My point concerns the absolute and percentage size of the tenure codes; social rented from council and social rented other. The percentage figure in the above table for England was 9.4 for the former and 8.4 for the latter.
In England, according to April 2011 stock data there are more tenancies of 'housing associations' in common language (referred to as private registered providers under the Homes and Communities Agency social landlord definition) than council tenancies. The research and statistics sections of the Department of Communities and HCA will confirm this generally known fact.
Hundreds of thousands of tenants of private registered providers (or more commonly, housing associations) apparently ticked the box meant only for council tenants!
Two classic examples of very large discrepancies relate to the West Midlands districts. In Walsall in 2003 some 25,800 council stock was sold to the Walsall Housing Group (and the remainder of about 1,700 was also sold but to separate tenant management organisations, now managed by WATMOS). In 2000 in Coventry all the stock of about 22,000 council houses was sold to Whitefriars Housing. In Table KS402EW, for Walsall 10% of all households reported a local authority landlord while for Coventry 5.4% reported this. The numbers for social landlord other were under reported at 14.1 and 11.6 per cent respectively.
The 'accurate in relation to stock ownership criteria' figure would be 0% for each district if this new statistic can be imagined. The share of social rented other might then be increased to 24% and 17% respectively, based on stock ownership criteria to provide a more accurate picture of the tenure situation in these districts.
Between 1988 and 2011 there were over 200 transfers of housing stock between councils in England to registered housing associations, which then became the new owners of social housing. So for England up to 200 local authorities may have an inaccurate Census return for social housing tenure. Many of these tenants retained their council tenancy conditions but with a new landlord.
Whether the likely error related to a mistaken belief or to the wording on the Census form (which I do not think was entirely helpful, but it is a terminological minefield), the problem of 'mistakenly reported social landlord' applies throughout the entire Census record to varying degrees.
The two coded variables do now need to be corrected nationally from smaller territorial units by linking the stock information of social landlords nationally available to the 2011 Census findings. It's a human mistake identifiable at various geographical levels. Astute social researchers can learn much from the process of correction to reduce the error in the future. Walsall and Coventry both confirm my point applies for their housing stock, hence my confident tone.
Senior Lecturer in Housing
438a Dawson Building
School of Social Sciences
Birmingham City University
0121 331 5166