The System Is Working

Environmentalist groups have a point when they say they are being muzzled by Elections Canada; trouble is, that’s exactly how the law is meant to work

As the media reported earlier this week, environmentalist groups are angry at Elections Canada, which has warned them that spending money to raise awareness of climate change in the run-up to the coming federal election would subject them to the rules on “third party” participation in election campaigns. Many are feeling that they will be required to keep quiet during the campaign, which rather defeats the purpose of being advocacy groups. Even the BBC has a story on this.

For its part, Elections Canada has issued a response claiming that the Canada Elections Act doesn’t prevent advocacy groups from advocating, so long as they register if they spend $500 or more and comply with the spending cap. Elections Canada adds that the registration requirement “leads to increased transparency” and has been in place “for nearly 20 years”. Helpfully, I suppose, the statement concludes with an acknowledgement that the rules “can be complex”, and Elections Canada is happy to answer questions about them.

The rules are indeed somewhat complicated, as I explain below. But the bottom line is simple enough. Despite the officials’ protestations, NGOs ― be they environmentalist or other ― have a point when they say that they are being muzzled. To some extent, that’s what the Canada Elections Act is designed to do; to an even greater extent this might be an unintended consequence of the Act’s pursuit of transparency, but an entirely predictable one. The issues are well known; I, for one, raised them in my statement to the House of Commons Select Committee that considered the latest round of amendments to the Canada Elections Act. The only surprising thing is the degree to which people still end up being surprised when problems of sort arise.


The Canada Elections Act‘s regulation of political spending is predicated on the idea that attention during election campaigns should be focused on politicians ― individual candidates and political parties, especially parties. Parties, if they run candidates in all ridings, are able to spend tens of millions of dollars on advertising ― which they are entitled to buy at favourable rates, in addition to an allowance of free airtime. Non-politicians ― that is, individuals, labour and student unions, corporations, and NGOs ― are known as “third parties” in the election law jargon and, as I explained here, their participation in electoral debates is viewed as anomalous, indeed suspicious, and is strictly limited.

One set of limits concerns the amounts of money third parties are allowed to spend, which are only a small fraction of the spending allowed political parties. The Supreme Court has upheld the limitation of third party spending during election campaigns, notably in Harper v Canada (Attorney General), 2004 SCC 33, [2004] 1 SCR 827, although there is good reason to be critical of that decision, which I have even rated as one of the worst in the last fifty years. (As I noted here, the High Court of Australia was also quite skeptical of Harper in a recent decision.) Last year, Parliament enacted further limits that apply even before the formal campaign begins, and their constitutionality has not yet been tested; Harper, in my view, does not dispose of the question.

In addition to spending limits, “third parties” are also subject to onerous registration and reporting requirements. Some of these are the cause of the latest dust-up. Specifically, Division 1 of Part 17 of the Canada Elections Act imposes such requirements on “third parties” that incur more than $500 of expenses on, notably “partisan activities” and “partisan advertising” during the “pre-election period”, which begins on June 30 of the year for which a fixed-date election is scheduled and ends with the start of the election campaign. During the election campaign itself, governed by Division 2 of Part 17, “election advertising”, as well as “partisan activities” count for the spending thresholds that can trigger registration and reporting requirements.

The definitions of “partisan” and “election advertising”, found in section 2(1) of the Canada Elections Act, are very broad. The former term “means the transmission to the public by any means during a pre-election period of an advertising message that promotes or opposes” a party or a candidate, further defined in section 2(7) as “naming”, “identifying” (“including by … logo” or picture, as the case may be, and “providing a link to an Internet page that” names or identifies the party or candidate. “Election advertising” includes the same things as “partisan advertising”, but also “taking a position on an issue with which a … party or candidate is associated”, even without naming that party or candidate. Since issues with which no candidate or party “is associated”, come election time, are about as common as colour pictures of a Maple Leafs Stanley Cup parade, the definition of “election advertising” encompasses pretty much any advertising that has anything to say on matters of government or policy.

Now, some means of communicating with the public are exempted from these definitions. In particular, the exemptions cover anything that the media will print or broadcast without charge to the speaker ― things like quotes in news items, interviews, and op-eds. Also exempt are organizations’ communications with their members, shareholders, or employees, as well as “the transmission by an individual, on a non-commercial basis on the Internet, of his or her personal political views”. Note, though, that on its face the latter exemption doesn’t cover ― indeed, it rather pointedly excludes ― a group’s or an organization’s online communications, even if not paid for (for example, tweeting under the organization’s handle). And of course, any communication that the media are not interested in carrying free of charge will count as an advertising. In effect, for groups and organizations, the media are the gatekeepers of their ability to communicate with the public without having to register as “third parties”.

So what’s the big deal about registration? Well, although you won’t know it from the Elections Canada statement linked to above, registration doesn’t just mean filling out a form. There are a number of other requirements. To begin with, unions and corporations cannot register before their board has adopted a resolution authorizing them to incur expenses on “partisan” or “election advertising” (sections 349.6(5) and 353(5) of the Canada Elections Act). All “third parties” are also required to have a “financial agent” who will be responsible for collecting money to be spend on “partisan” or “election advertising” and for spending it (sections 349.7 and 354). These transactions must be done through a separate bank account (section 358.1) After the election is over, a detailed report on the money collected, advertising taken out, and costs incurred must be filed (section 359). And this is not all. Those “third parties” that spend more than 10,000$ are also required to file interim reports during the course of the election campaign and, most significantly, to appoint auditors (section 355) and file the auditor’s report on their spending (section 360).

Needless to say, this is all quite costly, at least in time, but also ― especially for those third parties that spend more 10,000$ ― in money. Big trade unions, whose budgets are extracted from workers who don’t get a say on whether to contribute or on how the money is spend, may not be especially troubled by these costs. But for NGOs, whose income comes from voluntary (albeit taxpayer-subsidized) donations, and which need to be much more careful about how they spend it, compliance with the Canada Elections Act may be too expensive. From their perspective, the sensible if unfortunate thing to do may well be to keep quiet for the duration of the election campaign, or even starting with the beginning of the pre-campaign period.

This means that for a period of almost four months preceding the election ― the period when the most people pay attention, even if it’s still sporadic and fragmentary attention, to political and policy issues ― civil society organizations may indeed be prevented from expressing their opinion about politicians, except to the extent that the media will let them. Again, the bigger and better-known you are, the less of a problem this may be for you. Smaller groups, whose views are (naturally and fairly) of less interest to the media, will find it more difficult to get across to the voters. The more unusual voices, in other words, are the ones who are the most at risk of being silenced ― in effect if not, perhaps, in intent ― by the Canada Elections Act.

And of course even for larger groups, having to pass through the media means sound-bite-sized interventions have a much better chance of getting across to the voters than anything more serious. Say that a politician or party is anti-environment, or pro-worker, or something equally inane, and the media may well pick it up. But they’re not going to run a detailed report card assessing the competing parties’ platforms on some issue ― but publishing it on an NGO’s website, let alone running it as an advertisement would mean having to comply with burdensome registration and reporting requirements under the Canada Elections Act.


No wonder, then, that environmentalists are feeling muzzled and frustrated. And of course groups pursuing other agendas may be feeling that way too ― or may come to feel that way when the occasion arises. They have more than a little justification. And they shouldn’t be the only ones feeling wronged. The voters should be too. You may not miss the presence of a particular set of activists in the election campaign, but the rules that silence them silence the activists on your side too. You may not be all that interested activists generally have to say, but you should be interested in politicians’ feet being held to the fire.

The ostensible rationale for registration and reporting requirements is that they serve to promote transparency, in addition to assisting in the enforcement of spending limits applicable to “third parties”. It is on that basis that the Supreme Court upheld those requirements that apply in the course of the election campaign ― although not those applicable in the pre-campaign period, which weren’t yet in the Canada Elections Act ― in Harper. Yet one needs to weigh the value of transparency against the costs that its pursuit imposes on those subject to the Canada Elections Act ― and, as I have just explained, on the voters who are being deprived of important contributions to the electoral debate.

The Harper majority’s analysis on this point was quite perfunctory. There is no real discussion of compliance costs and their deterrent effects. Instead, the majority is content to baldly assert that “[t]he appointment of a financial agent or auditor is not overly onerous. Rather, it arguably facilitates the reporting requirements.” [145] Even worse, the majority did not at all consider what I think is the crucial issue: the thresholds at which the registration and reporting requirements kick in. All it said was that the requirements “vary depending on the amount spent on election advertising”. [145] Yet one can accept the principle of imposing such requirements on heavy spenders while also acknowledging that the existing rules go much too far.

