Do we need a census when we have Facebook?

This month, we are being ordered to fill in our census forms. But why should we help the state force

For ten years between 1991 and 2001, officially I didn't exist. I remember receiving the lengthy form for the 1991 census as a student, having
recently arrived in London from Belfast. But those were the days of the hated poll tax and, like many young people, I didn't trust the motives of government statisticians. Why did they want to know all this stuff about who I was? What were they going to do with it? As there didn't seem to be any pressing reason to tell them, I didn't bother to send back the form.

Once a decade, the government has a go at counting everyone who lives in England and Wales. This time, the date falls on 27 March. The results are always interesting, and not only for the data. Since its origins in Malthusian concerns about population growth in 1801, the census has been a good way of measuring the fears and priorities of those in power, as well as the public's attitudes to those powers. It is a flagship national project and the biggest peacetime exercise carried out in the UK - the jewel in the crown of the Office for National Statistics (ONS).

It is also a feat of logistics. This year, by posting out forms and collecting data from 25 million households simultaneously, 40,000 field collectors will try to put together an authoritative, everyone-at-the-same-time snapshot of who is living here. It doesn't come cheap. The estimated total cost of this census will be £482m, up from £254m in 2001.

In many ways, it remains good value. The data it collates on where we are, how old we are and how we live is enormously useful for researchers and planners of all kinds. It is with census data that central government works out how much money to give local authorities and where schools, hospitals and roads should be built. Charities and religious groups pore over it to decide how to allocate grants or where to open churches; companies use it to work out where to build supermarkets and houses; academics and demographers look to it as the definitive record of shifts in the make-up of the population. The census tells us who we are and how we have changed - and it creates a conversation around this information.

The last one, for example, confirmed that our society was ageing. It informed us that, for the first time, there were more people over the age of 60 than there were children. This resulted in a wave of media interest in everything from Harley-Davidsons to day spas.

Yet this year's census is likely to be our last. For some time now, there has been talk that its methods are ineffective and a little passé. In July last year, Francis Maude, the Paymaster General, said the census was "out of date almost before it has been done" and that the cash-conscious coalition government was looking around for cheaper alternatives.

In 2008, a Treasury select committee report concluded that the census was largely obsolete and that patching together various databases held by bodies such as the Department for Work and Pensions could track population movements more cheaply and effectively than census collectors. Because the data held by government departments and on electronic databases managed by credit-checking agencies is constantly changing, the argument goes, the information that they generate could make for a more responsive population count. Other countries, the committee noted, are moving away from weighty, once-a-decade questionnaires. Finland has a "population register", made up of the data its citizens leave behind in their brushes with national government, forms they fill in and bills they pay. Last year's US census was one of the shortest in the country's history. The data collectors shrank their questionnaire to just ten questions; to add colour and texture, they combined the answers with data from a survey that is posted to a quarter of a million US households each month.

None of these alternatives, however, produces quite the kind of information that is delivered by a full census. Academics are excited about census data because it is not worked up from a small sample of the population; it is a full questionnaire, in which everyone is asked the same set of questions at more or less the same time. To many of its defenders, it is also a weapon in the armoury of the modern welfare state. Governments still rely on finely grained data to dole out resources to the areas and groups that need them most.

The coalition is in a bind. It would love to save money by replacing the census with administrative data, but it knows that we already see it as part of an all-knowing Big Brother state - and we are suspicious of what it is doing with our data. This makes it harder for the government to justify sharing information between departments. But such scepticism is not entirely justified. The information that the census collects can't be passed around to any other department; it remains confidential for 100 years.

A bad case of ticks

The first box I remember ticking was "White Irish". It was on an equal opportunities form for a job managing a hostel for homeless families in Southwark, London, which I filled in three months after the 1991 census. I got the job. Southwark Council had good reason to collect information on my ethnic origins; the borough had long had a large and spirited Irish immigrant community and earlier generations had been discriminated against, so it made sense to keep an eye on how many Irish were getting jobs or failing to get them. But when officials identify people by ethnicity and fit them into boxes, they risk forcing them into stereotypes and colluding with those who like to define them as a species apart.

