Glimmers

on foreign election interference, thresholds, and the value of mockery

One of the Parliamentary committees that is exploring foreign election interference allegations, The Procedures and House Affairs Committee (PROC), held a session on April 25 with party election campaign chairs for the last two elections. The focus was on responses to foreign interference during those campaigns.

https://www.ourcommons.ca/DocumentViewer/en/44-1/PROC/meeting-65/notice

During the first hour, the Committee heard from two individuals, Azam Ishmael, who was the Liberal Party campaign chair in 2021, and Jeremy Broadhurst, who served as National Campaign Director for the Liberal party during the 2019 election (and is currently a senior adviser to the PM).

Both individuals were subject to the obligations of the Security of Information Act, having possessed security clearances to receive classified information. They could not discuss the contents of any classified information or briefings they received.

So glimmers. But, still, interesting ones.

Conservative MPs pressed in the first hour on the question of a security briefing provided to senior Liberal party staff regarding the candidacy of Han Dong during the 2019 federal election campaign. Faced with allegations reported by Sam Cooper for Global News that Mr. Dong was a witting accomplice in Chinese state election interference attempts in 2019, Mr. Dong has left the Liberal caucus to defend his name and has sued Mr. Cooper and Global News.

We learned three things about this briefing, all from Jeremy Broadhurst. One was a confirmation that it happened (on September 28, 2019). A second was that the Prime Minister was briefed on the matter the following day—this was fast and partly occasioned by the fact that Mr. Trudeau happened to be in Ottawa on the 29th. A third was that the briefing for the PM was conducted by Mr. Broadhurst alone and it was a purely bilateral meeting. We don’t know the contents of the security briefing provided by CSIS, its gravity, or what was discussed between Mr. Broadhurst and the PM. What we have is Mr. Broadhurst’s characterization of the CSIS material as involving an “information” briefing that contained no actionable intelligence. That is our shoreline of knowledge in the public domain. All else is speculation. That didn't stop Conservative MPs, who wanted to know why Mr. Dong was not removed as a candidate.

In response to a subsequent line of questioning from the Conservatives, this time taken from the contents of a tweet from Global’s Sam Cooper alleging that Mr. Dong has been warned by a Liberal riding organizer that he was a CSIS target, Mr. Broadhurst was emphatic in stating that the story was incorrect and that no riding official would have been privy to the CSIS briefing—thus pushing back against the allegation that the CSIS briefing had been leaked to the candidate.

The second hour featured testimony from the Conservative side. The CPC national campaign manager for the 2021 election, Fred Delorey, and the 2019 national campaign manager, Hamish Marshall, were the witnesses. Again, both had interesting things to say. They came across as less partisan than the Conservative MPs on the committee; more interested in finding solutions that would enhance future election security. One of them, I think it was Mr. Delorey, joked that he would love to win back the 2021 election but that wasn’t going to happen, and that it was important to find solutions for election interference going forward.

Mr. Delorey’s picture of what was known, when, was revealing. He recounted that there had been rumblings during the 2021 election campaign about foreign interference but that it was all anecdotal and never rose to a level of serious or focused concern. It was only after the results were in, that the Conservatives believed they saw an odd pattern—as Mr. Delorety put it—”some results seemed off.” The party then put together information from the grassroots level and created a report, focused on certain riding outcomes in Vancouver and lower mainland B.C, and the greater Toronto area, that had experienced voter swings. The key concern that emerged involved adverse messaging on social media regarding the Conservative party and its candidates, especially messaging on the WeChat platform heavily used by Chinese Canadians.

To his credit, Mr. Delorey stated that it was difficult to know who was behind the messaging and whether the Conservatives lost seats in ridings with a significant Chinese Canada population because of foreign interference efforts (including via social media) or whether it was the hard line approach taken by the then-Conservative leader, Erin O’Toole, towards China and, arguably, a failure to properly delineate criticism of the Chinese regime from that of the Chinese population. (Pierre Poilievre was credited with doing better in that regard). Mr. Delorey also acknowledged that even without the surprise (to the Conservatives) outcomes in these contested ridings the overall election result would not have changed and the Liberals would have formed a government.

Delorey was critical of the way in which information exchanges operated between the political parties and national security agencies during the election period. For both the 2019 and 2021 elections, classified briefings on election interference were provided to security cleared (at secret level) representatives of the political parties. For 2021 the “Critical Election Incident Public Protocol” was amended to include a provision that “political parties will be instructed on how to report any interference that they may experience during the election.” But in Delorey’s view, the information flow was strictly one-way. He contested the validity of the statement in the Rosenberg report that the classified briefings to party officials were regarded as highly useful.

Something is amiss here.

Morris Rosenberg indicates in his report on the operations of the election interference warning mechanism that he interviewed the party representatives involved in the briefings and found that “the party representatives were pleased with the thoroughness of the briefings and the openness of NSA (national security agencies) representatives. They appreciated the opportunity to ask questions.”

The Rosenberg report is available here:

https://www.canada.ca/content/dam/di-id/documents/rpt/CEIPP-rpt-eng.pdf

What did the briefings cover? We have only a general overview conveyed in the Rosenberg report. Topics include best practices that might be employed by political parties for cyber security—a major concern; and threats that could affect the parties, including disinformation threats by foreign actors. Concerns about domestic threats of violence to candidates was given more attention in 2021 in the midst of a rise of extremism around COVID-19 measures. The briefings were clearly broad-ranging.

