The data maze

What's going on in the supported housing sector? Are we really witnessing massive increases or decreases in provision for different groups as the figures seem to show? Each year RSLs produce their Regulatory and Statistical Returns - RSRs to the aficionados - for the Housing Corporation. These are then analysed by the Department of Land Economy at Cambridge University, where Christine Whitehead is based, and the results are then published under the label Dataspring.

A Dataspring discussion paper titled Snapshot of the RSL Sector in 1999, published by the Housing Corporation, seems at first sight to show some startling results (Housing Today, 5 October). "Supported housing for single homeless people increased by 45 per cent over the year" (from 1998-1999) "and accounted for over 25 per cent of all supported housing. The numbers of dwellings for those with HIV and AIDS and young people at risk or leaving care also increased, by 17 per cent and 13 per cent respectively. The numbers providing for older people and those with physical difficulties declined significantly by 26 per cent and 47 per cent respectively. The number accommodating refugees was also reduced by 31 per cent."

However, the discussion paper goes on to explain that behind the figures there have been changes in the ways in which data about supported housing is collected in the RSR. First of all, supported housing is now recorded by total units which includes all stock types combined, as opposed to being divided up between self contained units and hostel/shared housing bedspaces as before.

Secondly, in order to fall within supported housing status, housing provided by RSLs must meet three criteria set out by the Housing Corporation in their Performance Standards. These are that:

• A landlord/tenant relationship between the RSL and the person receiving support must have been established
  
• The level of support provided by the RSL must be over and above what would be provided in general needs housing
  
• Responsibility for providing support, either directly or indirectly must have been formally taken by the RSL

This means that some housing normally considered as supported housing, for example sheltered housing "with only the presence of a warden" does not meet all three criteria and is therefore categorised in the RSR and in the Dataspring analysis as general rented housing.

The situation is further complicated by the fact that prior to 1996 in the RSR there were two categories of older people, "elderly frail" and "elderly other". These were then replaced by the category "older people with support needs". In 1998 there was a reclassification of some of the housing for older people by several large RSLs which meant that the majority of stock classified as occupied by households requiring supported housing in 1997 were not in this category in 1998.

For people with physical disabilities, the supposed major reduction in unit numbers may be partly to do with adaptations to the general rented stock and the development of accessible and lifetime homes.

Where the trends appear to be going the other way, for example in supported housing for single homeless people, further changes to the RSR in 1999 may hold the key. For 1999 41 per cent of supported housing units "were described as general needs units where residents were receiving floating support on 31 March 1999". Since floating support relates to an individual living in their own home rather than being tied to a specific designated supported housing unit, the increase in units for homeless people may therefore relate to an increase in the number of homeless people receiving floating support rather than an increase in supported housing schemes for this group. In addition, for single homeless people, the high turnover of units, and of people receiving support, is likely to have pushed up the numbers significantly. The same might apply to young people at risk or leaving care.

For refugees, the figures show a nearly one-third reduction in the number of supported housing units despite reports of the number of refugees increasing. At first sight it is hard to understand how this can be.

So, what are the messages for RSLs from these figures? One is that, as ever, we need to be careful to read too much into figures, where the classification changes significantly from year. In fact the National Housing Federation believes that there are no huge swings in the supported housing sector at the present time.

A second, is that we do need to find more effective ways of being able to ascertain what changes in the actual pattern of provision, rather than changes in classification, are taking place. RSLs need to know this to ensure that they are building homes to meet today's and tomorrow's needs and aspirations rather than repeating outdated housing and service models that people in housing need no longer want. In this respect the underlying trends behind the figures are important. One of the lessons from the Rough Sleepers Initiative was that some hostels for homeless people remained empty because what homeless people wanted was a permanent home with flexible support tailored to meet their individual needs not another hostel with a time limited stay. The trend which these figures show towards floating support for homeless people living in general needs housing, is a sign that this lesson has been learned.

However, why has provision for refugees reduced? This surely needs more investigation and, at first sight does not reflect well on RSLs as landlords.

A third lesson is that the current classification of sheltered housing is confused. At a time when much of the sheltered housing sector is supporting an increasingly ageing population it seems misleading that changing the classification seems to show that the number of supported housing units for older people is declining. The main message here is that sheltered housing needs its own classification rather than being confused with general needs housing on one hand and supported housing on the other.

The fourth and perhaps the most important lesson however relates to the importance of good, consistent and intelligible data collection systems at both local and national levels, BEFORE Supporting People comes in.

If service planners and commissioners relate their estimates of need on current data for supported housing units and classifications it will make no, or only limited, sense. If classifications are regularly changed without clear explanations and linkages between before and after, then clear planning to meet future need will be difficult. It is vital, if Supporting People is to be a step forward, that the data maze is cracked. To rely on information from the RSR returns will not be good enough.