6 Conclusions, limitations and implications
Taking stock of modelling outputs
6.1. Summing together the output numbers projected according to our central assumptions, it appears that (in the absence of need-constraint), local authorities and RSLs have the capacity to develop some 6,400 homes p.a. over the period to 2034. However, while the RSL component of this figure factors in gross public subsidy similar to current levels (the £270M AHIP core programme), the total (real terms) annual cost of such a programme would be somewhat higher at around £348M . This figure incorporates the estimated annual grant cost of council housing development at this scale - £78M. This reflects the fact that, in contrast with our modelling for RSLs, our estimate of local authority housebuilding capacity has not been constrained within any specific grant funding envelope 74,75.
6.2. At an annual average of 3,894, projected annual output capacity of the RSL sector in the 25 years 2011-2035 would be only slightly greater than the current 'core programme' RSL output - around 3,600. This could be read as suggesting that the scale of currently un-utilised RSL housebuilding capacity is very modest. On the other hand, the recently initiated local authority housebuilding programme appears to be unlocking capacity for activity on a significant scale and which could continue to deliver appreciable numbers of new homes well into the future.
Distinguishing capacity modelling from output forecasting
6.3. It is, however, important point to note that our projections represent theoretical capacity rather than forecast output. The inbuilt assumption is that all social landlords utilise to the full their financial capacity and, in doing so, prioritise new housebuilding above other forms of investment (e.g. further upgrading existing stock beyond the SHQS).
6.4. Such an approach effectively moves away from the traditional model of giving relatively generous grant and rationing it in a bidding process. The system we are modelling is one where output and grant takeup is driven by ability to service debt and, in the case of RSLs, to maintain acceptable values on most financial performance ratios. This is our interpretation of modelling financial capacity. However, this is not the same as forecasting actual uptake. Some local authority and RSL governing bodies may opt not to use all of their financial capacity in this way, through being more risk averse than we have modelled, or having other priorities. This implies that actual takeup would tend to fall short of modelled capacity.
Implications for grant funding framework
6.5. The implications of our analysis could be seen as pointing towards a single grant funding stream for social housing. Given their recent record in developing new social housing at low grant rates, local authorities understandably see this as an attractive prospect. Nevertheless, in their initial resumption of a housebuilding role many councils have been aided in their ability to deliver competitive scheme costs (and grant rates) through exploitation of existing landholdings held on the Housing Revenue Account (and therefore potentially available for development at nil (or low) cost). As discussed in Chapter 2, in many authorities a continued municipal housebuilding programme on any appreciable scale and over any appreciable period will call for land acquisition exactly as is true for RSLs. Only via effective use of the landuse planning system to obtain sites at nil cost (see Chapter 5) will it be possible for authorities to continue to generate new housing on the 'efficient' basis seen in the recent Scottish Government grant-funded building initiative.
6.6. Case study RSLs voiced no 'in principle' objections to the proposal for a single, cross-sector, funding stream for affordable housebuilding - 'If local authorities can build more cheaply, they should do so' as one participant commented. However, RSLs stressed that any such system should constitute a 'level playing field'. There were, for example, concerns that local authorities could enjoy an 'unfair' advantage in being able to draw on existing landbanks and in having access to loan finance on terms considerably superior to those available to RSLs. It was also noted that the fairness of 'open competition' could be compromised by the local authority role in strategic land assembly. Similarly, councils could be seen as having certain inherent financial advantages - notably access to finance at public works loan board rates and VAT-exemption.
6.7. It is notable that, as shown by our research, most new local authority housebuilding is being managed in-house rather than developed collaboratively with RSLs. Whether such arrangements necessarily reflect the most efficient use of resources is not clear. There could be an argument for requiring grant-receiving authorities to demonstrate best value in procurement practice, to develop audited 30-year HRA business plans, and to file to the Regulator, 5-year financial projections to illustrate the future impact of housebuilding investment on rents and other key components of HRAs. This last requirement might even be linked to a contractual form of regulation as recently proposed by one respected expert 76.
6.8. Another important issue is that local authority capacity to develop at relatively low grant rates has been highly dependent on councils' ability to take on new debt funded via cross-subsidy from the main body of their rented stock. Providing that it does not extend beyond prudential bounds, leveraging existing council housing in this way could be seen as economically rational. However, since another consequence is to push up rents, this whole policy shift could be seen as requiring the Scottish Government to adopt an explicit national rent policy - i.e. a formal set of expectations on 'acceptable' rent levels and differentials to serve as a framework within which social landlords would be required to operate. For example, if we take the situation in 2015 where achieving SHQS compliance has required a local authority to increase rents to a level well above the national average, will it be acceptable for that authority to impose the further rent rises which would be required to cross-subsidise new construction? Against what benchmark could such a judgement be made?
6.9. As regards RSLs, the possible introduction of grant rates substantially lower than those applied historically will probably call for most or all developing associations to embrace as standard the 'cross subsidy' financing model whereby the cost of new development is underpinned, in part, by rental income from the main body of the landlord's stock. The ability to do so might be seen as the hallmark of a 'mature' sector where, over time, landlords have accumulated the capacity to reap the benefits from historic public investment and effective asset management. Nevertheless, this regime will pose new challenges for regulation, given the need to ensure that RSLs avoid both over-gearing and unacceptable rent increases.
