risk.” Delivery slows or stops.

Increase the allocation to the 3,500–4,500 range and the arithmetic changes. The secondary reaches a proper five- or six-form size. The bridge cost is diluted. The same infrastructure package starts to look fundable from the development itself in a higher-value district. The current figure sits precisely where the economics are most hostile.

How the system enables it

Local plans are not neutral technical exercises. They are political documents constrained by the tests of “soundness” — justified, effective, deliverable. Infrastructure delivery and viability are central to the “effective” and “deliverable” limbs. An authority that wishes to limit growth does not need to invent a new policy. It can simply scale strategic sites into the zone where the required mitigation is disproportionately expensive relative to the housing numbers. The viability evidence then does the political work.

This is not conspiracy in the crude sense. It is incentive alignment. Councillors face concentrated local opposition to large housing allocations and diffuse, long-term benefits from increased supply. Officers must produce a plan that survives examination while managing political risk. Under-scaling a site so that its infrastructure appears undeliverable is a low-visibility way to achieve constraint without an explicit anti-housing stance. The inspector sees a technical problem rather than a political choice. The housing target is notionally met on paper; the homes themselves remain contingent on infrastructure that is unlikely to materialise at that scale.

Public choice theory has long observed that bureaucracies and elected bodies respond to the structure of costs and benefits they face. Housing supply in England is the classic case of concentrated costs (local amenity, traffic, school places) and diffuse benefits (national affordability, labour mobility, intergenerational equity). The local plan process, with its heavy reliance on site-specific viability and infrastructure testing, creates multiple points at which those concentrated costs can be made decisive.

The pattern

Fernley is not unique. Across England one finds strategic allocations that sit just below the threshold at which a secondary school becomes efficient, or just large enough to require a costly highway scheme whose funding is left uncertain. The result is a plan that looks ambitious in the housing trajectory tables and fragile in the delivery evidence. When the homes fail to appear, the authority can point to “infrastructure constraints” or “viability challenges” rather than to the initial scaling decision.

The deeper problem is institutional. A system that requires local authorities to demonstrate deliverability while simultaneously exposing them to intense local opposition creates a predictable incentive to design sites that are difficult to deliver. The more sophisticated the infrastructure and viability evidence becomes, the more precise the targeting of that incentive can be.

Conclusion

The Fernley allocation, at its current scale, occupies the least efficient point on the cost curve for the infrastructure it triggers. Whether this was conscious design or emergent outcome of political negotiation is secondary. The effect is the same: a housing proposal rendered precarious by the very mitigations it necessitates. Larger would have been more viable. Smaller would have avoided the major infrastructure trigger. The chosen figure maximises the chance that the arithmetic, rather than an explicit refusal, prevents the homes.

This is how planning systems can appear to plan for housing while quietly arranging its frustration. The numbers are not neutral. In the hands of institutions facing the wrong incentives, they become a method of control.