Shopping for Votes: How Politicians Choose Us and We Choose Them (35 page)

Rove was eventually sold on the idea and Republicans plunged deep into this brave new world of “micro-targeting.” Six months and $3.25 million later, amassing thousands of “data points,” they realized that the old way of sorting voters, according to their income or neighbourhood, wasn’t all that helpful. Matthew Dowd, the chief campaign strategist for the Bush team in the 2004 presidential election, gave an illuminating interview to PBS radio’s
Frontline
about how consumer and lifestyle data turned old political assumptions upside down.

“People used to sort of divide it by income, and it was like, lower-middle-class, poor people were Democrats; upper-middle-class, rich people were Republicans. That’s sort of gone… You could be a wealthy Democrat as easily as you could be a wealthy Republican today, or you could be a lower-middle-class Democrat as [easily as] you could be a lower-middle-class Republican,” Dowd explained. “What we learned early on in 2001 and 2002, through some analysis, was that 85 percent of Republican voters don’t live in Republican precincts… It means there’s a whole bunch of Republicans that live in traditionally Democratic precincts around this country, and the only way to find them is individual profiles or calling or doing all that sort of things.”

Magazine subscription lists turned out to be a wealth of information for the Republican micro-targeters, Dowd told PBS. “Their social or household habits will tell you a lot. If somebody gets
Field and Stream
, they’re much more likely to be a Republican voter than a Democratic voter. If somebody gets
Mother Jones
, they’re much more likely to be a Democratic voter, or
Rolling Stone
, for example, they’re much more likely to be a Democratic voter.” The same thing was true with TV viewership. “Somebody that watches
CSI
is much more likely to be a Republican. Somebody that watches a soap during the day is much more likely to be a Democrat. All of these things aren’t completely 100 percent true, but when you go through it all and you factor it all out, you can find Republicans fairly quickly that way.”

Gage sorted voters into clusters such as “Flag and Family Republicans” or “Tax and Terrorism Moderates,” which would then help Republican strategists sort where to direct their direct mail and persuasion efforts. And if some marketing-type tweaks to the Bush platform were required, the Republicans obliged. Twenty years after Margaret Thatcher and Ronald Reagan had first demonstrated how to shape a political product to suit the consumer-citizen, it was now common practice for all political parties in the US and Britain. Micro-targeting advocates would argue that this high level of attention to voters’ concerns was a way to put democracy back into elite-driven politics in the US. “In 2000, we very broadly talked to people on broad issues,” Bush’s campaign manager, Ken Mehlman, told the
Washington Post
. “In 2004, instead of talking about what we thought was most important, we talked about what the voters thought was most important.” Bush won re-election in 2004, with the help of constituencies his campaign had singled out for special treatment, and so the fervour for micro-targeting got under way.

The Democrats, meanwhile, had been plunging into the online world themselves, especially under the party chairmanship of Howard Dean. As a one-time presidential hopeful in the early 2000s, Dean had led the way for other politicians of all stripes in the field of online fundraising—at one point, raising more than $800,000 in a single day in 2003. As chairman of the party, Dean also gathered around him an array of internet-savvy advisers, who used the early fundraising successes as a jumping-off point to revolutionize data-gathering for the Democrats and send the first black president, Barack Obama, to the White House in 2008.

Seeing what Rove and the Republicans had accomplished with micro-targeting, especially when it came to voter turnout, the Democrats pulled together a company called Catalist, which assembled masses of data from progressive-leaning organizations and individuals in the United States. The basic idea was the same: if the Democrats could use tiny data points to identify voters who could be swayed to their side, the campaign organizations would know where to concentrate their efforts and, most importantly, how to propel those people to the polls on election day.

On top of this huge database of information on potential voters, the Democrats applied some old-fashioned knowledge of persuasion. Veteran political hands know that voters are most likely to be persuaded by people who think like them or move in the same circles. And this is where the Democrats’ huge outreach on social media, such as Facebook, also played a role. Through social media, people reveal their friends and their tastes online, making it all the easier for politicians to tailor their appeals around them. What’s more, the Democrats could set up chains of persuasion, assigning committed voters in one social group (an environmental organization, for instance) to lure on-the-fence voters (consumers of “green” products, for example) to Obama’s team.

The respected magazine
The Atlantic
devoted a three-part series in 2009 to a dissection of “How the Democrats Won the Data War,” including a confidential report from Catalist. It showed that in turnout alone, micro-targeting had pulled off miracles for Obama. People who were reached through Catalist’s data network showed a turnout rate of 75 percent, compared to the 60 percent national average. And the outreach was impressive: Catalist made contact with nearly 50 million voters about 127 million times before the votes were cast in November 2008. Not only was this a powerful partisan argument in favour of micro-targeting, but a democratic one, too. What’s not to like about any system that could boost people’s participation in the most fundamental of civic exercises?

Voter turnout actually went up in the United States overall in the early part of the twenty-first century, a sharp contrast from Canada, where turnout was in decline at the same time. About 54 percent of the voting-age population cast a ballot in the 2000 presidential election and by the 2008 vote that figure had climbed to 62 percent, according to the Center for the Study of the American Electorate. In contrast, Canada’s voter turnout fell below 60 percent in the 2008 election for the first time in more than a century. In the 2012 presidential election, America’s black voters voted in higher proportions than white voters—a historical first, and a sign of the supremacy of the Democrats’ mastery of micro-targeting and data analysis.

