The widespread accounts of electoral fraud and voter suppression in the 2011 federal election are now a major volume in Canadian electoral history (Have some Pierre Poutine, anyone?). The chapters of this tragic farce are being written by various authors, one of whom is the Council of Canadians (chaired nationally by Maude Barlow), which has given its support to a court case challenging the results in seven ridings (Winnipeg South Centre, Elmwood-Transcona, Saskatoon-Rosetown-Biggar, Nipissing-Timiskaming, Don Valley East, Vancouver Island North, and Yukon) where there were extensive reports of electoral fraud and voter suppression, and in which individual voters filed applications to the Federal Court of Canada to have the election results overturned as a result of these irregularities, and new elections ordered.
This process has been underway since late March 2012 and there have been repeated attempts by the lawyers representing the seven Conservative candidates in these ridings to quash the proceedings -- none of them successful. The case is now slated to be heard by the court commencing December 10, 2012.
Nonetheless, attempts to derail the proceedings continue. On September 18, 2012 lawyers acting for the Conservative candidates filed a motion to have the applicants in the case post a $260,409 deposit to cover the legal costs of the Conservatives in order for the case to proceed. This is despite the fact that the Elections Canada Act specifies a security deposit of $1,000 per applicant (the lawyers are attempting to exploit a loophole that allows the court to increase the deposit at its discretion). It seems a transparent attempt to drive up costs so as to force the applicants to withdraw, a familiar bare-knuckles legal ploy (the court decision on this application is expected shortly).
In another serious attempt to derail the proceedings, lawyers acting for the Conservative candidates filed a motion on September 11, 2012 to exclude the study that Frank Graves of Ekos Research Associates conducted on behalf of the Council of Canadians, A Study of the Incidence and Effects of Misleading Calls in the 41st National Election. The Ekos study, part of an affidavit filed with the court, is a major piece of research that underpins the points under contention in the legal case. The motion accuses Graves of "clear bias" against the Conservatives and lawyer Arthur Hamilton claims that Graves has "repeatedly and publicly aligned himself with political interests, parties and actors that oppose" the Conservatives.
Before examining what the motion to exclude Graves' evidence argues, let's look at what Ekos has done and why it is important. Ekos set out to answer three questions:
- Were voter suppression techniques used in the seven ridings that are at the heart of the legal case?
- If such techniques were used, did they deliberately target voters who were supporters of a particular party or parties?
- If such techniques were used, were they effective in discouraging voters from casting their ballots?
These "subject ridings" (individually and collectively as a group) were compared with a "comparison group" of people selected from 106 other ridings across the country.
1. Were voter suppression techniques being employed?
The figure below summarizes the results (3,297 respondents in the subject ridings; 1,500 in the comparison group) of six polling questions. The overall margin of error is +/- 1.7 percentage points 19 times out of 20. The six questions asked were:
- Voter ID call? -- Did respondents receive a call asking how they intended to vote? Such "profiling" is a widespread (and legitimate) practice in elections.
- Late polling contact? -- Did respondents receive a call late in the campaign telling them the location of their polls? Such "get out the vote" (GOTV) tactics are also a widespread (and legitimate) practice in elections.
- Polling station change? -- Did respondents receive a call telling them that the location of their polling station had changed? Polling station changes are rare and Elections Canada does not make telephone calls to convey such information. Moreover, in the subject ridings there was only one actual polling station change.
- Polling station correct? -- Was the information on polling station changes that respondents received correct? [*Note: in this case what is shown in the figure is the proportion of respondents who answered "No."]
- Harassing calls? -- Did respondents receive a harassing telephone call?
- From "Elections Canada"? -- Did the caller identify themselves as calling on behalf of Elections Canada?
The response to the first (legitimate) question indicates that the subject ridings experienced significantly greater "profiling" to ascertain voting intentions than was the case in Canadian ridings as a whole. The results to the third and fourth questions show clear differences (which are statistically significant) in the subject ridings in polling station misdirection efforts in comparison to Canadian averages. Bear in mind that with only one actual change in polling station locations in the seven subject ridings, information about purported "changes" would almost invariably have been false.
