Shopping for Votes: How Politicians Choose Us and We Choose Them (36 page)

No matter how impressive these new Canadian databases were, however, they lacked much of the detailed information available in the United States. In the first place, the United States had a much richer consumer marketplace from which to mine this information. As well, Canada’s privacy laws made it almost impossible to obtain data on an individual, household level. So while an American marketer could look at the databases to get precise pictures of specific households, the big Canadian marketing databases were limited, by and large, to postal codes.

Canadian politicos, with rare exceptions, also lagged behind the Americans in micro-targeting. A few smart Canadian political operatives had already been dabbling in data. John Laschinger, for example, had been an early pioneer in accumulating political-marketing information when he started “prospecting” private-sector lists in the 1980s to accumulate lists for direct-mail targets for the Progressive Conservatives. The arrival of the permanent voters list in 1993, especially in digital form, saw the Liberals adapting software so they could manipulate the voters list for their own fundraising and get-out-the-vote efforts. The Liberals started to buy demographic maps from Compusearch in 1993, to feed the statistics into their own data about individual ridings and voters in local constituencies. For about $2,000 each, MPs received demographic profiles of their ridings as well as a computer program they could use to link the Compusearch maps to their own rudimentary databases. This particular milestone in data-accumulation politics only came to public notice when news emerged of sixty-one MPs paying for the Compusearch information out of their House of Commons budgets. The MPs protested that these maps helped them do their jobs better, communicating with constituents, but it was also true that the Liberals had election goals in mind. Liberal MP Ronald MacDonald told reporters that the Compusearch maps would help fine-tune the messages that the party wanted to get out to voters, as well as help determine which households should be receiving their literature.

But databases didn’t really become a major player in Canadian politics until around 2004, when the merger of the old Progressive Conservative party and the Canadian Alliance transformed the machinery of modern election campaigning in this country. The credit for the Conservatives’ early database plan has to go to Tom Flanagan, who was serving as Harper’s chief of staff before and after the merger.

The old Reform Party had organized its fundraising around small donations and in the process had accumulated the contact information for thousands and thousands of individual citizens. Gathered the old-fashioned way, with local poll captains and boots on the ground, this data gave Reformers an intimate knowledge of their local strength. These were their friends and neighbours, after all, many in rural areas. Reform’s entire grassroots mentality, heavy on populism, was well suited to gathering the kind of information that would prove essential to a modern database. Through waves and waves of pamphlets and postcards, they obtained valuable addresses of and contact information for potential donors and voters. As the party grew, and morphed into the Canadian Alliance under leader Stockwell Day, so too grew the list of donors and members. But the voter-ID and fundraising system, such as it is existed, was only as strong as the volunteers on the ground.

By the time Harper was running for the Canadian Alliance leadership in 2002, the party had about 300,000 names on its rolls, including anyone who had been a member of the Reform Party, dating as far back as 1987. With the help of campaign manager Doug Finley and his understanding of marketing technology, Harper and his team first used this database to raise money for his leadership campaign—all told, more than $1 million from 9,500 donors, whose average individual donation was about $116.

After Harper won the leadership, Flanagan realized the value of having an ongoing database, one with more information than simply supporters’ names, addresses and phone numbers. For all their skills in gathering the data, the Reform Party and then the Alliance had been in the habit of losing information in between elections, forcing them to start from scratch each time. By happy coincidence, the Canadian consumer world was also seeing a rapid expansion in professionals who were in the business of matching consumers to their data. One of them was an outfit called the Responsive Marketing Group. RMG specialized in direct marketing to help businesses and charities raise money; it had also worked with the Ontario Conservatives to set up an early database known as Trackright. (In 2009, RMG would be honoured by the Manning Centre for Building Democracy for raising more than $75 million for right-wing causes across Canada.)

Michael Davis, then head of RMG, went to Ottawa to sell Harper’s office on the idea of a systemized database, such as the one it had built for the Ontario PCs. Jim Armour, then Harper’s communications chief, sat in on one of the sessions and like the others was dazzled by the possibilities of this technology. Armour said he felt a bit how our prehistoric ancestors must have felt when they saw the first caveman walking with a torch. Suddenly, a whole new world was opened up to the Canadian Alliance team—a world in which they could quickly and efficiently find the voters they needed, through electronic sorting of phone numbers, addresses and contact information. Instead of relying on their small army of foot soldiers on the ground, they could assemble all this information in one big database.

With the help of RMG’s data-building skills, the Canadian Alliance started to put together the information colossus that came to be known as CIMS: the Constituent Information Management System. It’s pronounced just like
Sims
, the video game in which you build lives and lifestyles for simulated characters, winning points for how well you tend to their needs and desires (just like modern marketing in politics, you might say).

In those early days, CIMS was well fed, with information and big money. The Conservatives were spending as much money on CIMS as they were on advertising, in fact. And yes, this machine was being built by the same party that was leading a national crusade against the Liberals’ gun registry—the same party that would oppose the mandatory long-form census. Although Conservatives would rail against the “intrusive” government taking names and keeping lists, they were remarkably sanguine about their own political party doing the same—with much more information and fewer controls over the way it was gathered.

For privacy reasons, though, the Conservatives tightly guarded this information, even from each other. Only people at the highest level of the Conservative pecking order had access to the full array of information in this massive Fort Knox of voter contact information. Lower down the chain, partisans were given passwords that opened the door only to the information that had been stored about their local area or polling districts.

Reportedly, the quality of information in CIMS varied from riding to riding, depending on how diligently it had been collected by MPs or volunteers. CBC TV reporter Keith Boag was offered a rare glimpse into the database for a documentary he did called “The Data Game” in 2007. Boag gave the CIMS technicians the address and phone number of one of CBC’s producers. A frowning face appeared by producer Sylvia Thompson’s name; she had apparently been brusque with Conservative callers on some earlier occasion, so she was not viewed as a likely future prospect for funds or support.

