The numbers game

At the moment it is politically correct to state that the social housing numbers game is over and that the policy emphasis should be on strategies to cope with low demand and letting what is available. Yet No excuse not to build, published by Shelter, suggests that far more affordable homes are going to be required over the next decade. Why the difference?

In one sense the policy makers are right - although 25 years behind the times. It was the Labour government of 1974 which recognised that it was no longer true that providing any house anywhere would give value for money.

But at that time public housing was popular, and cheap, so problems were only identified in areas, such as Newcastle, where demand was very depressed. But the fact that there is low demand in some parts of the social sector does not detract from the need for additional affordable units.

The starting point for No Excuse Not to Build was the findings of the first Shelter report How Many Homes Will We Need? published last year and based on the 1992 household projections. This found that to accommodate newly forming households and to address the backlog of current unmet need would require some 115,000 additional affordable dwellings a year until at least 2011. The question the new report asks is "can this requirement be met from the existing stock or without the need for government subsidy?" It found that:

1. While vacancy rates are rising in some parts of the social sector, overall their level is still close to the feasible minimum. If vacancy rates could be reduced in areas where they are above average perhaps an extra 10,000 households in total could be housed - mostly in areas where demand is anyway low.

2. Difficult-to-let properties are not usually actually vacant - but do involve a lot of additional costs because of high turnover etc. But this is about obsolescence, and should be dealt with by effective management and investment in the existing stock. Such investment does not address the problem of newly arising needs. Indeed if, as seems likely, some parts of the social stock are reaching the end of their useful life - because of physical attributes, location or neighbourhood - more units will be required.

3. Meeting need for additional affordable housing through the private sector necessarily entails either reducing the vacancy rate, converting existing property to take more households or new building. The vacancy rate in the private sector is not high by international standards and cannot readily be reduced by policy measures. Reducing the size of units by conversion is not a long-term viable strategy as incomes, and thus demand for space, rise. The conversion of commercial buildings is limited to central London and some other inner-city areas. It is little more than a niche market and requires at least the usual levels of subsidy if affordable homes are to be provided. The existing stock can produce very few additional affordable units.

4. That leaves new building - and, as social landlords are all too aware, it is rare indeed to be able to make new units affordable without subsidy. The only available policy at the present time is to provide cross-subsidy through planning gain, which is shown so far to have produced relatively few extra units, many of which anyway involve additional government subsidy.

5. An enormous gap, between what the government is helping to provide and what is needed therefore remains.

The two main objections to this case are:

• The new 1996 household projections, published after the report went to press, are lower than those from 1992, so perhaps there will be fewer homes required for additional households.
  
• Surely the issue can be addressed through the private rented sector with the assistance of housing benefit.

Certainly, the 1996 projections reduce the overall number of households expected to form between 1991 and 2016 from 4.4 million to 4.1 million. But most of the fall is in areas where it is already obvious that demand will grow slowly. The total numbers in the Southern half of the country, where the majority of the need is located, have not been reduced.

Equally, there is evidence that households are moving out of the social sector more rapidly than expected in the early 1990s, which could reduce the requirement for new units in that sector. But this could as easily increase the need for affordable homes because obsolescent and badly located dwellings must be replaced. Taken together a very conservative estimate of the need for additional affordable units might be only around 100,000 dwellings a year.

At present, the shortfall compared to actual social sector provision is being met from private renting with the assistance of housing benefit. But this does not address the need for additional units. If they are not provided directly, they can only come through reducing standards (not an option over the longer term) or by additional market sector building to replace existing units transferred to affordable provision in the private rented sector. But this implies increasing house prices to generate additional building activity especially in areas of high demand - putting up the cost of housing and probably increasing the proportion of households requiring assistance.

The only alternative scenario is that the incomes of those currently in need of assistance rise faster than house prices and rents and that the income distribution improves so that they can pay for adequate housing themselves.

If this were the reason why fewer than 100,000 affordable units per annum turn out to be required Shelter would cheer. But on the evidence so far this scenario seems unlikely.