by Ken Gibb

There is much controversy about the use of explicit or implicit state loan guarantees in the contemporary housing system. My rather undramatic theme is that we should not disregard guarantees entirely and that they can play a useful role as a policy instrument in specific circumstances.

First of all, we should note that there are several variants of guarantees out there, chiefly:

  • State-backed guarantees associated with lending on new build private housing in order to support new supply, often targeted at first time buyers.
  • ‘Help to buy’ general guarantees to support any sales often with a limit on the mortgage involved but effectively encouraging lenders to offer loans at much higher loan to value ratios. In both this and the previous case, the lender risk is reduced because an element of any future negative equity would be protected by the guarantee.
  • Affordable housing guarantee schemes offer to reduce the cost of borrowing for social and private rented new supply projects. The Government will guarantee payment obligations on new debt provided there is offsetting lower grant per unit. This will apply in Scotland as well as in England.
  • Scottish National Housing Trust joint ventures guarantee void loss (i.e. loan repayments) and capital losses on the public sector investment part of the affordable housing joint venture with the private sector.

Since its announcement there has been a pretty consistent line of criticism that the second of these will induce some element of a housing bubble and there is much more skepticism about the worth therefore of such a policy or whether it can help to bring the market and supply up to pre 2007 levels on any sustained basis. There has been more muted criticism of the affordable housing guarantee scheme though in this case this has been as much to do with concerns over the private rented element, the fact that borrowing is already cheap for larger social providers (i.e. via bond finance) and delays before THFC were in a position to market the new product.

On the other hand, there is more support for policies aimed at promoting new supply and for first time buyers. The choice in this case of a guarantee instrument seems less controversial. Equally, the NHT model in Scotland, after a hesitant start, has gained some momentum, has the flexibility to work to different variations, and appears to offer good subsidy value for money (albeit for primarily affordable rather than social housing and for much shorter holding periods).

Guarantees are of course widely used elsewhere – Sweden and the Netherlands being important European examples, indicating different models are in use and that they can reduce borrowing costs. They are also used in other parts of the UK economy – e.g. the loan guarantee scheme for small and medium sized businesses run out of DBIS (now called the Enterprise Finance Guarantee). This is an explicit response to a market failure – and that is really the key. It is one thing to identify the market failure at hand; quite another to design and implement the appropriate policy instrument in response. In short, the key point is that good policies including such guarantees or indeed subsidies more broadly must be carefully designed.

What sort of criteria might be considered?

  • Targeting effectiveness. Who benefits and it is acceptably directed towards certain key groups and does so in a comprehensive fashion?
  • Consistency with wider policy goals. Does the policy complement other programmes and overarching goals; or, does it substitute for them and otherwise counter wider pre-announced higher-level objectives?
  • Value for money for scarce public resources – are the subsidy per unit costs sufficiently low comparatively and in absolute terms?
  • Minimizing unintended consequences. Does the subsidy become capitalized in prices, does it distort choices (e.g. through poverty traps) or otherwise lead to substantive impacts other than those planned (e.g. an artificial bubble)?

Price subsidies can, in principle, lead to too much housing being consumed. Tax breaks can easily be regressive and distort investment decisions between housing and other classes of asset, and, guarantees can also cause problems by lowering the cost of borrowing relative to lending costs for those activities that do not attract guarantees. So, we can easily have market and other distortions if subsidy/tax/guarantee design is not thought through both in terms of the problem at hand and its wider, general equilibrium, effects.

Guarantees therefore may or may not help. It depends on the specific context and how well they tackle the problem at hand and meet the kind of criteria outlined above. As part of a wider set of interventions and targeted towards reuse of partially completed or otherwise unused sites, it has potential through the National Housing Trust. It may also be a cost effective way at the margin of promoting new supply for entrants to home ownership (though 5% down payments may yet be too low as a stable, prudent ratio for potential first time buyers]. The design of help to buy is of course highly problematic. It is also the case that the constraints around the affordable housing guarantee scheme may make it less effective than Government hopes (e.g. the offsetting grant levels and the directing of the scheme to additional programmes which are likely to be more marginal – may reduce take-up].

A range of well-designed housing models and subsidy instruments can operate in different contexts to achieve a variety of valuable policy ends.  The key is the careful thinking through of its design and subsequent implementation, learning from experimentation and perhaps examples from elsewhere before committing fully. Cross these hurdles and there is no reason why one of the models should not be a form of state-backed loan guarantee.