By Peter Ewart

(Click here for Part 1 and Part 2 in this series)

An angry, older man charged into a Target store in the U.S. demanding to know why the store kept sending ads featuring baby items to his teenage daughter. Was Target trying to influence his daughter to get pregnant? A month later, the man called back to apologize. It turns out his teenage daughter really was pregnant (1).

What crystal ball did Target polish to figure this out? “Predictive analytics” it is called. And a Target statistician, Andrew Pole, was an expert at it.

It is well known in business that new parents are “a retailer’s holy grail.” However, once the news is out that a baby is born, parents are barraged with ads and incentives from dozens of other retailers. The question that Pole tackled was: How to figure out a woman was pregnant early on in her pregnancy, and thus capture her as a customer for years to come. If Target had this information, it would provide a huge leg up on its competition.

For years, Target had been collecting information from customers on everything they bought. In addition, it gathered demographic data about each customer (age, marital status, location of home, estimated salary, and so on) using a variety of means, as well as purchased huge swacks of information from commercial data brokers about each customer’s credit card purchases, magazine subscriptions, web browsing history, car and house ownership, political leanings, and so on.

What Andrew Pole happened to notice was that women tended to buy large quantities of certain products during the various stages of pregnancy. Using statistical data and mathematical algorithms, he identified about 25 products that, when analyzed together with other data, “allowed him to assign each shopper a ‘pregnancy prediction’ score. More important, he could also estimate her due date to within a small window, so Target could send coupons timed to very specific stages of her pregnancy” (1).

The result was a bonanza of sales for Target, and a lot of catch-up for its competitors. As Eric Siegel, a consultant with Predictive Analytics World, boasts: “We’re living through a golden age of behavioral research. It’s amazing how much we can figure out about how people think now” (1).

The lesson has not been lost on the big political parties in the U.S. (and increasingly so in Canada) who are engaged in a virtual arms race to gather all sorts of personal information on voters, crunch the data sets using mathematical algorithms and statistical methodologies, predict individual voter intentions, and develop election strategies and techniques accordingly. Within that context, knowing whether or not a voter has actually cast a vote at a particular poll becomes hugely important, as it will impact what techniques the political party will utilize towards the voter.

And so, the way things are going with this unregulated, wild west style of information-gathering, not only will political parties know whether you actually voted, they will know how you voted with a great deal of statistical certainty.

This poses serious problems for the sanctity of the secret ballot, as well as for the entire democratic process. As the 2012 report from the Privacy Commissioner of Canada office points out: “If communications with elected officials are merged with voter data, the potential for differential treatment based on political support is a serious risk that could undermine the integrity and fairness of our representative system” (2).

To put the above quote in plainer language, armed with this data, your MP or MLA or government could potentially screw you if you vote the wrong way. Interestingly, the report also notes that, “on the other hand, there is nothing in current Canadian law that prevents this form of data-sharing or data linkage.”

This kind of activity has happened before in Canada. Back in the 1970s, the Liberal Party of Canada held the reins of power of the federal government. Under its rule, the RCMP broke into the offices of a several political parties and organizations in Quebec (including the Parti Quebecois), and stole the membership lists. We do not know the full extent as to how this information was used by the federal government to target these members, but at the very least it was a gross violation of their privacy rights as citizens.

Nowadays, breaking into an office and determining which citizens are supporting a particular party, is becoming unnecessary. Instead, through information gathering, Big Data, and predictive analysis, political parties and governments can accomplish the same result. This can create a chill on political discourse and dissuade people from speaking out, organizing or voting a certain way, and thus threatens the democratic process.

One thing for sure – people do not like all their personal data being vacuumed up by party machines and their privacy seriously compromised by predictive analytics and other invasive techniques. Target found that out fast. According to Andrew Pole, the company had to start camouflaging their baby ads because pregnant women were “getting queasy” and freaked out about how much Target knew about them right down to the specific stage of pregnancy. The company did this by mixing in other kinds of products, like lawnmowers and wine glasses, so that the inclusion of baby products looked random (1).

As the Privacy Commissioner report relates, the Conservative Party of Canada also found out the hard way. In October of 2011, the Prime Minister’s Office sent out Rosh Hashanah cards “to supporters with Jewish sounding names,” i.e. assuming that they were of the Jewish religion. This prompted an outcry as to why the PMO would have such information in its database. It was particularly objectionable to a section of the population that had undergone discrimination and persecution in the past.

What is important in all of this is that we, as voters, become aware of the sophisticated techniques that political parties, especially the big ones, are using to gather data, manipulate voters, and gain power in the 21^st Century. After all, it is our rights that are at stake.

Which brings us to the next topic in this series. Stay tuned for Part 4 – “Microtargeting.”

Peter Ewart is a columnist and writer based in Prince George, British Columbia. He can be reached at: peter.ewart@shaw.ca

(1) Duhigg, Charles. “How companies learn your secrets.” New York Times Magazine. Feb. 16, 2012.

(2) Bennett, Colin J. “Canadian federal political parties and personal privacy protection: a comparative analysis.” Office of the Privacy Commissioner of Canada. March 28, 2012.