In New Zealand, “third parties” are not required to register until they spend NZ$13,200 (ca. C$11,000); more detailed reporting requirements only apply once a “third party” spends NZ$100,000. (Even then, third parties aren’t peremptorily required to provide an auditor’s report, although they may be asked to do so.) These strike me as rather more reasonable figures than those in the Canada Elections Act, though even they should probably be multiplied several-fold to account for the fact that New Zealand’s population is only a small fraction of Canada’s.

It is difficult to believe that a “third party” spending a few thousand, or even tens of thousand of dollars is going to have any substantial impact on an election by itself. At most, it may be successful enough in getting other people ― voters, media, or politicians ― to discuss the issues it is raising. It is this discussion, rather than anything published on an NGO’s website or even a Facebook ad, that might, conceivably albeit theoretically, matter. In the abstract, this discussion might be enriched by more disclosure. In practice, the very real costs of the disclosure requirements end up preventing the conversations from happening at all. I fail to see how the voters benefit from this.


As Elections Canada points out in its response to the environmentalist groups, the “advertising during the election period has been subject to the Canada Elections Act for nearly 20 years”. This is true. (As noted above, rules on advertising in the pre-election period are new.) For about half of this time, it has been known, at least to those who study these things, that the rules tend to hobble not business interests, but labour unions and civil society groups. Colin Feasby wrote about this in 2010; I did (in the context of Québec elections, which are subject to similar but even more draconian rules) in 2012; also in 2012 Tom Flanagan came out in support of rules like those in the Canada Elections Act, whose enactment he had opposed, with the declared intention to muzzle unions; I updated Dr. Feasby’s findings in an article published in 2015. And in my statement to the House of Commons Standing Committee on Procedure and House Affairs when it was studying amendments to the Canada Elections Act last year (which, among other things, introduced restrictions on “third parties” in the pre-campaign period) I specifically mentioned both the registration and reporting requirements’ tendency to muzzle civil society, and the needless low threshold at which these requirements apply. Needless to say, that had no effect on the resulting legislation.

Yet at every election the impact of restrictions on “third parties” seems to surprise. It happened in Québec in 2014, when the Chief Electoral Officer tried censoring a short documentary a group of citizens had produced to oppose the election of the Parti québécois and the enactment of its “values charter”. Eventually, the Chief Electoral Officer changed his mind; but he was wrong to do so. It happened again in Québec in 2018, now with environmentalist groups being targeted. And now it’s happening at the federal level. The system, one might say, is working. It was designed to shut down political debate not dominated by politicians or the media. That’s what it’s doing.

It will be obvious that I don’t think it’s a good system. Like the National Post’s Chris Selley, I think the rules need to be changed. Whether any restrictions on political spending are justified is debatable but, as noted above, one can accept the premises of Canada Elections Act and still support relaxing its requirements a great deal. Ideally, the next Parliament will take up the issue. But there is also room for litigation. Certainly rules on pre-campaign spending, whose constitutionality has not yet been tested all the way to the Supreme Court can be challenged. But perhaps even the registration and reporting rules upheld in Harper could be attacked, provided that the courts are forced to consider solid evidence of their pernicious effects.

Australia 1:0 Canada

Canadians have much to learn from the Australian High Court’s take on election spending limits for “third parties”

The High Court of Australia has just delivered Unions NSW v New South Wales [2019] HCA 1, a decision that should be of interest to readers who are concerned with freedom of expression in the electoral context ― a topical issue in Canada, given the recent imposition of further restrictions in this area by the recently enacted Bill C-76. The decision resulted from a challenge by a number of labour unions to New South Wales legislation that reduced the maximum amount a “third party” ― that is anyone not a candidate at an election or a political party ― is allowed to spend on campaigning in a nearly-six-month period preceding an election, from 1,050,000AUD (jut under a million Canadian) to 500,000AUD. The High Court unanimously held that the legislation was contrary to the implied freedom of political communication, which it had previously read into the Australian constitution‘s provisions requiring “representative” government.

The plurality judgment, by Chief Justice Kiefel and Justices Bell and Keane, finds that the third party spending limits are unconstitutional. That they restrict the ability to communicate is not in dispute. And while the plurality is prepared to assume that these limits are imposed for the legitimate purposes of levelling the campaigning playing field and preventing the wealthy from “drowning out” the voices of the less fortunate, they are not justified. Experts consulted prior to the enactment of the legislation provided no particular justification for recommending that the then-existing spending limits be reduced. A Parliamentary committee, however, recommend that the legislature look into the actual spending needs of third parties, and this was not done either. As a result, there is no reason for saying that the reduced limits are “reasonably necessary”.

Justice Gageler agrees with the plurality’s disposition of the case. He is persuaded of the legitimacy of the state’s pursuit “of substantive fairness in a manner compatible with maintenance of the constitutionally prescribed system of representative and responsible government”. [91] This might, in principle, justify much lower spending limits for third parties, which campaign on single issues, than for parties that must address a broad range of issues in their quest to form a government. However, “[i]t is not self-evident, and it has not been shown, that the cap set in the amount of $500,000 leaves a third-party campaigner with a reasonable opportunity to present its case”. [101] Absent such a showing, the restriction on the freedom of communication is not justified.

Justice Nettle’s conclusion is similar. He accepts the legitimacy of the objective of creating a level electoral playing field ― one on which political parties will be primary players ― and agrees that a legislatures may from time to time review the measures it takes to ensure fairness, including by lowering spending caps previously enacted. However, there must be a justification for whatever measures it takes from time to time. Such a justification is missing in this case. Although it was recommended that more evidence on the needs of third parties be collected, “for reasons which do not appear, that recommendation went unheeded. It is as if Parliament simply went ahead … without pausing to consider whether a cut of as much as 50 per cent was required”. [117]

Justice Gordon, like the plurality, assumes that restrictions on third party spending pursue a legitimate purpose, which she characterizes as the privileging of political parties and candidates. However, in the absence of evidence about the actual need for restrictions set at their current level, “the Court … cannot be satisfied that the level of the expenditure cap is reasonably appropriate and adapted to achieve the asserted constitutionally permissible end”. [150] It was for the State to show that the restriction it seeks to impose was justified, and it has not done so.

For his part, Justice Edelman considers that the reduction in the spending limits imposed on third parties, even as the limits imposed on political parties rose, cannot be explained by the purposes of maintaining a fair and corruption-free electoral system. Rather, it must have had an “additional purpose”, which “was to ensure that the voice of third-party campaigners was quieter than that of political parties and candidates”. [159] In other words, the reduction’s aim was “to burden the freedom of political communication of third-party campaigners”. [160] Justice Edelman considers that, although laws that rely on the relative silencing of some views in order to ensure that all can be heard are legitimate, to aim only at silencing some voices “is incompatible with the maintenance of the constitutionally prescribed system of representative and responsible government”, and legislation so motivated is “invalid”. [160]


Needless to say, I am not qualified to comment on whether the High Court is correct as a matter of Australian law. What I can do is compare its decision with that of the Supreme Court of Canada in Harper v Canada (Attorney-General), 2004 SCC 33, [2004] 1 SCR 827, which addressed much the same issues. (Readers will recall that I am not a fan of Harper, to put it mildly, and included it in my list of the Supreme Court’s five worst decisions of the last half-century in this blog’s recent Twelve Days of Christmas symposium.)

This is most obviously so on the issue of deference to the legislature on the issue of the appropriateness of a limitation of the freedom of (political) expression, and the evidence required for the government to make this case. The Harper majority insisted that courts should approach legislative choices with deference. In its view, “[t]he legislature is not required to provide scientific proof based on concrete evidence of the problem it seeks to address in every case”, and that “a reasoned apprehension of … harm” [Harper, 77] is sufficient to restrict fundamental freedoms protected by the Canadian constitution.