6.10. For both councils and RSLs the prospect of new housebuilding needing to be substantially cross-subsidised from rents raises a 'political' challenge in relation to securing 'buy-in' from the existing tenant population (especially given the regulatory expectation for tenant consultation on annual rent changes). As discussed in Chapter 2, there is some experience of addressing this explicitly among local authorities. For many social landlords, however, this would represent a new departure. Assuming confirmation that Ministers plan to proceed with a 'cross-subsidy' funding model it would seem appropriate for the Scottish Government to provide a lead here by consulting with national tenant representative bodies on consequential implications.
Wider policy implications
6.11. The implications of our analysis are also profound in that they envisage a future scenario where local authority housebuilding regains a significant and enduring role in national housing policy. This reverses the established order as seen for a quarter of a century. Whether this is justified in terms of the quality of local authority housing management possibly remains to be proven. The significant number of poor ratings recorded in the 2003-10 round of regulatory inspections could be seen as raising some doubts here. This raises a question which might be asked in relation to both local authorities and RSLs: should eligibility to take on management of newly completed homes be made conditional on some demonstration of acceptable management performance - in terms of service quality, as well as the traditional service efficiency performance indicators? In the absence of an ongoing large-scale inspection programme, this could call for greater weight to be placed on tenant satisfaction ratings, as generated via a reformed framework incorporating robust and consistent standards 77.
6.12. The kind of approach described above would also have similarities with the English Arms Length Management Organisation programme where qualification for additional resources has been conditional on 'excellent' management performance as calibrated through inspection - a strategy which undoubtedly enhanced social housing management efficiency and effectiveness as seen in England over the past few years 78.
6.13. At least until 2015, the Scottish Government should also need to be convinced that new housebuilding activity does not provide a managerial distraction (as well as a competing funding requirement) for any social landlord still facing major challenges in delivering its SHQS programme. This would apply, in particular, to a number of larger local authorities 79.
Some limitations of the analysis
6.14. The capacity models used in this report rely on a smooth extrapolation for many items. The methods used are particularly simplified for some items in the 45 group of RSLs. We are not modelling all of the potential managerial feedbacks responding to weaknesses in the finances of individual RSLs. For example, in some cases where a number of ratios looked adverse, management would be quite likely to take various actions both in terms of real decisions (e.g. rent levels, spending budgets) and in terms of accounting treatment of certain items, in order to improve these ratios. Greater pressures from lenders could reinforce such tendencies.
6.15. More broadly, responses might extend into a wider process of mergers and consolidation in the sector. This could have mixed effects on the capacity to build. On the one hand, some weaker associations could become part of larger groups and thereby eligible to participate in development. On the other hand, the strength of some of the merger partners might be diluted in this process. We have not, of course, attempted to model the impacts of such transactions.
6.16. There are limitations of the partial HAG envelope controlling mechanism included in the model. It acts to some extent to ration activity where the grant envelope is not sufficient to accommodate all the viable building options, or to boost activity where not all of the grant is used. But it does not bring about an exact utilisation of exactly the right amount of grant in each time period. For this reason the model has to be used with some care, and more attention may need to be paid to this if the focus is on particular early time periods.
6.17. Chapter 2 discusses evidence from stakeholder interviews, particularly in the lending industry, which tends to suggest that not just the cost but also the terms of lending could become more restrictive, for example in terms of loan covenants and repayment horizons. It is not clear whether this might mean that our financial performance time criteria come to look too liberal. We looked at the particular issue of security for loan finance represented by the stock valuation in the group of 30 RSLs, which tend to be the more active group with higher levels of debt. By running a form of rolling LSVT-type valuation on the existing stock for this group we obtained estimates of the existing use 'social housing' value per dwelling, and compared this with existing and projected debt levels. This confirmed the case study RSL view (see Chapter 2) that for most landlords loan security is not a limiting factor in financial capacity.
Possible enhancements to the models
6.18. We have not so far attempted to model the impact of increased activity in terms of LCHO and intermediate rent provision, although there are some expectations of this increasing in the future. In England, profits from LCHO development have constituted a major part of surpluses in the sector, but these proved vulnerable in the post-2007 downturn. It would be possible to model increased intermediate sector activity and track its potential impact on capacity to build for rent, although this could be quite a complex addition to the model.
6.19. The current treatment of need as a possible filter on activity in some geographical areas is a defensible refinement to the basic financial model. It draws on a model commissioned by the Scottish Government which applies common assumptions across all of the Scottish local authorities, and projects forward over the same time periods which are of interest. This run of the model dates from 2007 and could therefore be updated. It deals with the need for additional affordable housing but does not attempt to estimate needs for new provision relating to regeneration and replacement of deficient stock. Our baseline capacity model does allow for continuing activity at recent rates in areas with low or no need for additional affordable provision.
6.20. Only for local authorities - and, even then, only in a fairly mild form - does the model currently allow for any form of rent convergence or restructuring. Any policy proposal for convergence or restructuring would raise many issues. It is also questionable whether such a restructuring would generate overall capacity estimates markedly different to those resulting from the uniform rent increases currently modelled, as long as the average increase was the same. The same comment would apply to any convergence in management and maintenance costs, unless this were conceived as a selective downward convergence.
6.21. Investment in improving energy efficiency post- SHQS may prove to be a substantial issue in Scotland. Our current assumptions make some limited allowance for this.
6.22. We have not made any special assumptions about the impact of Housing Benefit changes and prospective welfare reforms. The recently announced Housing Benefit changes may have some impact on rent arrears and write-offs. Future wider Housing Benefit reforms could have a larger impact. However, the recently announced limitations of Local Housing Allowance and Housing Benefit to the 30 th percentile level of private rents do not appear to come close to placing any additional ceiling on social rents, given that these are still a lot lower in Scotland than in England.