US author Sasha Issenberg released a book in 2012,
The Victory Lab
, that gave Americans their first close look at how the data wars were transforming the modern art of politics. In a blog post he penned for the
New York Times
just before the presidential campaign, he lamented how journalists were still covering “horse-race” polls and developments in elections, when the real strategy was being conducted with the use of reams of data and analysis that came down to vote-by-vote persuasion. “Over the last decade, almost entirely out of view, campaigns have modernized their techniques in such a way that nearly every member of the political press now lacks the specialized expertise to interpret what’s going on,” Issenberg wrote in a September opinion piece on the
Times
website. Fittingly, he chose a consuming metaphor: “It’s as if restaurant critics remained oblivious to a generation’s worth of new chefs’ tools and techniques and persisted in describing every dish that came out of the kitchen as either ‘grilled’ or ‘broiled.’”

The Victory Lab
was also an illustration of just how much the American politicos’ preoccupation with data had merged people’s identities as citizens and consumers. Their consumer data and preferences were being fed into big-data machines in a bid to influence their civic preferences. Issenberg wrote of how Obama, even as he campaigned for the Democratic nomination in 2008, was relying on the help of a firm called Strategic Telemetry, which was matching voter information with as many as eight hundred “consumer variables” to come up with targets for support in the Iowa caucus. Knowing how Iowans shopped, in other words, was crucial intelligence in knowing how and whether they would vote for Obama.

And once you’ve started comparing people’s political preferences to their shopping choices, you can come up with no end of fun findings, too. A firm called Scarborough Research learned that drinkers of Samuel Adams beer were more likely to vote for Mitt Romney for president in the 2012 presidential election, while Heineken drinkers would probably vote for Obama. Another consumer-research firm, Experian Simmons, learned that Democrats were more likely to shop at Bloomingdale’s, while Republicans shopped in greater numbers at Belk department stores. Democratic women were more likely to buy shoes at Lady Footlocker, while Republicans favoured Naturalizer stores.

The revelations and lessons in Issenberg’s book, meanwhile, were well noted in Canadian political circles. Issenberg paid a couple of calls on Canada after his book was released: first to a sellout crowd in Toronto in November 2012 and then to a roomful of conservative thinkers and strategists at the Manning Centre’s annual conference in Ottawa in March 2013.

Stephen Taylor, one of the most ardent advocates of digital politicking among Canadian Conservatives, appeared on the panel with Issenberg at the Manning Centre event. In Taylor’s view, no party had a choice in the modern era—voters were getting harder and harder to reach, because they were too busy, too tuned out of politics or simply not answering the phone. Gone were the days when you could figure out people’s political preferences with simple surveys. Taylor, who spent a lot of time studying digital campaigning techniques in the United States, believed there was a democratic argument to be made for these leaps and bounds in digital political marketing. “If politicians are able to connect with people directly on issues that they care about, one could argue that’s closer to the ideal of democracy,” Taylor said. “Data brings politicians closer to having an accurate mirror of the electorate, and that’s something we should strive toward.”

The presidential elections in the United States through the first decade of the twenty-first century, then, would give Canadian politicos yet more lessons in how to do marketing in politics, following a pattern that went all the way back to the previous century. The marketing world would come up with efforts to reach customers better; then, the American political world would borrow and adapt those lessons; and finally, Canadians would import the techniques north of the forty-ninth parallel. The pattern would remain the same for this digital overhaul of campaign methods. Now, though, Canadian students of modern electioneering would require an eye for very tiny details.

Canada, in Miniature

The arrival of postal codes in Canada in the 1970s paved the way for the huge marketing databases of the future, by sorting Canadian households into tidy, small geographic units. Through the 1980s, advertisers and non-profit groups were increasingly using direct mail and postal codes to reach their preferred potential customers or donors. Credit card companies could target wealthier neighbourhoods to send invitations for sought-after customers, who were able to pay the premium rates they were offering to prestige clients. Magazines could gather up subscription lists and then sell them to companies who might want to know where readers with specialized interests lived. If you manufactured bicycles, for instance, you might want to know who was subscribing to cycling magazines. Then came personal computers, followed shortly by email, which produced a whole new set of addresses and contact information for the population. The clever marketers in these decades would be the ones who knew how to sort this information for their own purposes.

In Canada, one of the first big firms in the data game was one called Compusearch Market and Social Research, which billed itself as the country’s “premier provider of geo-referenced demographic, consumer, and business information solutions.” Bursting into the marketplace in the early 1980s, Compusearch could supply would-be entrepreneurs with information about their potential customers. Someone wanting to open a restaurant, for instance, could ask Compusearch for an analysis of the best neighbourhoods in which to situate their new venture. And the data Compusearch was accumulating was also particularly useful to the booming direct-mail marketing business in Canada at the time. Sales from direct-mail ads had doubled between 1984 and 1989, from $4 billion to $8 billion, and the sellers with the most detailed consumer-contact information stood to reap the biggest profits. In a 1990 article in the
Toronto Star
, Compusearch’s director Michele Sexsmith explained that the firm gathered up information from Statistics Canada and even some municipal and tax records. The researchers then sorted this information to group households by age, income, education and so on. That information, in turn, was used to create direct-mail marketing lists, which were traded among list brokers “like so many baseball cards,” according to the
Star
story.

As technology grew more sophisticated, and the demand grew for sharper pictures of the consumer marketplace, more of these firms started popping up in the United States and Canada. Environics Analytics emerged as a big player in this business in 2003, boasting when it started up that it could sort Canadians into sixty-six lifestyle types with playful names such as “Cosmopolitan Elite,” “Electric Avenues,” “Les Chics” and “Lunch at Tim’s.” A branch of the long-established Environics polling firm founded by Michael Adams, this company’s micro-analysis promised to be a much more vivid picture of the marketplace than the kind provided by traditional pollsters. “For years, we’ve been working with clients to help them understand the social values of their customers. My books and articles are written from about 35,000 feet,” Adams said when Environics Analytics launched. “Social values segmentation takes us down to 1,500 [feet].”