The Diseconomies of Fail – No Large Garden Communities do Not Cost More Than Smaller Development

One of the oddest and most disingenuous arguments I have heard recently.  And in Planning too from CAUSE

Size matters, but has not been debated, yet larger proposals bring complexity, risk and an increasing requirement for infrastructure. Delivery of larger projects is slow (think Ebbsfleet and Northstowe).

Our Small is Beautiful paper suggests that viability decreases at more than 2,000 homes.

Looking at this paper:

There are significant scale diseconomies in building big new towns. The bigger the town, the longer it takes to build, and the higher the cost per dwelling.  The dramatic cost increase arises because of the finance costs on land. The increase in total
land costs as the holding period increases is astounding. Land bought now for £100,000 per acre escalates to £179,000 after 10 years and £1.8m after 50years.

In theory big settlements might benefit from scale economies. It might be possible to build a sewage works or power supply at a lower cost per dwelling for a 24000 settlement than for a 2000 one. But we have seen no evidence that this is the case and we therefore ignore it.

The only claim to scale economies made during the recent Examination of the North Essex. Garden Communities plan related to education. The NEAs variously suggested that settlement sizes of 5000 and 15,000 are needed to support an optimum number of secondary schools. No detail was provided, and CAUSE would argue that secondary schools can be provided every bit as efficiently in smaller well connected urban extensions as in large standalone new towns.

There are four prblems with this argument:

1.Economies of Urban  Agglomeration

Why do cities exist in the first place?  Why isnt economic activity scattered across an even plane? This question has been the driving force of urban economics in recent years.  it has even led to a noble prize.

As Wikipedia states:

As more firms in related fields of business cluster together, their costs of production may decline significantly (firms have competing multiple suppliers; greater specialization and division of labor result). Even when competing firms in the same sector cluster, there may be advantages because the cluster attracts more suppliers and customers than a single firm could achieve alone. Cities form and grow to exploit economies of agglomeration.

Indeed the optimum city size is when the economies of aggomeration matches the diseconomies of congestion.  the more you invest in transit infrastructure the higher this size can be,

CAUSE suggests an optimum size of 2,000.  However a CBA need to look at both costs and benefits in both the public and private sector.  Otherwise you get in the reducio ab absurdio that London is too big as it most economically should have stuck at the size of a roman town of 2,000 dwellings.

2. Fallacies of Composition

You arn’t comparing (say) once scheme of 20,000 dwellings with one scheme of 2,000 dwellings.  To hit the same targets you have to build 10 of these and look at the cumulative costs.  Not all of these could be built in year one.  The absorption rate in the local housing market would be the same.  You also have to phase the schools and other infrastructure.  You therefore cant assume in any way that the costs of holding land would be different.  They would be exactly the same.

3. Economies of Scale

The most disingenuous part of CAUSEs argument is it own name as they are the Campaign for Urban Sprawl in Essex and they propose more dispersed development over many smaller sites.  Their own proposals for Colchester and Tendring has been assessed as only supporting round 6,000 dwellings forcing the rest to go to scattered villages.   Which is not to state that this plan could not be accomodated alongside one or more large Garden Communities.

There are very clear economies of scale from developing at large scale as any civil engineer will tell you.  You might need for example one new sewerage main rather than 20.

This is particularly clear in the education sphere. Imagine an area where all secondary and primary schools are full.

This means that at at average household size you need around 2,400 homes to build a two form entry new primary school.  a small two primary feeder secondary would have 4,800 homes, more economically twice this.  The last this you would build is half a primary school.  This leads to what I call ‘development quanta’ with optimum site sizes and densities going up in discreet quanta.  You vary the design size at design year of the development to fall outside the plan period if you have an odd residual.

There is a mass of research evidence on the high infrastructure costs of servicing scattered sprawl as opposed to concentrated urban development.

For example from Siemens

Studies undertaken by the cities of Calgary, Canada and Los Cabos, Mexico identified significant savings on infrastructure costs could be achieved through more compact growth. Savings of 33% and 38% were identified for the capital cost of roads, transit, water and other infrastructure for Calgary and Los Cabos. Savings on operational costs were 14% for Calgary and 60% Los Cabos.

Also more scattered development means that infrastructure networks of greater length need to be developed.

Todd Littman

Is the acknowledged expert on this working with LSE cities.

our analysis indicates that by increasing the distances between homes, businesses,  services and jobs, sprawl raises the cost of providing infrastructure and public services by 10-40 percent.


There are also disecnomies of servicing scattered smaller development by transit. Rapid transit is only economic above a certain critical mass of housing within a single route at densities that secure residents within walking distance.  You wont get transit systems for scattered development around 20+ villages.

This also means the costs of mitigating traffic from highr car use from scattered development

4. The  False Assumption of Buying all the land at once at existing use value

Neither would a rational house builder pay up front for land in year 1 when the housing would be phased in 5 year phase three.  Thie option agreement would trigger payment in that phase.  It is therefore a ridiculous and non real world assumption to assume vast interest payments on buying all land up front.  It simply would never happen.  Even if bought at CPO the ‘non scheme world’ utilising the ‘point guarde principle’ would be the non garden community world what its infrastructure. in the light of caselaw, and reforms made in the  Neighbourhood Planning Act 2017, the cpo has to ignore completely the infrastructure and project and any special suitability of the site for the project.  These collectively mean that sites for Garden Communities have a much lower EUV than many scattered developments that would come forward anyway in the ‘no scheme world’ local plan.  So the conclusion on valuation is the precise opposite of that put forward by CAUSE.

All of these economic factors can and should be built into the revised SEA, then with the expert evidence based on international consensus on the costs of sprawl, the outcome will be different.

2 thoughts on “The Diseconomies of Fail – No Large Garden Communities do Not Cost More Than Smaller Development

  1. Isn’t there an extension to your fourth point – which is that even if you did have to buy all the land up front (perhaps because you didn’t want to be held to ransom later) you could put it to use in the meantime. Either by keeping is as farmland, or finding new interim uses ? If the price you pay is EUV and you keep the EU then it doesn’t count just as a cost.

    Does that make sense ?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s