Real-world elections are messier than you think

Both public opinion and the academic understanding of our electoral processes, especially here in Canada, has been slowly drifting away from the realities of real-world election administration. I think this is in part due to a sort of nationalistic pride Canadians have in their own democracy; Elections Canada is (justifiably) well trusted, and has its processes down to a fine science when it comes to administering traditional paper ballot elections. It’s easy to think we have it good when elections elsewhere in the world hinge on international observers scrutinizing the process, as their officials cannot even be trusted to be independent from the sitting government. But elections can only be so bulletproof due to a factor most academics miss: the people working them.

One recent area where this conflict in understanding clashed was the Supreme Court case in Terrebonne, which invalidated an entire by-election due to a clerical error made on 40 special ballot kits. This case surprised me (and some of my collegues) familiar with previous case law regarding election challenges: clerical errors are expected to a degree, and are even statistically accounted for.

Of course, there is some special context in this particular case that explains the decision: the margin in this particular election was small enough that this single clerical error could have reasonably changed the results, and insufficient efforts were made by election officials to remedy the error. But what surprised me is that the courts seemingly were under the impression that a subsequent by-election, given a similar margin of victory, would have also been anything other than a coin toss. This decision also seems to indicate that the magnitude of a clerical error is evaluated in proportion to the margin of victory from the official results. Does this mean that any district’s election that is won by an insufficient margin is now vulnerable to these kinds of legal challenges regardless, thus causing new by-elections to be held until the margin is sufficient? The courts might not see this as a real problem, but they’re also not the people hiring poll workers.

To take some commentary of my own that I wrote elsewhere:

The true difficulty of elections is not the process itself, but by the fact that once the writ is dropped, returning offices have to hire a staggering amount of staff on very short notice. This means an organization that starts with hundreds of staff all of a sudden is a semi-militarized structure with tens of thousands of staff.

It can be a contentious proposal to suggest that all of these ground level staff, many of which go into elections with no prior experience, are held to a perfect standard when the margin of a race is this close. There’s also all sorts of eligibility errors made in every election, like voting in the wrong electoral district because electors didn’t understand the boundaries.

There are procedural improvements that could be explored to reduce error rates involved in the final count (namely tabulators, something Elections Canada is particularly cautious about), but errors made by the elector are almost impossible to eliminate without draconian address verification. Some EPB (electionic poll book) systems used by provincial electoral management bodies have functionality that can geographically verify an address to be within a particular boundary, but that technology is hardly trivial to introduce given the myriad of problems involved in ensuring addresses are formatted correctly. Canada actually has a huge problem with addressing authorities, Elections Canada, and Canada Post having conflicting information, and it is typically only the addressing authority that has mapped out everything in GIS. These ambiguity issues makes autonomous validation prone to errors of its own.

For federal elections, which operate at such unimaginable scale where individual registrations cannot be manually scrutinized by ECHQ personnel, there’s no good solution here. And that is only one class of error that could be made at a polling place. I believe that if we are to see more elections overturned due to legal challenges of this variety, elections could get more expensive (and even less accessible) as officials are forced to deal with these problems.

I do not disagree with the decision in the aforementioned Supreme Court case, as I understand being resigned to a “coin-toss election” is not a great reflection of our democratic principles. But I do not know what the mitigation to these legal challenges on our end will look like, and I do not envy those at Elections Canada preparing for future challenges.