This approach is explicitly rejected in Unions NSW. While the Australian judges avoid directly criticizing the Harper majority, both the plurality opinion and Justice Nettle explicitly side with the “strong dissent” of Chief Justice McLachlin and Justice Major (joined by Justice Binnie). The plurality takes a dim view of the submission “that Parliament does not need to provide evidence for the legislation it enacts [and] is entitled to make the choice as to what level of restriction is necessary to meet future problems”. [44] When legislative choice are made in a way that burden the freedom of political communication, they must be justified. Similarly, Justice Gageler speaks of the need for a “compelling justification”, and insists that “[i]f a court cannot be satisfied of a fact the existence of which is necessary in law to provide a constitutional basis for impugned legislation, … the court has no option but to pronounce the legislation invalid.” [95] Justice Gordon insists that “the Court must … be astute not to accept at face value the assertion that freedom of communication will, unless curtailed by a reduction in the cap to $500,000, bring about corruption and distortion of the political process”. [148]

Another point of contrast between Harper and Unions NSW is the treatment of the so-called “egalitarian model of elections” designed in part to favour the interests of political parties and candidates over those of the civil society groups, disparagingly consigned to the status of “third parties”. According to Harper, election campaigns must focus attention on parties and candidates, including by ensuring that any other participants in the public debate, except the media, will behave unobtrusively. By contrast, the plurality opinion in Unions NSW explicitly rejects the submission that candidates and parties deserve preferential treatment, advanced in part on the basis that elections are “not a choice between ideas, policies, views or beliefs except insofar as such choice may be reflected in the electoral choice between candidates”. [39] Rather, the plurality says, “ss 7 and 24 of the Constitution guarantee the political sovereignty of the people of the Commonwealth by ensuring that their choice of elected representatives is a real choice, that is, a choice that is free and well-informed” [40] ― including by third parties. Justice Gageler, of course, takes the contrary view on this point. Justice Edelman’s position is more complex. He explicitly endorses “a Rawlsian, egalitarian model” [178] in which spending limits prevent some speakers from “drowning out” others. However, he also considers that it is not legitimate to target particular speakers for silencing apart from such an anti-drowning out purpose.

A last difference between Harper and Unions NSW worth highlighting is recognition by Justice Gageler of “the propensity of an elected majority to undervalue, and, at worst, to seek to protect itself against adverse electoral consequences resulting from, political communication by a dissenting minority”. [66] Justice Gageler refers to prior cases where the risk of a government legislating to limit political competition the better to maintain itself in office was explicitly adverted to. Such legislation, he notes, is incompatible with presuppositions of the Australian constitutional order. Although he finds that, in this case, “[t]here is no suggestion of abuse of incumbency” [85] by one party against others, this clear-eyed position is in contrast to that of the Harper majorityr, which ignored the possibility that incumbent governments favour legislation that excludes “third parties” from electoral campaigns in order to avoid unpleasant criticism and so reduce the odds of losing power.


There are more interesting things in the Unions NSW decision than I have room to discuss in this post. For example, Justice Gageler’s comments about the role courts in finding facts that are relevant to deciding whether a statute is constitutional are in contrast to the position of the Supreme Court of Canada in cases such as Canada (Attorney General) v Bedford, 2013 SCC 72, [2013] 3 SCR 1101, and should be very nutritious food for thought for those who are skeptical of the Bedford requirement of deference to trial judges. Justice Eledman’s comments on identifying statutory purpose (and in particular the role of general statements of purpose in the legislation) are also very interesting.

Overall, based on this one decision, I think that Canadians have a great deal to learn from Australians. Admittedly, the length of the High Court’s decisions is a deterrent ― Unions NSW is about 85 pages long, and I take it that it’s pretty short by Australian standards. That’s the cost of so many judges delivering full individual reasons. But the upside is that interesting ideas don’t get swept under the carpet in the process of getting to a set of reasons many judges can sign onto. I’m not saying the Supreme Court of Canada should go back to having each judge deliver his or her own reasons (though I wonder sometimes) but, at any rate, reading the Australian decisions may well be worth our while. In particular, the willingness of the Australian judges to keep a legislature accountable for imposing limits on the freedom of political expression without justification is a welcome reminder that their Canadian counterparts can do much more to protect individual rights in the electoral realm, and elsewhere.

Civics, Feelings, and Politics

Expatriates’ alleged lack of connection to particular ridings is not a good reason to disenfranchise them

When it held, in Frank v Canada (Attorney-General), 2019 SCC 1 (summarized here), that denying the franchise to Canadians who have been resident abroad for more than five years is unconstitutional, did the Supreme Court go with “feelings over civics”? Did it decide the case in a way that ignores the fact that Canadians vote not for national parties but for candidates in local constituencies, to which expatriates are not meaningfully connected, even if they maintain, as the Court’s majority said, a “profound attachment” to Canada? Over at Routine Proceedings, Dale Smith argues that that’s precisely what the Court did. I disagree.

As Mr. Smith sees it, “five of seven justices of the Supreme Court failed to properly understand the importance of constituency-based democracy”. He also faults the government’s lawyers “for not making the case adequately either”, “and virtually all of the commentary” on Frank, including presumably my comment, for ignoring the issue. Yet in his view, it ought to have been a decisive consideration:

[W]e vote for local representatives. We don’t vote for parties, or party leaders, no matter what we may have in mind when we go into the ballot box – we mark the X for the local candidate, end of story. For an expat, it’s not the connection to Canada that should be at issue – it’s the connection to the riding, because that’s how we allocate our votes.

One might, of course, reproach the government lawyers for failing to emphasize this particular rationale for disenfranchising Canadians abroad. The Frank majority, even on this view, is blameless, because it wasn’t at liberty to sustain the disenfranchisement on the basis of a justification that the government did not even put forward. Section 1 of the Canadian Charter of Rights and Freedoms provides that limits on Charter rights, including the right to vote, must be “demonstrably justified” ― and as the Supreme Court has long held, it is the government that must carry out the demonstration. But there are other reasons, based in both what we might (loosely) call civics and feelings, as well as some realism about politics, that mean that, had the government chosen to make connection to the riding as the hill its case would die on, this case would be every bit as dead as it now is.


Start with the civics. Mr. Smith is quite right that, in point of law, we vote for local representatives, not for national parties or their leaders. Whether this ought to matter as much as he suggests, I will discuss below, when I turn to politics. But it’s important to consider a couple of other legal issues.

First, though there seems to be a good deal of confusion or even obfuscation on this point, the Canada Elections Act already takes care of the need for a connection between a Canadian voting from abroad (who may be a short-term expatriate, a long-term one previously allowed to vote, such as a diplomat’s family member, or a newly-enfranchised long-term expatriate). Paragraph 223(1)(e) provides that, when applying to be registered as an elector resident outside Canada and requesting to vote by special ballot, a would-be voter must provide the Canadian address to which his or her vote will be tied. Once the choice has been made, section 224 prevents the voter from changing it. This prevents forum shopping, as it were, and seems a sensible regulation.

Now, there is a range of options for the prospective voter from abroad to choose from:

the address of the elector’s last place of ordinary residence in Canada before he or she left Canada or the address of the place of ordinary residence in Canada of the spouse, the common-law partner or a relative of the elector, a relative of the elector’s spouse or common-law partner, a person in relation to whom the elector is a dependant or a person with whom the elector would live but for his or her residing temporarily outside Canada.

It has been put to me that the breadth of this range is excessive and gives the elector too much choice. If Parliament agrees, it can eliminate some superfluous options by legislation; this should not be constitutionally problematic. But I don’t think that Parliament should do this. On the contrary, giving the voter the ability to tie his or her vote to a former residence or a family member’s one makes it more likely that the elector will choose to vote at the particular place in Canada to which he or she is feels the strongest connection, which will not be the same for all expatriates, and which each voter is much better positioned to figure out when registering than Parliament when legislating.

Second, one must keep in mind that when it comes to voters in Canada, the law does not require any sort of evidence of a connection between the voter and his or her riding other than the fact that the voter resides there. Perhaps that’s because residence is simply deemed to be determinative of the community to which the voter belongs. But this seems a very rough assumption, especially in today’s urbanized world, in which many ridings are quite compact and the boundaries between them, fluid. A voter might be live in a bedroom community or a residential neighbourhood, but work in a downtown in a different riding, and perhaps have other attachments in yet a third one. It is, to say the least, not obvious which of these the voter is genuinely connected to. Residence, arguably, is only the most easily administrable way of sorting voters into ridings (both at the point of counting them through the census and at the point of registering them), simply because it tends to be more stable than other connections. As Chief Justice Wagner, writing for the Frank majority, put it, “residence can best be understood as an organizing mechanism for purposes of the right to vote”. [28] It is nothing more than that.


This brings me to what Mr. Smith might calls “feelings”. He and others who defend the disenfranchisement of Canadians abroad are very quick to demand that expatriates meet conditions that are not imposed on other Canadians to qualify for the franchise. Whether it be some subjective connection to a riding or to Canada as a whole, or knowledge about the local state of affairs, or tax liability, or subjection some undefined but substantial number laws, not all residents will meet these conditions that are said to justify denying the franchise to expatriates. But no one thinks to inquire into whether they really do, and no one, I’m pretty sure, would accept (re-)introducing tests of this nature into our election laws. Expatriates are the only people whom people judge on such criteria.

