The first door is opened by an elderly woman, barefoot in a long cotton robe. Some chopped vegetables have been laid out neatly on a tea towel in the lounge. The room smells of the kitchen. The cricket is on TV, and she fumbles with the remote control for something less embarrassing, inadvertently putting on an adult channel instead. She hasn't yet spoken when the phone rings, and then she talks in Gujarati for a good 10 minutes.
Her visitor, the woman sitting on the sofa doing her best not to look at the TV, is Jill Masterman, an interviewer for the Office for National Statistics – the government department that publishes all those stories that tell us all about ourselves. How happy we are. How healthy. How much we earn, work, what we believe, how well we brush our teeth. Masterman wants to ask the woman about her life and opinions. She has been randomly selected from a list of addresses themselves randomly selected from a randomly selected postcode. (No one can volunteer to answer questions for the ONS: that would indicate a predisposition to being surveyed, which is the opposite of random.)
At last the phone conversation closes, and the interview begins. Masterman asks the ages of the woman's children. It is a straightforward question, but the woman shakes her head sadly. "Long time they are gone from here." At what age did she leave school? "I married when I was 18," she says. Her long fingers hang over the arm of the chair, scratch at the fabric.
So did she leave school long before she married? She rests her head on her hand. "Can't remember, can't remember," she says softly. The telephone rings again. A woman, in spectacles, is on TV now ("Text a foxy secretary"). Eventually the second phone call ends, and Masterman asks: "How satisfied are you with your life today?" This appears to be the final straw. "If your husband isn't here, if you live alone …" the woman tries. And then, "Leave it. I can't do anything." She asks to be allowed to sleep. It has taken 45 minutes to complete a fraction of the interview, and the prospects of ever finishing it seem slim, but Masterman tells the woman she'll come back another day. Every record is worth perseverance. The more she gathers, the greater the value of the information.
As a society, we are continually measuring and being measured. Counting is a defence against the uncertain or unexpected – a deterrent against another economic crash, perhaps, or a forewarning of social unrest. It is a way of keeping step with a social landscape that is not only changing fast but whose very idea of fast is perpetually outpacing itself.
![Household crime graphic](https://web.archive.org/web/20130802084523im_/http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2013/7/31/1375292949453/Household-crime-graphic-010.jpg)
The sight of a woman trawling through sorrows for a long lost date may seem an unlikely place to start, but the data collected from this and other doorsteps will help to form a picture of the way we are, and what we think about all sorts of issues, from the sharing of personal information with supermarkets to the introduction of fees in family courts, which will be fed back to the government and, who knows, influence policy. Some may challenge its statistics, but if you live in Britain there is no such thing as a day untouched by the ONS. In the last month alone it has revealed that more babies will be born out of wedlock than in by 2016, that one in four people living in Britain's largest cities are immigrants, that we are happier than we were a year ago, and that there was no double-dip recession after all.
Its research underpins the statistics that underpin our lives. If your house or car insurance rises (owing to new crime figures), if the government decides that the majority of public opinion is in favour of gay marriage, it is probably because of research by the ONS. But what do the people behind the statistics look like, and what is the process that turns their lives, snapped in situations as surreal as the one above, into data? The ONS is habitually secretive, because it takes the confidentiality of its interviewees extremely seriously, but for the first time it has agreed to be observed at work.
So here is Masterman, on the fringes of Southampton. She has arranged to interview a young woman, but it is a boy in school uniform who answers. His sister isn't home "and she hant texted my mum", he says. It is the first of the day's many no-shows or no thank-yous. The next door opens, and a cloud of captive cigarette smoke heads for the exit. "She's out," the man says. "I don't live here." And so the day goes on, Masterman driving around in her car with the sticker that reads "I don't suffer from insanity, I enjoy every minute of it."
Demand for information of the kind Masterman gathers is growing. "We are surveyed more often now," says Patrick Sturgis, professor of research methodology at the University of Southampton. "It is partly to do with the internet and the development of infrastructure to do this easily. We are asked more and more questions in different contexts." He has no citation but quips: "You could do a survey asking people if they have been surveyed." This is not so farfetched. The ONS canvasses opinion on how best to phrase its survey questions. (It has ditched the word "nowadays" because young people found it fogeyish.)
![G2ONS](https://web.archive.org/web/20130802084523im_/http://static.guim.co.uk/sys-images/Guardian/Pix/maps_and_graphs/2013/8/1/1375360561836/G2_ons_preg.gif)
By the end of the day Masterman has completed only one full survey, of a student who was 70% happy. At the last house there is no answer, but a knocking noise is coming from the garage. Masterman calls out and the door rolls up noisily. "I'm a bit busy," the man says, chalking his cue. This seems a bare-faced fib, until he explains that he is practising for pool trials. "And I have recently had a hair transplant," he says. He wants to help another day. They make an appointment, exchange numbers, and the exchange feels like a promise, the numbers seal the deal. After all, most people have a faith in numbers that exceeds their faith in words. Data journalism is often held to tell a truer story.
This trust in numbers may help to explain why the language used to talk about data has been naturalised. It gets scattered and harvested, gathered, cut and cleaned. Professor Luciano Floridi (Hertfordshire and Oxford), author of The Philosophy of Information, says there is an area of research that looks at information gathering as "information foraging", as if data really were the good of the land, freely available and nutritious to anyone who knows where to look (such as Twitter and Google Trends, which the ONS now tracks in order to supplement and counter-check its own reckonings). "We are no longer journeying round this strange tropical place known as cybersphere," Floridi says. "In the 90s the metaphor was the frontier – surfing, exploring. It was all about seeing the web as a geographical space, not a personal space. Because we are settling down, biological metaphors become more natural. We are natives here now."
