14 December 2018 Small is Beautiful 2 The wider context Introduction Last year we published Small is Beautiful, a paper which makes one simple but very important point: financial viability falls for new developments of more than a certain size which we estimate at around 2000 dwellings in North Essex. This is the number which can be realistically delivered within 10 years of land purchase. It is at this point that the cost of holding the extra land the cost of equity and debt exceeds the extra residual value from an extra acre. This follow-on paper addresses some of the comments we have received and considers some wider political issues. Land acquisition strategy Our critics point out that Small is Beautiful relies on very specific assumptions about how land is acquired. They argue that there are lots of ways of acquiring land, and it is not necessary to assume it is all bought on day 1. In a narrow sense they are right. Our model is a simple arithmetic one which does indeed assume that land is purchased on day 1 and is financed at 6% 1, at the low end of the finance charges used in 1 6% is higher than the borrowing cost of many developers, especially the larger ones. It is better regarded as a low weighted average cost of capital. Ie a weighted average between interest cost of say 3% and the cost
development appraisals. Our model is deliberately simple; for example it assumes the same infrastructure cost per dwelling for large settlements as for small. But as with many simple financial models, the underlying point holds true it is very difficult to capture land value uplift on big sites. We agree that land does not have to be bought on day 1. But its price must be controlled, otherwise the land value capture needed to fund infrastructure will be lost. Land value capture is at the centre of the garden community movement and without it the infrastructure cannot be delivered unless Government writes a big cheque. If the land is not purchased then an option or fixed price contract is needed, both of which involve a cost of capital. If land value is to be captured, a smart strategy is needed, and that will be easier when there is a big choice of option virgin land to choose from. In North Essex the authorities have identified land which they do not own for three big new towns. The land already had hope value which has increased now that large sums of public money have been spent claiming that the chosen locations are sustainable. The public sector has put itself in a weak negotiating position: landowner expectations have risen, the land has mostly been optioned to experienced promoters or developers, and in one area they face a consortium of landowners, which reduces competition. More recently the authorities have started to talk about relying on compulsory purchase powers and a dubious interpretation of Pointe Gourde 2. The land under consideration already has considerable hope value, and it will be very difficult to determine a no scheme value with any certainty. The NEGC model North Essex Garden Communities Ltd hopes to become a development corporation which will buy land, masterplan it and sell fully infrastructured plots to developers. But it does not have planning powers, and it does not own the land. One of the three new settlements, West Tey, at 24,000 dwellings, will take nearly 100 years to build 3. So how is land value uplift to be captured? It will be not easy. The Councils have promised upfront infrastructure investment which will further inflate market values. The table below shows the various twists and turns of the debate but it all comes back to the same point: It is very difficult to capture land value uplift for big new towns. of equity of say 15%. Note that quoted housebuilders have very expensive equity due to the cyclical nature of the business, and are currently trading at p/e ratios of below 10. 2 Pointe Gourde Quarrying & Transport v Subintendent of Crown lands (Trinidad) establishes that compensation for compulsory acquisition of land cannot include an increase in value which is entirely due to the scheme underlying that acquisition 3 Assuming delivery at 250 dwellings per annum, a figure that is supported by the Planning Inspector. It is higher than the average of 161 dpa experienced at present but lower than the 500dpa in the Hyas viability study.
The landowner model The Gateway 120 Consortium 4 claims that: its members already own enough land for a large garden community (17000 dwellings) and can deliver it for development as and when needed. The landowners will bear the landholding costs themselves and the scale diseconomies argument therefore does not apply. The small is beautiful argument works rather differently for landowners developing their own land. They will have a cost of capital even if they do not always see it. Their land will be tied into 30-100 year commitments during which other opportunities will be restricted. We have analysed the deal for landowners in a paper which formed part of our August 2017 consultation response. The landowner development model has to work for local communities too. Assurance is needed that the infrastructure will be delivered in a timely manner and funded from land value uplift rather 4 A consortium of landowners in the Marks Tey area who are promoting their land for a large scale garden community as an alternative to NEGC
than the public purse. This is difficult for normal 10-15 year schemes and becomes unworkable for 30-100 year ones. There is deep distrust of developer promises, and legal structures strong enough to overcome this will be expensive to document. The long-promised Beaulieu Park railway station provides a vivid local example of the public sector bearing the infrastructure cost while the private sector benefits from the land value uplift. If a landowner model for large scale garden communities is to succeed much more transparency will be needed - both on the nature of the landowning consortium and the structure of the funding arrangements. The land owners will still want to receive fair value for their land under this model, and therefore land value capture remains a problem. The scale diseconomies will still be there, but will emerge in a different way. Justifying the big settlements We are alarmed by the lack of economic analysis to support the decisions on the size and location of the new garden communities. Central government is promoting garden towns of >10,000 homes and garden villages of 1,500 to 10,000 homes. The entire strategy is underpinned by the assumption that land value uplift will fund infrastructure (a core garden community principle), but it seems that no-one has done the figures. There is too much reliance on scaling up 1 hectare blocks, a useful technique for smaller projects but wildly misleading for larger ones. In North Essex the sustainability appraisal fails to justify the key decision to build three big garden communities, and the financial viability evidence submitted ignores the cost of funding the land a