Indeed, it is not so much a judgment as prejudice. Expatriates are simply assumed to fail such tests ― and arguments to the contrary are dismissed as “feelings”. Mr. Smith guesses that Canadians who live abroad cut themselves off from communities where they used to live, or have family, or intend to return (or all of these things). Why? My personal experience, for what that’s worth, is that I keep up with the news from Québec and Montreal (and occasionally write on Québec-specific issues), more than from other provinces. Do I specifically track the news for Notre-Dame-de-Grâce-Westmount, where my parents live and I will vote in October, if I can be bothered? Not particularly, but then again, I wouldn’t even if I actually lived there. To say that I’m not a suitable voter for this reason would be applying a groundless double standard.

And speaking of double standards (and, I suppose, of civics), it’s worth noting that pursuant to section 222 of the Canada Elections Act some long-term expatriates are already allowed to vote: namely, members and employees of the Canadian forces, federal provincial public servants, employees of “international organization[s] of which Canada is a member and to which Canada contributes, as well as anyone who “lives with” such voters. The rationale for this is, presumably, that all such persons ― not just public servants, mind you, but their family members too ― are deemed to maintain a connection with Canada that other expatriates lack. Yet even assuming that this is so, is it remotely plausible that such persons (who, if anything, probably tend to be more mobile than the average voter even when they live in Canada) maintain their special connections to their home ridings? I really don’t think this is plausible, and so, the invocation of the riding connection as a justification for disenfranchising some, but not all, expatriates is another sort of unwarranted double standard.


Let me finally turn to politics ― and, specifically, to the need to be realistic about it. If we want to understand the rules of elections and government formation in Canada, we must keep in mind that each voter only casts a ballot for a local representative, not (directly) for a party or Prime Minister. But if we want to figure out whether Parliament is justified in preventing a person or a class of persons from voting, I don’t think it makes sense to pretend, as Mr. Smith asks us to, that this is all that matters. The reality, as he more or less acknowledges, is that what we “have in mind when we go into the ballot box” ― or at least the voting booth, for the less acrobatic among us ― very much has to do with parties and, especially, their leaders, for most voters.

Political parties themselves know this. The big ones tried to prevent to keep the small ones from getting their names on ballot papers, until the Supreme Court wisely put an end to that in Figueroa v Canada (Attorney General), 2003 SCC 37, [2003] 1 SCR 912, because candidates not identified with parties get fewer votes. They give pride of place to party names, logos, slogans, and leaders in their advertisements. They make sure their MPs have lookalike websites in party colours. Local candidates are often little more than props for a leader’s tour. I’m too lazy to look for the relevant research (if it exists) right now, but as a not-so-wild guess, I’m inclined to think that many voters don’t even remember the name of their local candidate when they go vote. This may be regrettable, but the parties themselves have ensured that it doesn’t matter; what does matter is the party identification on the ballot paper.

One key reason for this is that election campaigns are largely national events, not local ones. (By way of thought experiment, imagine we didn’t hold simultaneous general elections, but renewed the House of Commons with staggered elections, one riding at a time. Our politics may well be quite different ― and more local. But of course we don’t do that.) The centrality of leaders’ personalities to election campaigns makes this unavoidable, and an even starker phenomenon than in the past. But even to the (limited) extent that voters are preoccupied with actual issues rather than personalities, the issues are largely national in scope. This is perhaps especially the case in federal elections, since Parliament’s powers are, by design, largely those that concern the country as a whole. Admittedly Parliament doesn’t always keep to its jurisdiction. Even when it does, Justice Rowe points out in his concurring reasons that “federal policy can impact different geographically defined communities in different ways”. [89] Still, federal elections aren’t about the quality of your local school or the regularity of garbage removal from your street. Most voters, especially in federal elections, just aren’t especially concerned with riding-level matters. To say that expatriates, and only expatriates, ought to be disenfranchised because they aren’t is, once again, to apply an unwarranted double standard.


The existing law already ensures that Canadians voting from abroad cast their ballots in the ridings to which they have the strongest connections. At the same time, it does not require the existence of a very meaningful connection between any voters, including those resident in Canada, and their ridings. The idea that expatriates should be prevented from voting because they lack such a connection is thus a double standard. Moreover, Canadian elections, especially federal ones, aren’t local affairs anyway. For all these reasons, had the government argued that Parliament was entitled to deny expatriates the franchise because of their supposed detachment from the ridings in which their votes would be counted, it would have fared no better than it actually did in Frank.

Doing Right on Rights

Why the Supreme Court was right to find the disenfranchisement of Canadians abroad unconstitutional

In my last post, I summarized the majority, concurring, and dissenting opinions delivered in Frank v Canada (Attorney General), 2019 SCC 1, which held that the denial of the franchise to Canadians who have resided outside the country for more than five years is unconstitutional. As noted there, I believe that the majority, whose opinion was delivered by Chief Justice Wagner was correct so to hold. In this post, I explain why, and also make some observations about the strongly-worded dissent by Justices Côté and Brown. To make my biases clear once again, I remind readers that I am a Canadian abroad myself, and have been for six of the past eight years. While I vote rarely and reluctantly, I am emotionally invested in the issue of whether my right to do so can be taken away. Feel free to discount the following accordingly.


While I am the subject of emotions, let me say this. Justices Côté and Brown assert that the denial of the right to vote to Canadians abroad “is not a distinction based on moral worth”. [168] By my lights, that’s precisely what it is. On their own view, it is the product of “normative conceptions of what the Canadian political community is, and how it can best be protected and made to flourish”. [139] The conception embraced by Parliament in enacting the provisions invalidated in Frank sees the Canadian political community as excluding Canadians abroad, who are consequently less worthy of the franchise. Indeed, they are, according to the dissent, a threat from which the community must be protected, since it is “unfair to Canadian residents for their lawmakers to be elected by long-term non-residents who have no connection of any currency to their electoral district”. [153]

This is utterly wrong. As the Chief Justice rightly observes, many Canadians who live abroad maintain strong ties with Canada (and, I would add, often with the local community where they used to live and, in many cases, intend to return). As the Chief Justice also says, “[c]onversely, there may be citizens who have never left Canada but whose subjective commitment to the country is much weaker and who are less well versed in local issues”. [68] Indeed, though it would be impolitic for a judge to say so, “may be” in this sentence should read “obviously are”. As, for example, Ilya Somin and Bryan Caplan respectively have shown, voters are both ignorant and irrational ― rationally so, but ignorant and irrational all the same. Yet we would not generally accept disenfranchising voters on that basis; we do not inquire into the degree of connection a voter has with his or her local community, or the country, before issuing him or her a ballot paper. It is only, it seems, in the case of expatriates that these things actually matter. To me, this is strong evidence that what is at work here is not really a concern with the fairness or integrity of Canada’s electoral system, but a judgment, or rather prejudice, about the moral worth of those Canadians who are taken to have left the community, and must “rejoin[]” [153] before being allowed to take part in the community’s affairs.

Now, it is true that most Canadians abroad do not, in fact, go to the hassle and the expense of requesting a special ballot and returning it to Elections Canada. This means, of course, that the idea of non-resident voters swamping elections and deciding them at the expense of residents is far-fetched if not entirely implausible. But more importantly, as the Chief Justice points out, this means that those who do take the trouble value their involvement in Canada’s political life ― probably more so than a great many of their resident fellow-citizens. In the Chief Justice’s words, they “demonstrate[] a profound attachment to Canada”, [75] and it seems absurd to pretend otherwise on the basis of no evidence whatsoever.


This brings me to another issue: that of the correct approach to deciding whether legislation is unconstitutional because it unjustifiably contravenes the Canadian Charter of Rights and Freedoms. Justices Côté and Brown issue what they regard as a profound challenge to the way we not only discuss but also think about the relationship between Charter rights and policies that interfere with them. They note that it is commonplace to speak of such policies as “infringing” or even “violating” rights, only for these “infringements” or “violations” to be upheld, or “saved” by applying section 1 of the Charter. Yet, as they further point out, section 1 provides that the Charter “guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society” ― limits, not infringements or violations. So we should stop talking about justified infringements, and talk about limits instead. The majority is not interested, observing that the words “limits” and “infringements” have long been used interchangeably. (The Chief Justice is probably too polite to note this, but I am not: Justices Côté and Brown themselves spoke in terms of “infringement” as recently as seven months ago, in their excellent dissent in Law Society of British Columbia v Trinity Western University, 2018 SCC 32.)

