robocalling

I’ve been working on some research into the voter suppression phone call scandal in which voters were misdirected to non-existent polling stations. I hope to have some of the detailed results published on the weekend. I’ve looked at the Elections Canada financial records of 96 Conservative candidates (including most of those listed by the Sixth Estate) with a particular eye on the “research and survey” lines. I also looked at the financial reports for all 55 Third Party campaigns.

While we’ve heard a lot about RackNine, Responsive Marketing Group (RMG) and Front Porch Strategies, I’ve identified up to 40 different service providers who were involved in these 96 Con campaigns. They are based in both Canada and the U.S.. As well as these private companies, there are a number of payments made to the Conservative Fund of Canada and/or the Conservative Party of Canada for research/surveys as well as the local riding associations themselves who will report on their finances later in the year (follow the money).

As reported on several occasions, the Catch 22 Campaign used auto-dialing to contact voters in 16 ridings. I thought I would outline the process to give readers a better idea of the mechanics before publishing the research findings (which still need some analysis and follow-up research). 

Catch 22’s experience

I’m by no means an expert on auto-dialing but I was directly involved — hands-on — in the last week of the election. We made more than one million calls into 16 ridings. Our calls combined a voter intention “robo-poll” with a strategic voting message to our target voters.

Here’s an overview of what we did:

1) Hired an auto-dial service provider. As our public financial records show, we paid a flat rate per riding which included the rental of the riding phone numbers. Each list had tens of thousands of phone numbers.

2) Got an account with a log-in name and password. This gave us access to a web interface that let us “do it yourself” and save money. We also had a training session.

3) Created an electronic “campaign” for each riding. At this stage we designed the call (what happens when they push different buttons, the dates and times for the calls, etc.)

4) Wrote and recorded scripts for each campaign. Once we did that, we had to listen to and approve the recordings before they could be finalized and used. This was a built-in “final” safety check to make sure we didn’t send out the wrong message by mistake.

5) Turned it on.

6) Once the calls began, we could see their progress online in real time — i.e. see a summary of how many calls were made, how many minutes we used and the results of the “voter intention” robo-poll.

7) We were also able to produce a detailed report with the robo-poll results in the form of a spreadsheet file that could be turned into lists of Con, NDP, Green and Lib voters’ phone numbers.

Voila. That’s one way that a local (or any other) campaign can produce its own “friends” and “enemies” lists. Catch 22 did not use these lists as we didn’t have a “get out the vote” component to our campaign.

Of course, with similar complaints coming from many ridings, there is growing speculation that there was a conspiracy (and cover-up) and that the lists may have been downloaded from the Conservative Party’s Constituency Information Management System (CIMS). However, a process similar to the one that I outlined above could have been used to feed voter intention data to CIMS (or used locally). In this undated CIMS document, it’s pretty clear that local ridings upload all their voter information to CIMS (see Slide 4).

Cross-posted from the Catch 22 blog.

Gary Shaul

Gary Shaul is a life-long Torontonian and retired Ontario civil servant. He's been involved with a number of issues over the past 45 years including trade unionism, proportional representation, Indigenous...