Does it make sense to stop the Census?

By Jim Vine - on 11/12/2013

The Office of National Statistics (ONS) has opened up a consultation period to see if the 2011 Census should in fact be the last. The alternative option they present for future data gathering will not be able to deliver statistics down to the same fine-grained geographies that the current Census delivers. Our Community Insight mapping programme relies on detailed open data sources to help housing providers better target services in their communities. We have responded to the ONS with reasons why the Census needs to continue for the benefit of UK housing sector.

HACT's view of the potential alternate Census approach

Social housing – and, indeed, every other aspect of the housing system – is inherently long-term and inherently local. It also represents an area of substantial economic importance, in terms of both public money and private resources. Public spending directly on help with housing costs was £24.9 billion in 2010/11.  For 2011 the total balance sheet value of dwellings in the UK accounts was £4.3 trillion and annual consumption expenditure on housing was valued at £180 billion.  The right homes in the right places are also clearly an essential component of delivering both a society and an economy that works for individuals, for communities and for the nation as a whole. It is, therefore, vital that decisions about housing can be taken informed by high quality statistics that are comparable over time and are available down to small geographies.

In HACT’s experience, the loss of datasets reported down to OA (or at least LSOA) level would be of significant detriment to housing providers’ ability to plan both their core housing activities and their broader community investment work. It is a sign of housing providers’ appetite to make use of data that over 60 of the largest housing associations, owning and managing in excess of a million homes between them, have subscribed to Community Insight to gain easy access to key data sets (many drawn from the Census).

How housing providers benefit from the current Census detail of population and housing statistics 

As noted above, social housing is a sector that is hugely important, both socially and economically. Through tools like Community Insight, social housing providers are making increasing use of population and housing statistics to inform all aspects of their work.

The use of statistics in core housing activities, such as planning applications and regeneration plans, ensures that the provision of social housing is informed by the relevant context at the local level. However, given that housing markets operate over relatively large areas, it is perhaps in community investment activity where hyper-local statistics are of particular value. For most social housing providers, the neighbourhood or the estate is a natural unit for management and for working with communities. These providers are increasingly looking to maximise the impact of their resources, and are using statistics in support of this. Some providers are using statistics to identify the priority issues in each of their areas, so they can address the one or two most important issues in each area; others choose to focus their resources on a few ‘priority neighbourhoods’, which they identify through the use of statistics to highlight the areas with concentrations of problems. In either case, local level statistics are vital in ensuring that spending decisions are evidence-based rather than driven by impressions and gut reactions.

Furthermore, the provision of population and housing statistics at OA level underpins the development of tools that can report on arbitrary geographies without further input from the ONS (see below for further information).

Significant additional benefits of such detailed population and housing statistics for housing providers

Housing and population statistics are being used by social housing providers right now to make decisions that have the potential to improve life for communities around the UK. In one example, a Community Insight user in Wales used census statistics to identify that a hard-to-reach neighbourhood where they had been focusing on employability skills had an underlying problem of transport / accessibility; because few of the households owned cars (combined with local knowledge of limited public transport access to the nearest centre of employment) the housing provider identified that addressing the transport situation was a pre-requisite to getting people into work.

A significant further benefit of statistics being released at the OA level is that tools like Community Insight are able to build them up on-the-fly to make statistics for arbitrary larger geographies. In the housing sector, this has permitted housing providers to define at will hundreds of neighbourhoods that make sense for the people that live and work there. The ability to define an area and a few minutes later to have a detailed report on its key statistics has been revolutionary for housing providers, and an approach requiring ONS to build custom areas when requested (as described on p23 of the consultation paper) would inherently be vastly less responsive and more costly, and would be a very retrograde step.

The high-level impact if the most detailed statistics for very small geographic areas and small population groups were no longer available

For social housing providers operating in neighbourhoods and on estates, the ability to identify which issues are most important in a local area can underpin a range of decisions.  For the beneficiaries, this can be the difference between a tailored intervention that meets the need of a community, and one that has been generically developed, perhaps based on the issues that are most important at the local authority level.

One response to the removal of small geographical statistics might be for public and non-profit organisations to try and recreate them. Such an approach would inevitably be piecemeal, is highly likely to have poorer response rates than the census, and would not benefit from any of the economies of scale, resulting in far higher ‘per data item’ costs. Whilst this cost might be invisible in the Census budget line, overall nationally it would be a waste of resources to be paying more for poorer data, not just in terms of the direct additional resources to gather the data but also the wastage associated with making (worse) decisions based on weaker data. 

The social housing sector represents a very substantial set of assets that need to be deployed rationally, and makes significant investment in communities that can only be accurately targeted based on data at a hyper-local level. Given the scale of community investment recorded by housing associations – some £747 million in 2010/11 alone, representing billions over the 10 years of a Census – even a marginal improvement in the ability to deliver this investment where it can have the most effect would justify a large proportion of the cost of a Census on its own.

In housing as in other sectors, policy makers and regulators are increasingly concerned with ensuring that value for money (VFM) is delivered. The absence of local statistics will make it harder for organisations to fully engage with this agenda, as they would struggle to target their interventions where they would have the most impact; it would be harder both to maximise VFM and to demonstrate where VFM is being delivered.

If you would like to have your say to the ONS about how you use the current level of local Census data to help you, please get in touch before the 13 December 2013.


Add new comment