I take the dissent’s textual point. The word section 1 uses is “limits”, and it arguably makes more sense to say that the state can justifiably limit rights, not that it can violate them. But I fail to see what great change to the way we actually think about rights and scrutinize the state’s possible interference with them follows from this. Contrary to some, indeed quite a few, of the participants in this blog’s recent 12 Days of Christmas symposium, Justices Côté and Brown show no interest in reconsidering the test for verifying the permissibility of limitations on rights that was first set out in R v Oakes, [1986] 1 SCR 103. They apply the same proportionality analysis, warts, subjectivity, and all, under the label of assessing the “limit” on the right to vote as the majority does when considering “infringement” of this right. They have much to say about the fact that the right to vote is a “positive” one, requiring legislation to “breathe[] life into” it, [142] but that it is true of many of the rights the Charter protects (notably, but by no means only, the various rights of criminal suspects and accused) and, more importantly, it simply does not follow that the legislation that “breathes life into the right” may not fail to do so to a constitutionally required standard. (Similarly, Justices Côté and Brown point out that the legislation imposing a five-year expiry period on expatriates’ franchise replaced that which gave no expatriates the vote. So what? The test of constitutionality is not whether Parliament comes closer to respecting the Charter than it once did, but whether it respects the Charter now.)

The real methodological disagreement between the dissent and the majority (as well as Justice Rowe’s concurrence) has to do with the level of deference each accords Parliament. The majority insists, at the outset, that “[a]ny limit on the right to vote must be carefully scrutinized and cannot be tolerated without a compelling justification”. [1] Although, as noted in yesterday’s post, it later wavers a little on this point, its approach is, indeed, one that refuses to take the government’s claims about the need for or usefulness of the legislation it examines on faith. By contrast, the dissent waxes deferential, “eschews rigid and technical application” [124] ― it’s not quite clear of what, but presumably of the justification requirements ―, and plays up “Parliament’s policy-making expertise”. [126]

Yet here the dissenters’ professed textualism fails them. To repeat, section 1 of the Charter says that “only such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society” (emphasis mine) can be countenanced. The constitutional text, therefore, requires a demonstration ― not judicial acquiescence on the basis that Parliament knows best. And this requirement, in turn, suggests that contrary to what Justices Côté and Brown say Charter rights can only be limited in order to deal with an identifiable problem, to address some specified mischief, and not merely because a legislature thinks that in an ideal political community these rights would be limited in this way. The enactment of the Charter, as a law superior to ordinary legislation, has taken that moral judgment out of the Canadian legislatures’ hands. (To be clear, this is a separate question from that of the permissibility of what used to be called “morals legislation”, which proscribes what the legislature sees as immoral behaviour, such as taking drugs. There is no argument that expatriates act immorally by voting.)

In short, by all means, let’s be careful with our language, and speak of reasonable limits on rights instead justified violations. But let’s also insist that limitations on rights, to qualify as reasonable, must be ones whose justification is capable of being demonstrated, rather than merely asserted, and is in fact demonstrated, rather than taken for granted. Indeed, I think that this substantive concern is rather more important than the semantic one. In Frank, it is the majority, not the dissent, that follows an approach that is closer to that required by the constitution.


In my view, the majority decided Frank more or less as it should have. It correctly insisted that any exclusion from the franchise except the one based on citizenship must be justified. Contrary to the dissent’s strident warnings, it does not follow that no exclusions could ever treated as reasonable limits ― only that the justification process for upholding these limits cannot be elided by saying that some limits on the right to vote must obviously be admissible. This goes even for the denial of the right to vote minors (which, as Ilya Somin, for example, has argued, is not as self-evidently reasonable as the dissenters would like us to think). I think that it would be quite easy to sustain the disenfranchisement of three-year-olds, the dissent’s scaremongering example, should anyone challenge it; but as for the denial of the vote to teenagers, I for one wouldn’t mind seeing the governments put through their justificatory paces.

The majority is also right to be skeptical of the government’s arguments based on an ill-defined “social contract” of which expatriates are allegedly no part. I’m not too impressed with the Chief Justice’s definition of Parliament’s objective in terms of “fairness” ― fairness is too capacious a word for my liking, and the dissent’s suggestion that Parliament was pursuing a currency of relationship between the voter and his or her community seems closer to the mark. But one should also acknowledge that objectives that are largely symbolic (or, as Justices Côté and Brown see it, moral) do not lend themselves to easy definition, and so inevitably compromise the quality of the Charter analysis. The majority’s skepticism about the existence of a rational connection between the objective of fairness and disenfranchisement of Canadians abroad is also warranted. Indeed, I would have liked the Chief Justice to have been bold enough to say that, given both the possibility that expatriates maintain current links with Canada and their Canadian communities, and the lack of any assessment of whether any other voters do so, disenfranchising expatriates alone based on their deemed lack of connection to Canada is actually irrational. But the Chief Justice is not the first judge not to want to go there.

I’ll live with that, because the Chief Justice ends up making the right arguments at the “minimal impairment” stage of the analysis. Like I had in criticizing the decision of the Court of Appeal for Ontario in this case, he points out that Canadians abroad can and do maintain close links with Canada; that they are affected by Canadian laws and government policies; that, moreover, “Parliament can change laws on its own initiative and thus alter the extent to which Canadian legislation applies to non-resident citizens”, [72] so that one cannot invoke the limited scope of current extra-territorial legislation to deny expatriates the vote ― I had called this “let[ting] the statutory tail wag the constitutional dog”; and that in any event “attempting to tailor Charter rights to the extent to which citizens are burdened, or not burdened, by Canadian laws would be an impossible exercise”. [71] If all this is not enough to qualify the disenfranchisement of expatriates as irrational, than it certainly suffices, as the Chief Justice says, to show that it is grossly overbroad, and not “minimally impairing” of the right to vote.

The Chief Justice’s reasons are not perfect. He does, as the dissent points out, get somewhat carried away in patriotic praise for the Canadian democracy, and there are shades of what I’ve been calling “constitutionalism from the cave” in his claim that “a broad interpretation of” the Charter‘s guarantee of the right to vote “enhances the quality of our democracy and strengthens the values on which our free and democratic state is premised”. [27] Constitutional interpretation must aim at ascertaining the text’s meaning, not at strengthening values or anything of the sort. Still, what the Chief Justice’s opinion for the majority does is substantially in line with the constitutional text, despite its rhetorical imperfections. (And still on the subject of rhetoric, or style: can someone please ask the Chief Justice and his colleagues to stop using the initialism “AGC”? It is a recent innovation ― the Court hadn’t done it at all before 2010 ― and not a good one. It is ugly and has a jargon-y feel that is quite at odds with the Chief Justice’s stated desire to make the Court’s work more accessible, including to laypersons.) But as majority opinions of Supreme Court have gone in the last few years, the one in Frank is up there with the best.


The Supreme Court has held that Canadians who live abroad cannot be disenfranchised based on, in effect, stereotypes about their lack of relationship with their home country. This is a relief. In doing so, it has insisted on carefully and critically examining the government’s claimed reasons for limiting Charter rights, and this is a good thing too, one that doesn’t happen often enough. There is something to learn from the dissenting opinion, too, about our constitutional vocabulary, and I hope that this lesson is not lost just because the substance of that opinion is bitterly disappointing. But the constitution was upheld in Frank, and so right was done. May 2019 bring more of that.

“A Profound Attachment”

The Supreme Court holds that disenfranchising Canadians abroad is unconstitutional

Yesterday, the Supreme Court at last delivered its judgment on the constitutionality of disenfranchising Canadians abroad, Frank v Canada (Attorney General), 2019 SCC 1. By five votes to two, the Court holds that disenfranchisement is indeed unconstitutional. This is, as I have long argued (especially in criticizing the decision to the contrary by the Court of Appeal for Ontario), the right result. Full disclosure, in case this is necessary: I am myself a Canadian abroad, and while I would not have been disenfranchised at the coming election under the rules the Supreme Court has found unconstitutional, and am only an occasional and reluctant voter anyway, I am emotionally invested in this issue.

Section 3 of the Canadian Charter of Rights and Freedoms provides that “Every citizen of Canada has the right to vote in an election of members of the House of Commons or of a legislative assembly and to be qualified for membership therein.” The Canada Elections Act sets out the details of how this right can be exercised ― and denies it to some groups of citizens. One of these disenfranchised groups consists of Canadians who have not resided in Canada for more than five years, although those who are representatives of a Canadian government or members of the Canadian forces, as well as members of such persons’ families, are not subject to disenfranchisement.

