Everybody’s going on about big data these days, from the Guardian’s Sustainable Business podcast (http://www.theguardian.com/sustainable-business/big-data-sustainability-podcast) to the Financial Times (http://www.ft.com/intl/cms/s/2/21a6e7d8-b479-11e3-a09a-00144feabdc0.html#axzz2xXga2LUN) to AT&T (http://adage.com/article/cmo-strategy/t-big-data-shape-tv-creative/292313/).
IBM estimates that 2.5 quintillion bytes of data are generated every day and 90% of all the information in the world was created over the last two years (http://www.carbontrust.com/news/2013/10/big-business-big-data-big-sustainability). Le Monde declares that this is the moment “When Mathematicians Became Sexy” (http://lesclesdedemain.lemonde.fr/organisations/big-data-quand-le-statisticien-devient-sexy_a-12-1769.html).
Beyond the buzz, some critical questions are being asked: will big data truly make governments and firms more alert to consumers’ needs? How will its energy use affect the environment? And will it undermine privacy? I’ll focus on the last of these issues.
In a sense, big data is not really news. It began with the census—local, regional, and national attempts to find out who lived where, what they did, and whom they did it with (if you get my meaning). The first census may have been taken in ancient Egypt, China has had one for millennia, imperial Rome used them to track people who were liable for military service—and Jesus was supposedly born on census day.
The census is designed to permit large institutions to know (and control) the population. That was as true for churches in the first century as it is for companies in the 21st. And governments’ social programs are best devised and targeted to alleviate suffering and inequality when they are based on the living circumstances of people.
The census is no simple matter. Whereas the US variety classifies race as a social and political construct, it regards sex as “real,” in the same sense as houses or salaries (https://www.census.gov/population/age/). This is in keeping with gender’s place in the earliest meditations on citizenship. Aristotle reasoned that as ‘the state is made up of households, before speaking of the state we must speak of the management of the household,’ which he defined as the relationships of masters to slaves, husbands to wives, and parents to children (http://classics.mit.edu/Aristotle/politics.1.one.html). Hmm.
Ethnic categories characterize the Census as much as gender. The US divides people into five groups: White, Black or African American, American Indian or Alaska Native, Asian, and Native Hawaiian or Other Pacific Islander (https://www.census.gov/population/race/). But these ways of understanding populations have become less and less useable, because many people don’t identify with the categories on offer, mistrust the state, and choose not to provide the data requested.
The increased number and variety of people with Spanish-language descent has particularly compromised official racial divisions. The Census Bureau says: ‘People who identify their origin as Hispanic, Latino, or Spanish may be of any race.’ But many [email protected] even reject ‘Hispanic’ as a description. The best way to describe people who origin can lie in many countries, and who do not necessarily speak languages other than English, is far from obvious (http://www.census.gov/population/hispanic/).
In the 2000 Census, 42% of [email protected] selected ‘some other race’ as the category that best illustrated their social situation. But their ambivalence about this category has proven relatively powerless next to the desire of activists and corporations to identify a single group worthy of either welfare or consumption. Miami used to be cluttered with bumper sticks that read ‘Don’t Call Me Hispanic. I’m Cuban!’ Now, not so much.
The other main difficulty facing the Census is Republicans. Terrified by secular truth and loss of political power (in reverse order) they insist that our national Census be conducted purely by an empiricist head count rather than more accurate statistical extrapolations. Because that militates against including the transient working-class and ethnically marked folks, who tend to vote Democrat, it helps solidify the gerrymandered electoral system that has enabled recent Republican majorities in the House of Representatives.
Despite these drawbacks, the Census provides a useful snapshot of society. Whereas 80% of children grew up with their married, biological parents in the 1950s, that was true of just 12% of children by the end of the 1980s: 7% lived with an employed father and “home-duties” mother. The 2000 Census disclosed that heterosexual married couples with children were just 25% of the population. To the horror of anxious Christians, these numbers reveal that the straight family is a contingent form of life. The more strident and politically active that evangelicals become, the more such families decay.
Meanwhile, new, private systems of population surveillance and analysis further complicate the Census. The amount of data, the speed with which it is collected and analysed, and the way it is articulated to individuals have all increased in their reach and effect. The Census is just one game in town next to vast numbers of companies and divisions that are dedicated to obtaining information about us (http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html?pagewanted=all&_r=1&hp).
Target was rightly embarrassed by the 2012 revelation that it analysed purchasing patterns by women to determine whether they were pregnant, then proceeded to advertise pregnancy and baby products through direct mailing to their homes. That risked disclosing their situation to people from whom they might wish to keep such matters private, be they parents, grandparents, children, partners, or lodgers (http://culturalstudies.podbean.com/2014/03/11/anna-crowe-ken-page-privacy-international/).
The Federal Trade Commission alerted citizens to how much surveillance they were subject to in the name of consumer sovereignty in a 2012 report (http://www.ftc.gov/reports/protecting-consumer-privacy-era-rapid-change-recommendations-businesses-policymakers). The Commission claimed that if consumers gave companies information about themselves, their desires would be met more easily and rapidly, but acknowledged legitimate concerns about the use of such data by firms. Yet it failed to point out that beyond privacy lay the ownership and use of ordinary people’s intellectual property: their ideas and identities should be theirs to share or not, for monetary gain or not. So it’s not just that they must be allowed to keep things private—they should be paid by corporations that wish to collect, analyze, exploit, and sell information about them.
The same applies to those of us who play on-line games. In addition to paying subscription to corporations, we are frequently obliged as a condition of playing to sign away the intellectual property we create. Companies can exploit our avatars, moves, speech, and so on for free—in fact, we pay them to do so!
So the big news about big data is that we need to slow the process down: slow down its march into our lives, and slow down the use of information by companies to invade our privacy and use our intellectual property.