Many of the people living in that Southwark hostel had only recently arrived in the UK. Among them was a young Turkish woman who was on the run from a husband who had tried to kill her. I had been informed by her caseworker that she didn't speak English but then, after three months of watching my attempts to communicate with her using elaborate mime, she burst out laughing and confided that her English was almost perfect. The competition for permanent council housing was intense and she was simply trying to do everything possible to secure a home for herself and her son. She wanted to tick all the boxes.

As the politics of class has diminished, officialdom has become increasingly and intrusively concerned with understanding who we are and how we live. In its heyday, our allegiance to the welfare state was rooted in a contract forged between social classes. Without this, the state has been obliged to hunt around for what political scientists call "new cleavages", or new ways to understand us.

The question about ethnic origin was first included in the 1991 census. The point was to take into account the special needs of ethnic-minority groups and monitor racial disadvantage. Then, in the 2001 census, the collectors moved on to asking about religion. The reason for this was to find subgroups of the same ethnicity but different religions and to get a better understanding of how they wished to be identified. This year, we will be given the oppor­tunity to go beyond religion and ethnicity and define our national identity. In something of a sop to devolution, we will be presented with a choice of "Welsh", "Northern Irish", "British", "English", "Scottish" or "other".

All this box-ticking racks up. A century ago, the census fitted on a single sheet, but by 1991 it had grown to 34 questions. The following decade there were 41, and this year there will be 56. Census collectors in the US reckon that it takes ten minutes to fill in their short form. To fill in ours will take about an hour.

As the government has become more curious about who we are, we have become increasingly resistant to answering its questions. It is estimated that in 1991, the year I gave census collectors the slip, a million people joined me and did not show up in the final count. Nearly a million went missing in 2001, too, defying a huge publicity campaign to "Count me in", but this time the data collectors were ready. They buttressed their main census with an extensive sampling exercise a month later to get a handle on how many they had missed.

There is no doubt that many of those who weren't counted in the last census were migrants who had recently arrived in the country and simply didn't want to be counted. But it is easy and lazy to blame those at the margins. It's not just immigrants who shy away from census collectors. Many of us live in homes in large blocks or with entryphones, making it difficult for them to find out who lives behind each door. We move around a lot more for work or pleasure, flitting from one place to another and in and out of the country. Supporters of the census say that one of the frustrating things about the process is that those who are most likely to evade the census collectors - inner-city ethnic-minority populations and young, fitfully employed men living in cities - are the very people who stand to lose out most if their living circumstances don't show up in the data.

Just as ominously for the questionnaire sociologists, many of those who do fill in the form baulk at the boxes prepared for them. In 2009, when researchers from the Institute for Public Policy Research (IPPR) confronted a sample survey with the ethnic categories listed in the 2001 census, most respondents felt that they were a laughably inadequate way of describing who they were. The subsequent report, You Can't Put Me in a Box, stated that "young Britons in particular seem not to care much for tick-box approaches to identity".

Many of those who were interviewed seemed to be rebelling actively against the classifications the authorities were foisting on them in favour ofmore fluid identities of their own choosing. While they agreed that some aspects of one's identity were given or inherited, it was important to them "that their identity was something in which they had a choice and that that choice was a free choice". They appeared to be resisting religious classification, too.

Shortly before the 2001 England and Wales census was collected, an email was circulated suggesting that if 10,000 Britons identified their religion as Jedi Knight - the fictional faith espoused in the Star Wars films - the government would have to recognise it as a religion. It wasn't strictly true, but that didn't stop more than 390,000 respondents from doing it anyway, making Jedi the fourth-largest religion in the countries. Something similar happened in New Zealand the same year and sizeable Jedi populations have since identified themselves in Australia and Canada.