The government conceded that denying their right to vote breached section 3 of the Charter, but contended that the breach was justified as a reasonable limit authorized by the Charter‘s section 1. The majority ― Chief Justice Wagner (who wrote the majority opinion) and Justices Moldaver, Karakatsanis, and Gascon ―, as well as Justice Rowe, who concurs, reject this view. In dissent, Justices Côté and Brown say that the denial of the franchise to Canadians abroad is justified.


Relying on the Court’s decision in Sauvé v Canada (Chief Electoral Officer), 2002 SCC 68, [2002] 3 SCR 519, which invalidated the disenfranchisement of long-term prisoners, the Chief Justice writes that the right to vote must be given a “broad and purposive interpretation”, and “any intrusions on [it] are to be reviewed on the basis of a stringent justification standard”. [25] The Chief Justice rejects deference to Parliament, insisting that “reviewing courts must examine the government’s proffered justification carefully and rigorously”. [43] Unlike in cases that involve “complex” or “nuanced” choices among competing priorities, deference “is not the appropriate posture for a court reviewing an absolute prohibition of a core democratic right”. [44] Later, however, when considering whether the prohibition is “minimally impairing” of the right, the Chief Justice grants that “some deference must be accorded to the legislature by giving it a certain latitude”. [66]

As for residency requirements for voting, they are “an organizing mechanism”, “an important device” [28] that helps structure our electoral system, but have no constitutional value in themselves: “In clear language, the Charter tethers voting rights to citizenship, and citizenship alone.” [29] In any case, there already are Canadian citizens who are allowed to vote from abroad, suggesting that residence within cannot be an implicit pre-condition for having the right to vote.

Applying the test for the justification of Charter infringements set out in R v Oakes, [1986] 1 SCR 103, the Chief Justice begins by rejecting the idea, accepted by the Ontario Court of Appeal, that “preserving the social contract” whereby citizens’ obedience to laws is exchanged for a say in making them as a pressing and substantial objective capable of justifying the infringement of rights. While Sauvé had invoked the language of social contract theory, correctly understood, it stands for the proposition “that deeming that a citizen has ‘withdrawn’ from the social contract is not a legitimate basis for denying him or her the right to vote”. [52] However, the Chief Justice accepts that “maintaining the fairness of the electoral system to resident Canadians”, [55] which he seems to interpret by focusing on the existence of a connection between voters and the Canadian polity, is an important governmental objective.

At the second stage of the Oakes test, Justice Wagner finds that the government “has not definitively shown that a limit of any duration” on the ability of Canadians to vote from abroad “would be rationally connected to the electoral fairness objective advanced in this case”, [60; emphasis in the original] but declines to reach a firm conclusion. He argues, however, that neither the existence of residence requirements for voting in provincial elections nor the prevalence of such requirements abroad make their imposition by Parliament rational, and observers that “there is no evidence of the harm that these voting restrictions are meant to address”, [63] or even any complaints about those non-resident citizens who already are able to vote.

As often, it is the next stage, originally described as that of “minimal impairment” although the word “minimal” has not been taken literally, that is crucial. The Chief Justice finds that disenfranchising Canadians after five years abroad, “[f]ar from being a measure that is carefully tailored so as to impair voting rights no more than is reasonably necessary, … seems to have been simply a ‘middle-of-the-road’ compromise”. [67] There is no “correlation between, on the one hand, how long a Canadian citizen has lived abroad and when he or she intends to return and, on the other hand, the extent of his or her subjective commitment to Canada”. [68] Indeed, whether the issue is knowledge of and commitment to Canada, the impact of Canadian laws on a given voter, many Canadians abroad will be better qualified as voters than those residing in the country. Chief Justice Wagner concludes by noting that “[a] non-resident citizen who takes the trouble to vote by way of special ballot … has demonstrated a profound attachment to Canada. We have nothing to gain from disenfranchising such citizens.” [75] In the same vein, he notes that any positive effects of this disenfranchisement are speculative, while the negative impact on those disenfranchised is real and present.


Justice Rowe agrees that disenfranchising Canadians who live abroad is unjustified, but writes separately to emphasize the “significance and centrality of residence to our system of representative democracy”. [84] He details the history of residency requirements in Canadian election legislation (including the slow expansion of voting rights for Canadians abroad), and pointedly rebukes the majority by claiming that “residence has been historically and remains today more than just an ‘organizing mechanism’. It is foundational to our system”. [90] As a result, Justice Rowe says, while “[s]ection 3 [of the Charter] protects the right to vote … it does not follow as a corollary that there is a right to vote in the constituency or province of one’s choosing”. [91] Provincial and territorial residency requirements, in particular, would be subject to different considerations than federal ones (including because provincial laws are more local in nature and applicability than federal ones).

Whatever might be justifiable in other cases, however, Justice Rowe concludes that the disenfranchisement of long-term expatriates is not. He accepts that it pursues the objective of electoral fairness, although he notes that fairness for resident citizens is being pursued at the expense of non-residents. Justice Rowe also accepts that fairness can reasonably be pursued by preventing “those who are largely unaffected (non-residents) [from] participating in decisions that would affect others (residents)”. [103] Passing over the question of whether the disenfranchisement of Canadians abroad is minimally impairing of their right to vote, he moves on to the balancing of its salutary and deleterious effects. The former, he finds, are “negligible”, [106] since very few expatriates actually vote. The latter are not. Expatriates who are disenfranchised “may not feel the local consequences of particular federal policies in the constituencies in which their votes would be counted, [but] they stand nonetheless to be affected by certain federal laws and policies, perhaps in life altering ways”. [107] As a result, the disenfranchisement of Canadians abroad is not justified.


Justices Côté and Brown dissent. They not only disagree with the outcome reached by the majority and Justice Rowe, but want to approach the issue quite differently. They stress that the right to vote “is a positive right which, unlike most Charter rights, requires legislative specification in order for the right to be operative”, [113; emphasis in the original] so that the denial of the franchise to expatriates is not the product of legislative action, but of a “failure to extend the right to vote” to them. [128] This right is also not absolute: “Nobody suggests that s[ection] 3 entitles three-year-old Canadian citizens to vote.” [114] Indeed, they deny that the legislative provisions at issue “disenfranchise” long-term expatriates, since they had not been allowed prior to these provisions’ enactment, or ever. They also accuse the majority (and, implicitly, any number of past judgments) of “distort[ing] the limitations analysis” [120] by speaking of a “breach” or “infringement” of the right to vote rather than of a “limitation” on this right, as the terms of section 1 of the Charter would suggest. (A breach, they insist, is caused by a limitation that is not justified.)

Thus the real question, Justice Côté and Brown argue, is whether the long-term expatriates’ right to vote has been reasonably limited. The way to answer this question is to apply the Oakes test. However, while they make a point of agreeing with the majority that the burden of justification under this test rests on the government, Justices Côté and Brown insist, citing the dissenting opinion in Sauvé (without acknowledging that they are relying on the dissent) on “a ‘flexible contextual approach’ … one that eschews rigid and technical application”. [124, citing Sauvé at [84]] They also argue that it is wrong to look for “a concrete problem or mischief” that rights-limiting legislation is meant to address, because it is “undeniable … that Parliament can constitutionally legislate in pursuit of, or in response to, considerations of political morality or philosophy”. [126] There is “moral nuance inherent in defining and defending the boundaries of rights — that is, in justifying rights limitations” — and, like “Parliament’s policy-making expertise”, it must be “afford[ed] due respect”. [126]

Justices Côté and Brown define Parliament’s objective as “privileg[ing] a relationship of some currency between electors and the communities in which they are eligible to vote”. [132] (In doing so, they spend four extensive paragraphs cautioning against reliance on statements by individual legislators during the course of parliamentary debate… and conclude by pointing to statements that support their understanding of the objective.) This objective “is clearly inspired by a particular moral philosophical understanding of the relationship between citizen and state in a democracy”. [140] Indeed, electoral “legislation is never designed to solve a problem or address a particular mischief. Rather, it breathes life into the right [to vote] so that it may be recognized and exercised.” [142] While limitations on the right to vote require justification, Justices Côté and Brown attack the majority for considering that, other than citizenship, “all other specifications [of this right] are necessarily unconstitutional”. [142] Justices Côté and Brown note that other groups are excluded from the franchise ― they mention citizens who have never resided in Canada and minors ― and argue that these exclusions too must be regarded as examples of Parliament’s permissible pursuit of philosophical objectives. Indeed, they say, majorities in Sauvé and here have acted in furtherance of philosophical views of their own.