key error. The Planning Inspector has asked for credible evidence of viability and the Councils are doing more work, as yet unaware that more work will not do the trick. The large communities they are planning will not capture land value uplift, and thus will not be able to deliver the infrastructure that is needed. At the Examination in Public 5 it became clear that the authorities have no theoretical framework for making decisions on size. It was suggested that critical mass was needed for a secondary school but this was variously used to justify scales from 3000 homes to 15,000 with little supporting analysis or evidence. Size needs addressing. Are there really scale economies in secondary schools which might outweigh the huge landholding costs? Are people really happier in large towns than in smaller communities? Are Garden Community principles best delivered by large communities or small? Are smaller settlements sustainable? The Garden Towns strategy is underpinned by an assumption that it is better to build brand new settlements rather than supporting the facilities of existing villages. In November 2018 the CLA published a paper on sustainable villages which highlights the need for some villages to grow if they are to keep their churches, pubs, schools and surgeries. In some cases growth is stifled by overrestrictive planning policies, and there has been a failure to recognise the role of broadband and social capital in making village living more sustainable. 5 Examination in public of the North Essex Authorities combined Part 1 Local Plan January 2018
The CLA also demonstrates the ability of landowners to deliver significant numbers of houses in smaller settlements: only 2% of landowners have ever built more than 50 properties. We argue that it is the smaller landowners who will be more in touch with their communities and more likely to generate sustainable development. The big developments will be tied up with promoters and extracting community benefit will be more difficult. Evidence from the real world Our hypothesis is strongly supported in the real world. Despite many attempts 6 no large new towns have been delivered at the scale of West Tey since Milton Keynes and policy makers should be asking themselves why. Could it be that it is difficult to capture land value uplift on mega sites? Sir Oliver Letwin reports that large sites (>1500) are currently delivering only 25,000 dwellings per annum, less than 10% of the 300,000dpa target. Even within this group there is a heavy bias towards the smaller end: Government has encouraged Councils to be bold, and the result is a crop of Garden Towns that are much bigger than our 2000 threshold. Even the smallest (at 8000) is substantially bigger than the 3700 average of the large sites addressed by Sir Oliver. 6 For example Labour s Eco Towns.
A more realistic scale is illustrated by successful developments such as: Poundbury was started in 1993 and has grown to a population of 2500 (1100 dwellings 7 ). It is projected to grow to 4500 (2000 dwellings) by 2025. Great Notley, population 5500, roughly 2400 dwellings over 20 years Beaulieu Park, [3600] homes over 20 years on land already owned by Countryside properties Milton Keynes (100,000 dwellings) is regarded by some as a success story for large scale development. But it is worth remembering that its development corporation benefitted from lower land values and it still suffered debt write-offs and restructuring. Fuller & Peiser estimate the total loss at 0.5billion. 8 Why 2000 dwellings? 2000 is not of course a precise figure each settlement will vary and the choice of location is vital. But it is a size that should be deliverable in 10 years, a period over which land funding cost will be manageable. The cost of land bought for 100,000 per acre at the beginning of the project will have risen to 179,000 in year 10 and 1.8m per acre by year 50. See diagram below. 7 When converting population numbers to dwellings we assume the national average of 2.3 people dwelling 8 See The Milton Keynes Experience by Richard B Peiser & Alain Chang. Also Transferable Lessons from the New Towns DCLG 2006.
How will the Councils deliver their housing targets? OAN: Large settlements take so long to deliver that they largely fall outside the legal 15 year plann period. In North Essex reducing size would have little or no effect because only 2500 dwellings for each settlement are projected for delivery before 2033. The benefits of planning beyond 2033 have never been articulated. Any benefits are likely to be outweighed by: 1. the political costs the entrenched opposition to large settlements, which engulf entire communities, and legitimate fears that the infrastructure will never be delivered 2. the economic costs the planning blight to the value of people s homes as well as the difficulties in capturing land value uplift. Five-year supply. Smaller settlements would actually improve 5-year supply because they can be started earlier: They will not need a development corporation developers and landowners can co-operate on smaller scale projects so long as they honour the Garden Community Charter There will be less need for cross border co-operation between Councils Land assembly will be simpler because there will be fewer owners and more choice of location. CPO is less likely to be necessary Land value uplift can be shared through more manageable CIL and s106 agreements Small delivers faster. Faster planning, faster build-out. How can land value uplift be captured for smaller sites? CAUSE s alternative strategy for land value capture is based on: 1. Immediate implementation of CIL 2. A stronger s106 approach based on the Mayor of London s August 2017 SPG 9 3. Planning resource directed towards brownfield and smaller sites rather than new towns 4. Working with neighbourhood plan groups to identify smaller sites below 0.25hectares 5. Reducing the target size for any garden communities to 2000 dwellings 6. Introducing competitive tension for sites by increasing the number considered 7. Approaching the new Sustainability Appraisal with an open mind as required by law 8. Activating the Inspector s option 1 to get a legally sound plan in place as soon as possible 9. Giving proper consideration to CAUSE s Metro Plan as part of a wider small is beautiful strategy 10. Numerate decision making rather than planning judgement 10. Introducing a proper analytical framework into the new Sustainability Appraisal. Conclusion We can find no good analysis to support the scale of the proposed Garden Communities. If the Government is serious about increasing housing supply before the next election it should come down to earth and promote smaller and more realistic initiatives. William Sunnucks MA ACA MBA 9 Homes for Londoners Affordable Housing and Supplementary Planning Guidance August 2017 10 CAUSE will be pushing for numeric evidence on all key choices made in the new Sustainability Appraisal which currently being prepared for the North Essex Section 1 plan.