The CIMS database was built on a wide variety of information sources. It started, naturally, with the national voters list, the names and addresses of every eligible voter in the country. On this basic layer of data, the Conservatives added information they gained from door-knocking, party membership lists, phone-bank efforts and national and local surveys. The basic Elections Canada list has no phone numbers or birthdates attached; that vital information could be gathered manually, or by purchasing phonebook-style lists from data vendors. Everyone, from Conservative members of Parliament to candidates to canvassers, was expected to log information into the system. Well-trained volunteers were also instructed to keep an eye out for toys in the yard or small things such as a National Association of Stock Car Racing (NASCAR) sticker in the window of a pickup truck in the driveway. “We train our canvassers to take note of things that are of interest to us, in terms of them being able to segment [the voter information],” Richard Ciano, a longtime Ontario Progressive Conservative who was elected president of the provincial party in 2012, said in an interview.

CIMS kept track of who had donated or volunteered for the party in the past, and also names and contact information of anyone who had attended partisan events in the ridings. It noted whether you had a lawn sign, either for the Conservatives or for one of its rival parties. If you were generally seen as friendly to the Conservatives, a little smiley-face icon appeared beside your name in the CIMS database. If you angrily hung up on a Conservative seeking your support or slammed a door in a campaign worker’s face, a little frowning face was likely beside your name. If you had written a letter to the party condemning a policy issue—the gun registry or funding for the CBC, to cite two popular causes—that too was tracked. And of course, CIMS kept records of people who were undecided or leaning toward the party—and those who were not. CIMS could sort and manipulate all this data quickly to help Conservatives figure out where to fire up their base of support, and where they needed to apply some extra effort. And the more information the party accumulated, the more sophisticated and “micro-targeted” its efforts would become.

Through a CIMS list, the party knew where to send fundraising letters that mentioned the gun registry or the CBC, for instance. CIMS could also generate digital lists of phone numbers, which could then be supplied to companies such as RMG for Conservatives’ telemarketing efforts. During election campaigns, the Conservatives divided up this telemarketing work among a number of companies, not just RMG. The other Conservative-friendly firms in this business over the years included Campaign Research, Voter Track and Margaret Kool Marketing Inc. Other parties had their own favourite firms to do the same sort of thing. The Liberals used First Contact Voter Management and Voter Identification Solutions Inc., for instance.

Ciano, in addition to his political work, also helped run Campaign Research, a firm that probed consumer and political preferences, sometimes in the same survey. “Things like hunting, fishing, gun ownership, boating, outdoor sports, NASCAR, major league sports—those are all things that are tracked as part of consumer behaviour,” Ciano explained. “We can index that against polling information... That kind of thing is extremely useful to us in prospecting for support and donations.” As an example, Ciano said that his surveys might turn up the fact that people who attend National Hockey League games several times a year, or people who buy a certain type of clock, are more likely to vote Conservative. The party then would try to seek out commercially available lists to reach these consumers of hockey or clocks. It’s not just the collecting of the data that’s important, in other words—it’s the
analysis
that makes those connections between shopping and voting.

This was the business that politicians of all stripes had scrambled to keep exempt from the do-not-call list in 2005, in the dying days of Paul Martin’s government. And the databases existed in a legal grey area with respect to privacy. Unlike commercial firms in Canada, political parties were under no obligation to tell members of the public what information it was storing about them, nor did they require people’s permission to accumulate data on them. So these companies had a free hand, more or less, to keep dialling into the homes of Canadians, often through automated waves of calls, nicknamed “robocalls,” and to keep amassing intelligence in the process.

Like the pollsters of earlier eras, these telemarketing firms made the lion’s share of their money in the commercial world, but their reputation in the political world. And they served as yet another bridge between the consumer and civic worlds—bringing private-sector expertise into the practice of democracy. Technology was making this possible to a degree that previous political marketers never could have imagined.

Stephen Taylor was building online communities for the party in the early 2000s—before, as he says, anyone had heard of phrases like “social media.” Taylor founded a website called “Blogging Tories” that served as a one-stop shop for online Conservative opinion blogs. As the party’s database became more sophisticated, so did Taylor’s focus. He would mine through online data to find out who supported conservative and libertarian issues, to see what data-based characteristics they had in common. An adherent of the “always be testing” school of digital campaigning, Taylor pored through the analytics to see what online methods work best in getting people to vote, speak out or donate to the Conservatives. Passionate about this new online frontier, Taylor talks the language of digital marketers, sprinkling terms like “metrics,” “URLs” and “split-testing” throughout his explanations of his methods.

Split-testing, incidentally, is the means by which web marketers send out multiple versions of a digital message—sending customers to two different versions of a web page, for instance, or sending out two different types of emails to potential customers. The marketer can then test which web page or email had the better effect, by measuring how many web clicks they get or, better yet, how much business they generated. It’s not that different from TV commercials that urge buyers to call a 1-800 number and ask for Operator Number 222 or 537. There are usually no such numbered operators—it’s just a way for the advertisers to figure out who saw the ad in question, on which TV show and at what time, for instance. In this way, whether on television or online, marketing messages are constantly being tweaked and modified to maximize their potential audience or success. Political parties are no different from commercial marketers in trying to figure out what messages work best, according to Taylor. Taylor does the same kind of digital-analysis work for the commercial clients of Fleishman–Hillard, a worldwide public relations and marketing firm. Taylor has a foot in both worlds—the political world and the consumer world—just as Martin Goldfarb had in the 1970s and Allan Gregg had in the 1980s.