Ultimately, ensuring a current relationship between voters and their communities is a pressing and substantial objective because it “ensures reciprocity between exercising the right to vote and bearing the burden of Canadian laws” [152] and “protects the integrity of the Canadian electoral system, which is founded on geographical representation”. [153] This integrity would be undermined by allowing people to vote in constituencies with which they lack a community of interest.

Justices Côté and Brown also consider that the limitation of the right to vote from abroad to those citizens who have not been outside Canada for more than five years is reasonable and therefore proportionate to Parliament’s objective. They insist that, under the majority’s reasoning, no time limit on voting rights could be upheld, including for provincial elections. They add that the majority is wrong to ignore the treatment of expatriates’ voting rights by New Zealand, Australia, and the United Kingdom: “the majority’s patriotism risks descending into exceptionalism”, and blinding it to “some lessons” that “Canada would well have taken … from other countries” [166] (or at any rate from New Zealand, which enfranchised its aboriginal people and women well before Canada did). Indeed, the majority’s position is “highly political, rhetorical”, and “in tension with the majority’s own invocation of internationalism and of a ‘globalized’ world of connectivity and communication”. [167] As for the effects of the legislation, the deleterious ones are minimized since the denial of expatriates’ voting rights “is not … based on moral worth”, [168] while the salutary ones ― which consist in the attainment of Parliament’s objectives ― are considerable.


As I noted at the outset, I believe that the majority is correct (though Justice Rowe makes some valuable points about provinces and territories). The dissent, I confess, perplexes me. But this post is much too long as it is. I shall publish my comment separately ― and quickly, I hope.

Constitutionalism from the Cave

The constitution is a binding law, not just an incomplete statement of political ideals

The imbroglio with the Ontario legislature’s enactment of Bill 5 to restructure the Toronto City Council a couple of months before an upcoming election, the Superior Court’s declaration of that legislation unconstitutional, the threatened invocation of the “notwithstanding clause” to override that declaration, and the Court of Appeal’s restoration of what little sanity could still be restored by reversing the Superior Court’s decision has generated a great deal of commentary. Some of this commentary has been very imaginative indeed in coming up with constitutional arguments that would have advanced the commentators’ preferred policy agendas and forestalled the seemingly obvious legal conclusions.

Of course, such a creative argument had prevailed at the Superior Court, which (as for example co-blogger Mark Mancini, as well as yours truly, explained) ignored clear constitutional language on its way to finding that Bill 5 violated the guarantee of freedom of expression in the Canadian Charter of Right and Freedoms. Even more unorthodox reasoning was unleashed in attempts to argue that the Ontario legislature couldn’t actually invoke the “notwithstanding clause”, the constitutional text once again be damned. Mark has written about open letter in which professors who wouldn’t dream of treating originalism as a serious interpretive methodology suddenly turned original-intentist ― but that, at least, was an explicitly political text. Other arguments along similar, or even more outlandish, lines purported to be legal ones.

This outburst of creativity is, of course directly related to a certain way of seeing the constitution that is prevalent in the Canadian legal community (including, but not only, in the academy). On this view, the Canadian constitution ― especially, though not only, the Charter ― is not so much a law that courts must apply as a sort of shadow in Plato’s cave, a vague reflection of true constitutional ideals that the judges must discover and explain to us cavemen. The constitution’s text is not in any meaningful way binding on the courts;  it is only an inadequate approximation, one whose imperfections judges can and ought to circumvent in an unceasing quest to get a clearer view of the ideal constitution. And, of course, this ideal constitution, just so happens to enact the political preferences of the persons urging this view, and presumed (often not incorrectly) to be shared by the judiciary.

Perhaps the latest contribution to the post-Bill 5 constitutional free-for-all is illustrative. It is a post by Colin Feasby, over at ABlawg, arguing that section 3 of the Charter, though it ostensibly only guarantees the right of Canadian citizens “to vote in an election of members of the House of Commons or of a legislative assembly and to be qualified for membership therein”, really also applies to municipal and other elections, and to referenda. Now, I have learned a lot from Dr. Feasby’s law of democracy scholarship (which has also been cited and relied on by the Supreme Court!). But this post is nothing more than a call for the judiciary to wilfully re-write the constitution we have so as to bring it closer to an idealized view of what a good constitution ought to be according to Dr. Feasby (and many others).

Dr. Feasby argues that “the lack of constitutional protection for important democratic processes” ― such as municipal elections ― “is an unnecessary defect in our constitutional arrangement”. “Unnecessary” a word that I wouldn’t use, and as will presently be apparent, Dr. Feasby uses it advisedly. However, I agree with him to this extent ― the lack of constitutional protections for municipal institutions is indeed a shortcoming of our constitution, as I have suggested here.

Dr. Feasby, though, is not suggesting a constitutional amendment to remedy the defect he identifies. Rather, he “proposes a way that the Supreme Court of Canada can remedy this defect”. He argues that “courts have a role in ensuring that the democratic process functions so that the sovereign will of the electorate may be expressed without distortion”, which is true, if somewhat beside the point in matters where the will of the electorate is not, legally speaking, sovereign, including in municipal elections. The question, though, is how far the courts’ role extends. Dr. Feasby thinks it allows the courts to embrace what he describes as a “purposive and … generous approach to interpreting Charter rights”, and “impose a rule” according to which

Where a government, Federal or Provincial, delegates a legislative role to a democratically chosen body or where a government, Federal or Provincial, effectively delegates a decision to the electorate in a referendum, section 3 of the Charter applies.

In other words, “a body elected in processes governed by section 3 cannot delegate its power to an elected body chosen by electors with lesser constitutional protections”.

Dr. Feasby anticipates two objections to his proposal. First, he expects people to argue that it would get in the way of worthy reforms of municipal and other institutions. His response is that “so long as those changes are consistent with the principles that animate section 3 of the Charter“, nothing would get in their way. Fair enough, I suppose. The other objection Dr. Feasby foresees is based on the concern about section 3 claims being brought by people who are not in the intended electorate for a given election (say, the residents of a municipality). Such claims should simply be rejected ― as would that of “Canadian citizens resident in Alberta” demanding “the right to vote in Provincial elections in Quebec”. That too seems fair enough.

There are other, more serious, problems with Dr. Feasby’s argument, however. A practical one is that, even in the form given to it by Dr. Feasby, it reaches very far indeed. Municipalities, band councils, and school boards are not the only entities that might be described as “democratically chosen” entities to which governments delegate legislative powers. Various professional bodies (such as law societies) and agricultural marketing boards come to mind; so do, perhaps, universities, whose powers ― which include the ability to regulate large swathes of student and staff conduct ― are ultimately exercised by (partly) elected boards and senates. (Whether the universities are subject to the Charter in at least some areas is an open question, but there are good arguments for that view.) It’s not at all obvious to me that, “the principles that animate section 3 of the Charter” can be usefully applied to such bodies, even assuming that they can be to municipalities and school boards. And it’s not at all obvious that the argument for rejecting the claims of persons excluded from the relevant electorate ― say, the consumers of professional services or of agricultural products ― can be dismissed as easily as  those of Albertans looking to vote in the Belle Province.

This problem becomes even more pressing if we take up Bruce Ryder’s suggestion “that a province that is bound by s.3 democratic norms shouldn’t be able to do an end run around them after creating subordinate governments” ― seemingly regardless of whether these “subordinate governments” are themselves meant to be democratically elected. If this principle were taken seriously, it would amount to a non-delegation doctrine on steroids, preventing the exercise of legislative power by undemocratic bodies ― which means pretty much all of the administrative state. Anti-administrativist though I am, even I don’t actually favour this approach. In truth, I don’t suppose that Professor Ryder favours it either. He simply makes an argument that furthers his preferences in a particular controversy, and doesn’t worry about its implications in future cases. I’m afraid this is symptomatic of the treatment of the constitution not as a law, but as a series of results-oriented propositions subject to permanent revision from one case to the next.