At the ONS's sprawling cellular base camp in Newport, the corridors are long and all the doors bear long numbers, a decimalised warren. There is a number for everything here. In the Life Events – Mortality Analysis department a book, two inches thick, lists the code for every kind of demise – W27 means death by contact with a nonpowered handtool, L60.0 is an ingrowing nail, T63.4 a bite by a centipede, and so on. From birth to death, our lives are codified, digitised. Why?
"Data helps you make sense of a really complicated world. It helps you understand what's really going on, and separate that from anecdote, from spin, from misinformation," says Glen Watson, director general of the ONS. As he sees it, British lives are built upon the work of the ONS. The thickness of water pipes that deliver your morning shower; the arrangement of roads that take you where you are going; the location of schools, offices, supermarkets; the stock on supermarket shelves. His faith in the ONS is personal as well as professional. "I might – touch wood I haven't been – at some point in my life I might be diagnosed with cancer," he says. "The first thing I'd want to know, I'd go straight to the ONS page on cancer survival rates. We publish all of that." When he talks about data, he cups his hands as if he has caught a butterfly.
![Old people graphic](https://web.archive.org/web/20130802084523im_/http://static.guim.co.uk/sys-images/Guardian/Pix/pictures/2013/7/31/1375293112052/Old-people-graphic-001.jpg)
In the Labour Force Survey (LFS) office, an analyst is looking at the latest figures on a spreadsheet. Row upon row of digits fill the screen. Each row is a person. Each column is their answer to a question. There are 76 columns, but they would go on for ever if you wanted them to. "You can't fall off the edge of the world here," says the analyst, Mark Chandler. Whatever an interviewee's answer, there is a code for it. Person 1 has given the answer 31 to a question, which means they are retired from paid work. Someone else is a 2: self-employed. Here's a "25" – not in work, looking after their family or home.
These answers have been collected by interviewers like Masterman, and by the ONS call centre in Titchfield, Hampshire. Listening in to telephone conversation after conversation, you can hear people combing the breadth of their circumstances in order to check the right box. Which part of this gentleman's body is worse for arthritis? "Take your pick," he says. The plastic of his handset creaks as he gets comfortable. "Today it's the shoulder. It could be the knee, the arms, the legs, the hands."
One operator is wrestling with the fact that a man doesn't know his daughter's qualifications. A woman on another line is saying: "She's dead, my lovely." In Bournemouth, Liverpool, Skelmersdale, answerphones are clicking in. Another operator is trying to find out if a woman is looking for work. "It's a difficult question, because if the job is rewarding I would like to work," the woman says. So is that a yes or a no? "A no," she decides, because no one should have to work until they are 70. "It has been known to find people dead on the toilet at work when they're 65," she says. There is no box for that.
It is a funnelling process, a sifting through a lifetime of all that's made a person the person they are, shaking out every irrelevance, until the only thing left in the sieve is an affirmative or negative. To listen in to those conversations, and subsequently to see fragments of them pop up in an enormous grid of numbers, is a bit like witnessing people being fed into a giant Hadron Collider: lives churned and turned into statistics, a great big crunching and mashing (to use a favourite word of data analysts), a sort of digital dismemberment. You half-expect to hear screams, see a pool cue fly out and some veg on a tea towel, the spectre of a dead person on a toilet. And what is this great collider creating? What is the output of these weeks of combing streets and dialling numbers in search of co-operative people, and the subsequent weeks of chopping and reading the information they give?
The analysts have a printout showing employment. "It's going to add a dot here," says Chandler. He points to a space beyond the last little sphere in the sequence. This is where the new speck will go, small as a particle, vast as a miniaturised planet. And then, of course, the dot will be cut in myriad ways and replotted, scattering new solar systems across countless graphs.
"It sounds a bit creepy. A bit Orwellian, when you put it like that," says Sturgis. "But of course these are anonymous answers. We are not interested in individual cases. What we are doing is making a map."
But is it a double-edged comfort, that for all we are being surveyed, the surveyors are interested in only the type of individual? On one hand, Floridi says, "You can hide behind typologies. There is safety in that new sort of anonymity, which is no longer anonymity but opaqueness behind proxies which pick you up well enough, but which don't really identify you as unique." But there is danger in this kind of anonymity too. "Next thing you know, because you are treating everyone as a kind of person, it becomes part of our culture. I fear that it might be changing our social interaction. And that would be sad."
If Floridi is correct, people will be kinds of people – figures – not only in public, on social media or in the files of administrative data, but in private encounters too and, perhaps, even in their own estimations. Data collection has many useful purposes – it is not hard to see all those digits as the building blocks of our society - but it is also, in some degree, an exercise in collective vanity. As a public, we enjoy looking in the mirror that is held up to us. As individuals, we like to spot ourselves in the picture, find ourselves on a distribution line: are you more or less happy than the national average of 7.4 out of 10? All these charts, tables and graphs are society's selfies, snapshots taken from so many angles. Clearly we can't get enough of them; they are infinitely fascinating, and may help us to know our world better than ever before – so long as we don't lose ourselves looking.