It's not hard to guess their motives. "Do it because you love Star Wars," the email concluded. "If not . . . then just do it to annoy people." This year, the Jedi will be joined by hard rockers who are campaigning to have heavy metal listed as a religion. At the time of writing, the Facebook group Heavy Metal for the 2011 Census had more than 33,000 followers.

Suspicious minds

Campaigns for silly religions are more than just wrecking exercises. They speak of our need to be part of something, preferably something that kicks against the categories suggested by the census. The young Britons interviewed by IPPR may have felt boxed in by the classifications offered by the authorities, but this didn't stop them devising boxes of their own.

The only problem with the official categories, they felt, was that they were "so broad and impersonal as to preclude or hinder any sense of who they thought they were as a unique individual". When invited to, they were happy to replace them with categories of their own. When asked by the researchers to describe themselves, they were more likely to name their values or personal characteristics than
demographic attributes. "Rather than 'male, 32, Scottish', people told us they were 'warm, bright, funny'; less census category, more personal ad," the report's authors said.

Our enthusiasm for identifying ourselves in different ways should hardly be surprising. Our reluctance to be categorised by the census has gone hand in hand with an increased willingness to divulge information about ourselves in other ways. On our Facebook profile pages, we rush to choose our own categories, answering questions in quirky ways that we feel comfortable with and then bunching together around what we like. It is as if we have become our own questionnaire sociologists, profiling ourselves according to name, age, gender and relationship status before gleefully compiling lists of our interests. We seem to be happiest when filling out empty boxes and sculpting an identity for ourselves. Everything is constantly updated, and whenever any of it changes, the news is broadcast to everyone we know.

Mostly we seem to be telling the truth. Chris DeWolfe, the co-creator of MySpace, reckons that 98 per cent of his American users correctly identify where they live and report important changes, such as getting married or moving house, soon after they happen. For the most part, however, it is the same people who are throwing their data around online who are reluctant to give it to the state. During last year's census in the United States, return rates were initially so low in a few of the hipper areas of New York that it gave rise to a media panic. Perhaps, noted a journalist, the people were too busy tweeting.

It could be that we are tiring of the demands of the questionnaire sociologists. Many of the questions we are asked in polls and surveys feel loaded and a little cheap. Online, it is easier to wriggle out of other people's categories and express who we are in the way we want. The census allows for this: we can fill in the boxes labelled "other". On the net, however, almost everything is "other". The boxes we like to fill in are those that give us free rein to write what we please. It makes it easier for the watchers, too. There is less need for them to ask us questions when we are telling them everything online. Blue-chip companies can afford to buy sophisticated databases that tell them what we are buying and where; our supermarket loyalty cards, too, make it easier for companies to follow us. They can track us even more closely using the grains of information that we leave behind as we navigate the internet.

Like big business, the state seems to know that its audience has been slipping away for several decades. Without much in the way of mediating institutions between the state and the public it has lost, officials are desperate to get a better hold on who we are. But government often looks like a poor relation compared to the corporations, with their vaults of credit-card data and armies of researchers. It is doing its best to catch up. The Beyond 2011 Project, a working group within the ONS, is looking at the future of the census and examining the alternatives. This year, we will be able to submit our census data online. But the problems with the census are not just about technology - they also mirror the decline in social solidarity. In essence, they are a symptom of a quiet falling-out between the public and the public authorities; a mismatch between the ways in which they want to understand us and the ways in which we wish to understand ourselves.

We complete our profiles so carefully on Facebook because they allow us to define ourselves as we want, and for those we wish to be seen by. The result is a vast ocean of information. This is data of a kind that questionnaire sociologists can only dream about and it could yield fascinating insights. It can't provide the hard data that governments need to know, but it doesn't seek to force us into rigid categories, either.

This time, now that I'm older and a little more responsible, I will send back the census form. But I am still suspicious about why the government's data collectors need to ask so many questions and I'm no more enthusiastic about the way in which they want to understand me. That, in the end, should tell them more about national identity than the tick box they will give me to express it.

James Harkin's "Niche: Why the Market No Longer Favours the Mainstream" is newly published by Little, Brown (£20)