The number of "harassing" phone calls was also higher in the subject ridings (and the difference was statistically significant) but as Ekos points out this is an imperfect measure since there is a degree of subjectivity in terms of judging what constitutes a "harassing" phone call. There was no difference between subject and comparison ridings in terms of the proportion of callers who claimed to represent Elections Canada.
Together, these results indicate that voter suppression techniques were being employed in the seven subject ridings, and that their incidence was greater there than the Canadian average.
2. Was voter suppression targeting particular parties?
Now this is where things start to get really interesting. In the case of purported polling station changes, the figure below illustrates the differences reported by supporters of different political parties. The incidence of misdirection reported by Conservative voters in the subject ridings and the comparison group is virtually identical, whereas there are wide (and statistically significant) discrepancies between these values reported by NDP, Liberal, and Green party supporters.
Similarly, in terms of incorrect polling location information shown in the figure below, the difference in incidence of misdirection reported by Conservative voters in the subject ridings and the comparison group is rather slight, whereas it is much greater in the case of that reported by NDP, Liberal, and Green party supporters.
Together these are a clear indication that voter suppression in the subject ridings targeted non-Conservative voters. It would appear that the very extensive voter profiling effort (used to determine who supported which political parties) conducted in the subject ridings at the outset may then have been selectively used in voter suppression tactics.
3. Were voter suppression tactics effective?
Having established that voter suppression was taking place, determining how effective it was is a more complicated undertaking. The figure below shows the percentage of voters who, if they received information on polling station changes, were subsequently motivated not to vote as a result.
When this information is subsequently factored into the breakdown of supporters of different political parties (remembering that voter suppression was differentially targeted at supporters of Conservative versus non-Conservative parties) and the proportions of the populace in those ridings that were supporters of those parties (using actual 2008 data as a proxy) Ekos concluded that the suppression effect was in the range of 1.2-1.8 per cent [see the report itself for further details of how these calculations were done.] Applying the margin of error to these results gives a span of 0.8-2.2 per cent. In other words, had voter suppression not occurred in these seven subject ridings, the Conservative advantage in them would have been reduced by this amount on average. Is this significant in the context of the 2011 election results?
The figure below (generated from 2011 election data, not from the Ekos study) shows the percentage of votes received by the Conservative candidate and the closest non-conservative candidate in each of the seven subject ridings [red indicates a Liberal opponent; orange an NDP opponent). Except for Vancouver Island North (where the percentage gap was 3.06 per cent) in all other six subject ridings the percentage gap is within the 0.8-2.2 per cent voter suppression range as determined by Ekos.
Thus, in my view, there is the plausible contention that had voter suppression not taken place, in at least six of the seven ridings that are the subject of the current court case, a different political party could have emerged as victorious in the 2011 federal election. Again, in my view, this would appear to constitute significant evidence that the election results in those six ridings could be considered as suspect and that in order for the democratic will of the populace in those ridings to be known, the results of the 2011 election should be discarded and new elections ordered in order to determine what the electoral will of the citizens in those ridings actually is.
If you don't like the message, attack the messenger
Another tried and true legal ploy is to avoid discussing the substance of an issue by attempting to discount the credentials of the person conveying the information.
The motion filed by lawyers Arthur Hamilton and Theodore Frankel to "exclude the evidence of Frank Graves" is filled with unsubstantiated assertions, confusions, non sequiturs, irrelevant statements, unsupported assumptions, hyperbole, and a seeming inability to differentiate between casually humorous remarks offered on a Twitter feed and information within substantive, empirically-based studies. It is peppered with statements such as, "Even more troubling, however, is Mr. Graves’ tendency to explain away his incendiary and partisan statements with an appeal to supposed statistical data."
As one illustration, consider an example offered in the motion which alleges that Graves' contention that the subject ridings were "exposed to a program of voter suppression which was targeted and effective" (quotation from Graves) -- is "replete with unstated assumptions and impermissible factual conclusions" (quotation from the lawyer's motion). Thus the claim that:
(i) there must have been a 'program' behind the patterns;
(ii) the program was put in place to suppress voters from exercising their franchise;
(iii) the program was targeted towards a specific sort of voter by a directing mind; and
(iv) the program somehow affected the results of the Election;
"Are all decidedly outside the realm of a statistician's special expertise in reading the results of a telephone survey (quotation from the lawyer's motion)." None of this is in the least meaningful or cogent. The Ekos study clearly demonstrates that non-random voter suppression took place, and Graves has 32 years experience in impact evaluation research and related topics under his belt. In like mind, the logic of the entire motion reads as if viewed through some sort of warped looking-glass at a Coney Island funhouse: it might make sense but only if you'd received a "boot upside the head."