But Delorey, who was not a direct recipient of the security briefings himself, found the process a “black hole.” Hamish Marshall, reflecting on the earlier 2019 election experience and the inauguration of the party briefing process, believed that the security agencies were uncomfortable in dealing with political parties and had little understanding of the nature of the federal election cycle and process.

Rosenberg himself says that party representatives felt that the briefing process in 2021 (in the midst of a short election period) was treated as “an afterthought.”

The missing piece here is potential testimony from Walied Solomon, who is a lawyer and chair of the law firm Norton, Rose Fulbright, and was the Conservative Party’s security cleared representative during the 2021 election period. Contradictory testimony has been given about Mr. Solomon’s efforts to report election interference allegations, and it would be valuable if the dust could be cleared. The PROC spent some time wrangling over a process to ensure Mr. Solomon could be scheduled to give testimony.

Clearly there is work to be done in ensuring that the national security agencies can properly engage with political parties on election interference threats both during an election writ period and in intervals between elections.

Inevitably, the issue of the threshold for issuing a public warning of election interference was raised, notably by Hamish Marshall. Marshall’s view was that the threshold was both vague and impossibly high. I suspect he is right on both counts. I also suspect that being vague and impossibly high is also necessary. Perhaps even that the threshold is largely irrelevant (but more on these thoughts below).

Just as a reminder, the threshold for the issuance of a public warning of election interference is defined in the Cabinet Directive from 2019, established in advance of the 2019 federal election, that first established the “Critical Election Incident Public Protocol.” The Cabinet Directive is here:

https://www.canada.ca/en/democratic-institutions/services/protecting-democracy/critical-election-incident-public-protocol/cabinet.html

The entire passage describing the threshold is worth reproducing:

6.0 Threshold for informing the public

A public announcement during the caretaker period would only occur if the Panel determines that an incident or an accumulation of incidents has occurred that threatens Canada’s ability to have a free and fair election.

Determining whether the threshold has been met will require considerable judgement. There are different considerations that could be included in making this judgement:

• the degree to which the incident(s) undermine(s) Canadians’ ability to have a free and fair election;

• the potential of the incident(s) to undermine the credibility of the election; and

• the degree of confidence officials have in the intelligence or information.

The Panel brings together unique national security, foreign affairs, democratic governance and legal perspectives, including a clear view of the democratic rights enshrined in the Canadian Charter of Rights and Freedoms.

A disruptive event or incidents of interference may emanate from domestic and/or foreign actors. Attribution of interference attempts may be challenging or not possible within the timelines permitted by events, given that attempts to unduly influence the election may involve misdirection and disinformation. Further, it is possible that foreign actors could be working in collaboration with, or through, domestic actors. Ultimately, it is the impact of the incident on Canada’s ability to have a free and fair election that is at issue in the determination of whether the threshold has been met, and if a public announcement is required. For clarity, Canadians – and democracy – are best served by election campaigns that offer a full range of debate and dissent. The Protocol is not intended to, and will not, be used to respond to that democratic discourse.

This is a smartly crafted statement. Its hard to find fault with it, except to wonder where it leads in practical terns.

The Rosenberg report gives careful attention to the threshold issue and makes two important points. One is that the idea of the panel of five issuing a public warning, presumably to come from the Clerk of the Privy Council, was always seen as a last resort measure. The related issue is that too much attention has been paid to the threshold issue at the expense of a broader understanding of the what Morris Rosenberg calls a “robust information sharing environment” for the defence of elections, including interaction with civil society and social media platforms, with political parties and with the Chief Electoral Officer. Intelligence gathering efforts by national security agencies may also lead to threat reduction measures, which are in the mandates of both CSIS and the RCMP.

Rosenberg defends the need for the warning protocol and a threshold, while also suggesting that defining the threshold needs further study.

I would go in a different direction and suggest that the idea of a public warning by the Clerk during an election is largely a fantasy, rooted in a historic concern drawn from the experience of the US election in 2016 involving wholesale Russian state/proxy interference efforts. It was understandable that the Canadian government would want to draw a lesson from the American experience of seven years ago and find a way to defend Canadian elections through a public alert system.

But the real threats to election security have shifted, and Canada is not the target that the United States presents to foreign actors. So the 2016 analogy may no longer be apt, and the threshold issue largely irrelevant.

What are the real threats now? They operate in the “grey zone” of disinformation operations on social media platforms, where attribution is difficult and distinguishing between foreign and domestic hands may be well nigh impossible. The best defence is not a warning by a panel of senior civil servants, based on secret intelligence which cannot be divulged, but public education, buttressed by trust in the Canadian electorate’s ability to make its own political choices.

Then there is the other “grey zone” often channeled thorough social media—threats of violence, harassment, and intimidation of candidates, political parties and voters. These are domestic threats, not foreign interference threats, and have to be dealt with as domestic political problems. We will have to look to policing and law enforcement (including around hate speech) as tools, and anticipate heightened physical and cyber security during campaigns.

Political parties will also have an important role to play during election campaigns in making decisions on moderating polarizing platforms and rhetoric. Keep those dog whistles in your pocket. The public needs to turn its back on expressions of political extremes.

While we contemplate real tools, it is also worth keeping in mind the French government’s approach during elections in 2017. If you see a disinformation campaign worthy of being mocked (and most have an element of sheer stupidity)--well mock away.

Free advice to all political parties.