Surveying Your Identity Or Surveillance

Arguments for a digital census

“From the time of his first term of office he had established personal contact with the population, visiting each of the settlers at census time, entering their homes and discussing with them their needs and aspirations.” — Jean Talon, 1665

Every morning, I wake up to the din of newsletters and junk email from hundreds of sites I’d once registered with. With this long digital trail of mine, relatively few details are stored—at most the amalgam of my digital census would resolve few things about me, my full name, gender, ethnicity, sexual orientation, and birthday and maybe a list of places I’ve lived. If there were something big about such big data stores, it certainly isn’t personalization but repetition. My search history, location data and status updates might offer some other clues into my behavior and identity but though objective, it would be unreliable.

I make a concerted effort to not be tracked, not out of any sense of paranoia but out of privacy and pricing. Given how much data apps farm while running in the background, I’ve turned off Background App Refresh and Location Services on my iPhone. I’m not paying my cellular bill so apps can sell my data to interested third parties.

Self Knowing

By now, we recognize the value of accurate anonymized data in helping inform policy and improve well-being. From John Snow, who mapped cholera outbreaks in London in the mid 1800s to the Boston Department of Transportation using smart phone accelerometers to map potholes or “street bumps” in their city, there are methods to gather data for sociological needs that are beneficial and non-intrusive.

And the public is enchanted by such findings, from the decline of crime correlating to abortions to the rise in death rates for middle-aged white people relating to substance abuse and suicide. Exactly 350 years since Jean Talon introduced the first Canadian census, we’ve argued for information gathering as a method of national self-knowing. So in this age of widespread digital data collection, what if any societal value is derived from ad tracking and surveillance?

Stephanie, a queer friend and data scientist recently told me about an exhibition she’s designing entitled, Historical Museum of Binary. The installation suggests that in the near future, we would display cultural artifacts like memes and surveys, that point to our limited and binary identity today. I’d agree. My short form profile contains binary fields—young, male & non-white. And linked to this imprecise identity, my spending, browsing and other digital habits are tracked over the course of time. This imprecision weakens the significance of logged user behaviors that often I’m served completely targeted ads with unsurprising irrelevance.

Responding to lapses in my user profile, Facebook tried to fill in data gaps relating to gender and began offering wider selections for gender and relationship status beyond a binary or primary. And though only a small percentage use one of the 70 “Other” gender designations, GLAAD President Sarah Kate Ellis recognizes “this new feature is a step forward in recognizing transgender people and allows them to tell their authentic story in their own words.” The user can then decide who they share their gender with. It seems a first step in proper representation while offering the platform visibility and behavioral insight to tackle harassment. I asked @lvhntr, a trans friend what she thought of the whole implementation:

“The somewhat offensive college-era holdover ‘it’s complicated’ is still an option, with an option to say who ‘it’s complicated’ with, but nothing that allows you to specify multiple meaningful relationships with different people. It still says to me that Facebook is ultimately about an assimilation into whatever it’s paradigm of ‘real experiences’ are.”

If Facebook offers a crude approximation of identity, the human data driven matchmaking service OkCupid fares a bit better on relationship options. Sean Rad, founder of Tinder, would be happy to know that there is a sapiosexual designation. Their data has surfaced our inherent racisms and other stereotypes of dating by sorting through quiz responses and favoriting habits. As Christian Rudder, co-founder of OkCupid suggests, the data is “making the ineffable totally [fuck]able.”

Identity API

Our ability to judge identity appears evolutionarily hardwired where even as infants, we’re able to recognize age and gender, while distinguishing race appears slightly more dependent on exposure. But in a world where hundreds of identities collide every moment, how useful are our naturally evolved instant judgements. To counter discrimination against women and others, research has found blind auditions improve recruiting outcomes in symphonies and tech corporations, suggesting that our human firmware is outdated for our social economy. If our perception is patently flawed, and our digital census imprecise, how might long-form digital profiles improve conditions relating to identity? As it were, hiring is biased and diversity is problematic in its underrepresentation of women and minorities. Worse, the imprecision suggests that Asians fare better, but Asia is expansive and just two nationalities are represented in technology, Indians and Chinese.

Even as the demographic of users online widens, surveillance remains problematic and such intrusions on our privacy, from government surveillance to identity theft pose a challenge to wider census taking, However, I’m not talking about background data collection from a cellular backdoor or a location app.

There is an important distinction between tracking behaviors, the sort that surface “walnut moments” gleaned from your day to day web use, and voluntarily entering anonymized personal information into secure stores, an API for your identity. I’d hope we are not far off from having an open authorization that allows you to choose which parts of your identity you wish to share or make public when you logon to services, perhaps just your alias and gender but not nationality. Anonymizing is still quite hard and sometimes reversible, as we found out from the researchers who reversed identified users from anonymized Netflix records using IMDb. I’d like to believe that it is possible to gather deep data while retaining anonymity, and moreover, that such data on macro identity can inform social and platform policy.

The early extensions of gender and sexuality profile suggests a digital future where our identity is cataloged beyond binaries towards something of higher fidelity. Just as audio and visual sampling has increased in bitrate to levels surpassing human cognition, we should expect that identity and behavioral sampling will at some time approximate the true diversity and variation found in life. Lest we stay convinced that there are only two genders, male-female; two orientations, gay-straight; or two races; white-black. The Historical Museum of Binary will look back at societal artifacts from our present day with a sense of awe at how primitive our digital censuses were.

I spoke to Navdeep Bains, the Sikh Canadian Minister of Innovation and Technology who helped restore the mandatory long form census in Canada. As he correctly pointed out as, we live in a world driven by data based decisions and without accurate and widespread information, our policy making frameworks are simply ideological gambles. Former Prime Minister Stephen Harper’s government had scrapped the census to much uproar, and given the gap in data, we don’t have a way of knowing how his policies have impacted the population.

In 2012, Estonia surpassed Canada with their e-census, which appears more complex in scope and widespread at 66% participation. Their government is continuing work on anonymization and visualization of data with hopes that in the future, the census itself will be unnecessary, rather updated from other government service agencies. While precise data is not tantamount to the hard work of sociological change, it is a necessary requisite in spelling out present conditions. Countries and companies paving the way for improved census and decision making suggest a future where diversity and wellbeing is less a slogan and much more a science.