In fact, the heart of the motion is the claim that "Mr. Graves has repeatedly and publicly aligned himself with political interests, parties and actors that oppose the Respondent parliamentarians, all of whom are members of the Conservative Party of Canada."
The essence of this legal stratagem is to de facto exclude anyone who may have political convictions that differ from the ones being represented. Why? Because such a person can't possibly be a credible witness. Because the assumption is that methods, facts, data, evidence, reason, and logic are all irrelevant. That political beliefs irredeemably taint everything, and so to actually examine evidence would be prejudicial. As the thousands of scientists who protested on the steps of Parliament Hill maintained, in Canada we have arrived at the "death of evidence." Only political convictions matter and the only permissible evidence is that of your political pals (everyone else is, by definition, excluded).
In an article by Glen McGregor for the Ottawa Citizen, Graves said that "the motion is full of falsehoods and he considers it defamatory. While he cannot sue for libel because the statements are protected by the privilege that attaches to court proceedings, Graves said Monday he is considering a complaint to the self-governing body for Ontario lawyers, the Law Society of Upper Canada."
"It's not very gratifying but it's all I've got available to me," he was quoted as saying. "They've really lost their judgment here. The stuff they say is just false."
What are we to make of this? Two bleak points emerge:
It is a dismal testament to the state of Canadian democracy that voter suppression and electoral fraud appear to be pandemic across the country. Even amongst the "comparison group" of voters from other ridings, 11.6 per cent of Canadians report receiving calls that incorrectly identified their polling station in the last federal election. Could all of these have been made by bone-headed GOTV volunteers legitimately working for political parties? I leave it to readers to judge whether such extensive voter misdirection is credibly ascribed to human error: it beggars the imagination.
And in the seven subject ridings that are at the heart of this legal case, the incidence of apparent voter suppression and electoral fraud is even greater: 17.6 per cent of voters report receiving such calls. The good news is the large majority of Canadians are savvy and don't fall for such blatant attempts at suppression and misdirection. The bad news is that in some cases, even the small proportion of voters who fall victim to such tactics are enough to tip the balance between political parties under Canada's antiquated first-past-the-post (FPTP) electoral system. Close vote splitting amongst multiple parties is now a feature of a majority of Canadian ridings. Thus, such electoral manipulation, far from constituting fringe political shenanigans, can affect the outcome, not only in individual ridings, but also in the general election results as a whole. FPTP already produces unpredictable and unrepresentative outcomes, even when conducted fairly. When you add electoral fraud to the mix, the results start to resemble a rigged horse race. If one more argument for the necessity of electoral reform in Canada were needed, the current bleak state of Canadian democratic affairs certainly provides it.
The second dismal point is provided by the current federal government, which appears prepared to stonewall at all costs rather than address this situation. While it could be argued that the current progress of this legal case only represents the adversarial, zero-sum, bare-knuckles, no-holes-barred legal system in action, in which attorneys representing both sides employ any every legal stratagem to best represent their clients -- and this is certainly true -- nonetheless, the decision to deploy such DEFCON 1 legal maneuvering is a conscious and deliberate one. It doesn't need to happen this way.
The legitimacy of the Canadian government is based on fair elections, and the perception of fair elections. If the results have been rigged, the legitimacy of the government itself is subject to question. Even if there is the widespread perception on the part of the Canadian electorate that the deck has been stacked, then the legitimacy of the government is tarnished accordingly. One might hope that every Canadian government would be profoundly concerned by such a state of affairs, would set aside any partisan impulse, and do its utmost to address such actual or perceived democratic distortions and clear the political air. That it would feel an obligation to its citizens to ensure that democratic legitimacy was strictly observed and to immediately address any shortcomings. This would be the hallmark of a responsible and accountable government. Sadly the hyper-partisan, win-at-all-cost attitude of the Harper Conservatives appears to indicate that none of the foregoing is true.