Equally symptomatic of this way of thinking is the fact that Dr. Feasby apparently does not see coming another objection, a principled rather than a practical one. This objection is, quite simply, that his proposal is a perversion of the constitutional text, a blatant attempt to expand it beyond what it was quite clearly designed to do, and what it not only originally meant but still means. Even if one believes that the constitutional text should be read according to the meaning of its terms today, “an election of members of the House of Commons or of a legislative assembly” does not mean “an election of members of a municipal council”, or “an election of the benchers of a law society”. Dr. Feasby invokes the “living tree” view of the constitution, but he advocates something different than just reading the text in light of evolving social mores or trying to use “progressive interpretation” to “accommodate[] and address[] the realities of modern life”, as the Supreme Court put it in Reference re Same-Sex Marriage, 2004 SCC 79, [2004] 3 SCR 698, at [22]. It’s not as if municipal election were a new phenomenon unknown to the framers of the Charter. It’s just that Dr. Feasby thinks that the Charter is only an imperfect statement of an “aspiration … to be a truly free and democratic society”, which can be given whatever contents a court, under the guidance of progressive advocates, can come up with in a given case.

Needless to say, I do not share this view. It is contrary to the terms of the constitution itself (specifically, section 52 of the Constitution Act, 1982, which provides both that “[t]he Constitution of Canada is the supreme law of Canada” (emphasis mine) and that “[a]mendments to the Constitution of Canada shall be made only in accordance with the authority contained in the Constitution of Canada” ― which says nothing about the Supreme Court amending the constitution in the absence of the political consensus required to do so. It is destructive of the Rule of Law. And it is especially galling because many of the same people who advocate this view of the constitution not as binding law but as merely suggestive of (their) political ideals demand that political actors ― such as the present Ontario legislature ― that do not fully share these ideals comply with judicial decisions based on them. I think it’s right to demand that political actors comply with the law, including the law of the constitution. But why on earth should elected officials comply, not with the law, but with the philosophical preferences people who are not elected to anything? There can be no real constitutionalism in Plato’s cave. It’s time to climb out.

Deuxième Moisson

Tout comme il y a quatre ans, le DGE essaie de censurer une intervention de la société civile dans la campagne électorale québécoise

Les campagnes électorales ont leurs habitudes, leurs rituels. Les autobus, les slogans, les débats des chefs. Certaines de ces traditions sont communes à bien des sociétés démocratiques, d’autres sont plus locales. Une qui est particulièrement québécoise ― mais ne devrait pas pour autant être source de fierté ― c’est la lettre du Directeur général des élections (DGE) sommant un représentant de la société civile qui tente de se prononcer sur les enjeux de l’heure de se la fermer. Le rituel vient d’être renouvelé, comme le rapporte La Presse, avec cette fois Équiterre, dans le collimateur du DGE pour avoir diffusé les résultats d’un questionnaire remis aux principaux partis politiques et portant sur leurs politiques en matière d’environnement.

Je racontais un tel épisode, impliquant les producteurs d’un court documentaire critique du Parti québécois et de sa « Charte des valeurs », alias la Charte de la honte, lors de la campagne électorale de 2014. J’ai dit, à l’époque, que les penseurs et juristes « progressistes » qui ont cherché à limiter le rôle de l’argent en politique en limitant sévèrement les dépenses autorisées en période électorale récoltaient là ce qu’ils avaient semé. Ils s’imaginaient que les limites de dépenses feraient taire les riches, mais en réalité, elles s’appliquent d’abord à avant tout aux étudiantsaux syndicats ou aux individus impopulaires. En 2014, on a visé les défenseurs du pluralisme. En 2018, on vise les environnementalistes. La tendance, encore une fois, se maintient.

Il faut souligner qu’il y a quatre ans, le DGE avait alors fini par faire marche arrière ― au bénéfice de la liberté d’expression, mais au mépris de la Loi électorale. En tordant le sens des définitions pourtant claires de ce qui est et n’est pas une « dépense électorale » (prévues aux articles 402 et 404 de la Loi), le DGE a réussi à éviter l’opprobre médiatique qu’allait provoquer un épisode de censure. Mais la Loi électorale, elle, n’as pas été changée pour permettre à la société civile d’intervenir dans les campagnes électorales. Il n’est pas impossible, je suppose, que le DGE se démène encore pour ne pas censurer Équiterre, même si ce sera, comme je l’expliquerai à l’instant, très, très difficile. Cependant, même si la manoeuvre réussit, la censure ne sera que partie remise jusqu’à la prochaine campagne électorale. C’est à la Loi électorale, et non à son application par le DGE, qu’il faut s’attaquer pour régler le problème une fois pour toutes.

L’article 402 de la Loi électorale définit comme « dépense électorale »

le coût de tout bien ou service utilisé pendant la période électorale pour:

1° favoriser ou défavoriser, directement ou indirectement, l’élection d’un candidat ou celle des candidats d’un parti;
2° diffuser ou combattre le programme ou la politique d’un candidat ou d’un parti;
3° approuver ou désapprouver des mesures préconisées ou combattues par un candidat ou un parti;
4° approuver ou désapprouver des actes accomplis ou proposés par un parti, un candidat ou leurs partisans.

Cette définition s’applique aux dépenses des candidats et des partis aussi bien qu’à celles de la société civile, et il n’est pas surprenant qu’elle ratisse large. La production et diffusion du questionnaire d’Équiterre tombe sous le coup de cette définition, puisque celui-ci vise à diffuser certains aspect des programmes des différents partis et aussi, par l’usage de symboles visuels (coche verte, crois rouge) à approuver ou désapprouver les mesures préconisées par ceux-ci.

Deux problèmes se posent cependant. D’une part, il y a à la fois l’insuffisance et la vétusté des exemptions prévues à l’article 404. Contrairement à la disposition équivalente de Loi électorale du Canada, celui-ci n’exempte pas les communications d’un groupe (par exemple, un syndicat) à ses membres et n’est pas technologiquement neutre, exemptant la diffusion de nouvelles ou éditoriaux « dans un journal ou autre périodique » ou encore « par un poste de radio ou de télévision », mais pas par de nouveaux médias opérant sur internet. En 2014, le DGE a fini par décrire le documentaire en cause comme étant un « média citoyen » pour l’exempter de l’application de l’article 402. C’était, selon moi, à tort, puisque la Loi électorale n’exempte que certains médias, et n’autorise pas le DGE à en inventer de nouvelles catégories exemptées. Quoi qu’il en soit, je ne vois pas comment on pourrait user du même procédé pour aider Équiterre.

D’autre part, la Loi électorale limite excessivement les dépenses électorales des membres de la société civile. En fait, elle les interdit presqu’entièrement, ne faisant qu’une exception minimaliste à l’alinéa 13 de l’article 404, qui permet à un individu (ou un groupe de personnes ne possédant pas la personnalité morale) de s’enregistrer pour, ensuite, engager des dépenses d’au plus 300$ ― mais sans pourtant « favoriser ni défavoriser directement un candidat ou un parti ». Équiterre, si je comprends bien, est une personne morale, et ne pourrait se prévaloir de l’exemption, même si sa part du coût de la production du questionnaire dont on lui reproche la diffusion s’élevait à moins de 300$. De plus, il me semble clair que le questionnaire, même s’il se veut non-partisan, vise à favoriser l’élection de partis ayant des politiques environnementales qui reçoivent l’approbation d’Équiterre et à défavoriser l’élection des autres.

Ces restrictions sont draconiennes. Il est ridicule d’interdire aux acteurs de la société civile de prendre part au débat pré-électoral pour peu qu’ils choisissent d’obtenir la personnalité morale. Il est ridicule d’avoir un plafond de dépenses ― non-indexé, contrairement à celui des partis et candidats! ― de 300$. Il est ridicule d’exiger qu’une personne voulant engager des dépenses tout à fait minimes doive préalablement s’enregistrer auprès du DGE. Il est ridicule d’interdire les interventions qui favorisent ou défavorise l’élection de partis nommés. Même si l’on accepte le principe général de la limitation de dépenses et celui de la primauté des candidats et des partis en période électorale, les restrictions imposées par le législateur québécois sont ahurissantes. Elles ne sont pas justifiées. Elles sont, selon moi, inconstitutionnelles, même si la Cour d’appel du Québec en a déjà décidé autrement.

Ainsi, je pense que le DGE fait son travail en s’en prenant à Équiterre. Il applique la Loi électorale. Cependant, les dispositions en cause n’ont pas lieu d’être. Le législateur québécois devrait s’empresser de les revoir de fond en comble, sinon de les abroger. À défaut, ou d’ici là, c’est malheureusement à Équiterre d’en contester la constitutionnalité. Cette contestation ne sera pas facile, mais, selon moi, elle aura des chances réelles de succès. La Cour suprême a certes avalisé les dispositions de la Loi électorale du Canada limitant la participation de « tiers » aux campagnes électorales, mais, comme je l’ai déjà souligné, celles-ci sont bien plus permissives que celles de la loi québécoise. En attendant, le décret ordonnant la tenue d’élections générales demeure un bâillon.