Sense and census necessity

John Perry

As a wealth of data continues to pour out from the 2011 census, we need to remind ourselves how valuable it is. Government suggestions that this census might be the last one are nonsensical and should be resisted.

Not long after the last census took place, stories began to emerge that the government believed that the £500 million cost simply wasn’t worth it. It was even suggested that the 2021 census could be replaced by scrutiny of supermarket store cards. Government was later reported to be looking for ‘more effective, less bureaucratic’ ways of collecting data. These moves were announced before the detailed results even began to emerge. Will the obvious usefulness of the data now being released help to change their minds?

Local government had reason to be sceptical about the census in the wake of the 2001 results. In that year, it under-counted the population and under-represented immigration. As the census coincided with a surge of new migration, councils quickly got frustrated that their pleas for extra help to provide services to growing populations fell on deaf ears: old data were constantly cited by government as grounds for maintaining the status quo.

The Office for National Statistics (ONS) seems to have learnt its lessons. The 2011 data came out quickly, the coverage of the latest census is extraordinarily high and the results flesh out in great detail the trends that we’ve known about more generally from other surveys. Most important of all, whereas sources like the International Passenger Survey log inward and outward movements of people on a regular basis for the UK as a whole, the census tells us exactly where they come from and go to, which is the information that local service providers need.

Here’s an example of how the data can be used. HACT and OCSI have developed a free resource for housing providers, called Population Insight. It takes highly detailed data on diverse communities and immigration and provides it in a form that housing providers can tailor to show the detail for local areas or even particular streets. For example, one of the headline changes across Britain has been the growth in the foreign-born population from seven per cent in 2001 to 13 per cent in 2011. But some places are hardly touched by immigration while others have been transformed. Take Boston in Lincolnshire: it was 98 per cent ‘white British’ in 2001 but now over ten per cent of the population is Polish. The council and social landlords in Boston need to know where the Polish and other incomers actually live, and whether their representation in the housing stock is typical of the area or not. To respond further, they may want to know how many households in a neighbourhood don’t speak English. Population Insight processes census data to tell them this.

Here’s another example. Important government forecasts, such as projections of the future growth in household numbers, are based on inter-census surveys and other records. The ten-yearly census enables statisticians to rebase their figures. The current projections for England use estimates of numbers of households in 2008; the next set early this year will be able to take account of the census’s finding that household numbers are actually lower than was thought. Now, if ministers aren’t keen on the census they are nevertheless keen on using these projections, as the gap between household growth (currently projected at 232,000 per year) and new housebuilding (barely topping 100,000) is regularly cited. Don’t they therefore have a strong interest in the accuracy of these figures?

It beggars belief that supposedly cheaper measures like scrutinising storecards can provide similar high-quality data to those from the census. We risk entering an epoch when we no longer know about our own population with any accuracy, since people who don’t have storecards, bank accounts or other accessible information, or who have recently arrived from abroad, could be missed out. In any case, information will be reduced to that which is commercially viable for private operators to record and provide.

Danny Dorling, a geographer who has never been afraid to use census and other data to criticise governments, has sketched out how the dropping of future censuses might affect local authorities.

The way that Dorling and other researchers make use of the census often uncovers inconvenient truths. But however uncomfortable, we ought to insist that we still need to know about them.

John Perry is an associate at HACT. He writes extensively on housing issues and is a Policy Advisor to the Chartered Institute of Housing.

This article first appeared on Public Finance opinion blog

What did the census feeding frenzy tell us about housing?

What can we find for housing in the midst of the census “feeding frenzy”?

Oxford Consultants for Social Inclusion’s Director Tom Smith says that up to half a million people in social housing do not know their landlord is a housing association.

Image: Image Broker/Rex Features

The December publication of the latest data from the 2011 census kicked off a veritable feeding frenzy of analysis from media pundits, thinktanks, researchers, public service providers and even local pub drinkers. Headlines focused on the impact of migration over the last decade, with Oxford University’s migration observatorythe IPPR and Manchester University leading from the front.

However, there are also some big social issues hidden in the housing data. The census confirms the impact of two of the biggest housing stories over the past decade: the rise of private renting and the fall in owner occupation, and the scale of the transfer from council-owned social housing to housing associations. There is also a lesson in relying on census data, where other administrative data sources can be used for cross-referencing.

• Growth of the private rented sector and fall in owner-occupation, driven by population increases

Results from the census show the dramatic increase in the scale of private renting, by almost 50% over a decade. Between 2001 to 2011, the number of private-rented households increased by 1.63 million in England and Wales, from 12% of all housing in 2001 to 18% in 2011 (if you include in this sector those living rent free). Unsurprisingly, given difficulties of accessing mortgage finance, the proportion of people owning their own homes has fallen significantly. Although still the biggest component of all housing, accounting for 64% of all housing in 2011, owner occupation has fallen from 69% of all housing in 2001.

This is further evidence of the scale of what has been described as Generation Rent. Although the total number of households in England and Wales increased by 1.7 million over the decade, 95% of this increase was in the private renting sector.

Local variations are associated with rapid population growth. The national figure hides some big local variations. Across London, less than half of households are now owner-occupied (49.5%); the capital also saw the largest regional fall in owner-occupation levels over the decade; from 56.5% to 49.5%. In Hackney, only one in four (26%) of households were owner-occupied.

The biggest changes are seen in those local authority (LA) areas with the biggest population growth. The local authorities with the largest population increases are also those authorities with the biggest shift towards private renting, and away from owner-occupation. In the 10 authorities with the largest population increases over the decade, the number of private renting households increased by 139,000 over the decade, but owner-occupied households only by 5,600.

The data from the census underscores the challenges facing central and local government and the housing sector. The increased pressure on housing through population growth, low levels of house building, and difficulty in accessing mortgages, is leading to big shifts in how we live. Generation Rent is here to stay.

• Census data shows – but undercounts – the shift in social housing

The second significant story seen in the census housing data is the scale of the shift from council to registered social landlord housing. In total, 18% of all households were self-reported as renting social housing in 2011, similar to the 19% in 2001. But within these figures, the impact of stock transfers from councils to housing providers is clear. The proportion of social housing now managed by registered social landlords is significantly up, with that owned by local authorities correspondingly down.

However, the census undercounts the true scale of the shift. In 2011, census data identified 1.8m properties run by registered social landlords in England, well below the 2.3m identified in housing provider returns to the Tenant Services Authority. Similarly, the 2.1m council owned properties from 2011 data is well above the 1.7m identified in Department of Communities and Local Government data.

In other words, up to half a million people in the census appear to have misreported their housing as council owned rather than registered social landlord. This is understandable: many tenants may well continue to see their house as “council housing” even after stock transfers – and it is unlikely that the data on ownership and private-renting is similarly skewed. But it highlights the importance of cross-referencing, or “triangulating” in the research jargon, census data against other sources where available. Although the census is the most detailed information source we have on the state of the UK, we can’t simply take the data as gospel.

More detailed data from the census will be published on 31 January 2013 and will be available to housing providers under the Community Insight tool developed by OCSI and the housing charity HACT.


Original article can be found on the Guardian Housing Network

Click here to view a copy of the full OCSI Census 2011 review on housing