Canada's electoral history from 1867 to today

What problem does PoliCan solve?

by Maurice Y. Michaud (he/him)

Frustration!Those who know me well also know how easily I can succumb to the idea of diving into a seemingly impossible big project. As I often joke, "Why make simple what can be made complicated?" Closer to the truth, though, is that I got absorbed in reading about more than 155 years of Canadian electoral history. In so doing, I discovered that how our electoral system works now is not how it has always worked. I realized how serious the problem of declining voter turnout is, and how it is getting worse. I met some fascinating political figures I had never heard about, from a mass murderer to people out on their luck, or from people dying while in office or in unusual circumstances, to others who are or were extremely unsuccessful perennial candidates. I had to concede that some politicians I considered ignoble had in fact several redeeming qualities, while others I revered had nasty flaws I didn't suspect. And you know what? Switching party loyalties is not a phenomenon invented just yesterday.

There are two official sources for each province and territory, plus two at the federal level: the independent non-partisan agency that runs elections in the jurisdiction, and the jurisdiction's legislative library — not to mention that, today, there are more unofficial yet reliable online sources than there were when I started this project. The closest thing that existed prior to PoliCan was the Canadian Elections Database, put together under the direction of Dr. Anthony Sayers at the University of Calgary's Department of Political Science, but it does not seem to have been updated since 2020. What's more, and with respect for the huge amount of work that was done there, it does not include by-elections and referendums, it has several omissions and inaccuracies, and it does not offer as many ways as PoliCan to explore or aggregate the data.

Over time, PoliCan has virtually turned into an encyclopedia because I had become frustrated with the need to consult so many sources presenting similar data but each in their own way. Those sources ranged from online databases with their own organization and navigational schemes (that are often clunky and assume either some prior knowledge of the dataset, or that one is only interested in their own riding), to PDF documents bringing information together in every way imaginable, as well as those good old-fashioned hard copies obtained from legislative libraries.

So I ask you... Who exactly is the one who's making things complicated? Currently, when one considers all the jurisdictions in Canada since Con­fe­de­ra­tion, there are many nooks and crannies where the results of an election are hidden. For some jurisdictions, to get the information you are seeking, you have to know the year of the election and it better not be too far back in time! For others, you have to know the riding that interests you. And then, rarely can you easily link results with people; you have to do more searching on other websites to know more about the people. But with PoliCan, results, people, ridings, and political parties are all easily interconnected, and that information is searchable in a simple, intuitive manner — thereby only requiring that you have a high interest in seeking the information, but little to no prior knowledge of what that information will be. In fact, because of all the cross-referencing I have done to make those connections, PoliCan is often more accurate than some official published sources.

What's more, for those of you who, like me, are interested in one day seeing TRUE electoral reform in Canada, you have the results of 408 general elections that can be run through a simulator to see how those results might have looked like had a proportional representation system been in place. Although the unfairness of our current electoral system has received considerable air time in recent years, the data organized as they are in PoliCan help at proving that the results of general elections in Canada have been distorted ...since 1867!

I did not start this project with such high ambition. I only wanted a realistic simulator for alternate outcomes if the electoral system had been different. I gradually went from entering the results of a few selected elections, to all general elections federally and in selected provinces since Canada's centennial year, to every single electoral event since Canadian Confederation in 1867. Through this process, I started wanting to know more about all these actors involved in our democratic process, and thus built the means to keep that information accessible for future reference. While, on the back end, the data is organized by jurisdiction and event, they can now be uncovered through a variety of points of entry. If many years ago your great-grandfather Murdoch ran in a federal or provincial election (not sure which level), just knowing his name should be enough to find him.

What's included and what's not

PoliCan does not include any results at the municipal or county level, so time spent serving at that level is NOT tabulated. Fur­ther­more, the time individuals spent in the Canadian Senate (or a province's equivalent when it existed) is also excluded because that time is NOT in elected office. It thus follows that, if someone died while in service at either these levels, they are NOT marked as having died in office in PoliCan. However, PoliCan will eventually provide means to easily identify and search for profiled individuals who have served at either those levels.

If you are wondering what is the size of this database in technical terms, this short table sums it up.
Tables 22
Records 482,935
Result Records 134,071
Data Points 7,318,982
Weight 39.06 MB
Acronym denoting the software bundle that makes PoliCan possible: L for the Linux operating system, A for the Apache client server, M for the MySQL relational database software, and P for the open-source PHP scripting language.

Database (relational)
A collection of "data items" with relationships between them, where these items are organized in a set of tables. Think of a table as a spreadsheet: the intersections of the rows and columns form cells, and the cells in each column contain similar items. Two tables can then be joined during a database query if they have a column of items in common. Example: In the table listing all the elections, one row describes a specific by-election and one of the cells in that row indicates the number of the jurisdiction where the by-election was held ( J ). In another table listing the jurisdictions, the row numbered J describes that jurisdiction. Since the two tables can be joined on that common item, the row in the first table contains only items specific to that by-election and none about the jurisdiction.

As defined in "Database (relational)," a table is the equivalent of a spreadsheet designed to receive a set of related data items.

For all practical purposes, a record is one row in a table.

Result Records
A record found specifically in the results table to take note of the number of votes a candidate received, the riding in which they ran, their political allegiance, if they won that race, etc. If a candidate ran seven times, they have a total of seven result records in this database.

Data Points
For all practical purposes, a data point is the content of one cell — in other words, at the intersection of a row and a column.

The number of megabytes that the data occupy. It should be noted that all these data are non-formatted plain text and thus very light on their own. This weight does not include that of the files developed to maintain and present the data.

The electoral results — that is, the actual number of votes for each race — are in the public domain and have been found from a multitude of public sources. However, supplemental information on profiled individuals is taken from the following online sources.

Accomplished and to be accomplished

It may seem astonishing to you that one person, alone and without pay, could have managed to cobble together all these data, and in two languages at that! Truth be told, it has taken me years in spurts and bursts, for I do have a full-time job aside from this project. That being said, the restrictions on activities imposed by the COVID-19 pandemic have certainly helped me push this project over the finish line — except it will never really be finished as long as Canada and its provinces and territories continue to have democratic elections. And while I'm optimistic by nature, what jumps out at me are the data that are missing.

For instance, it is often easier to find the birth and death dates of people elected 50 years ago or more than it is for those elected in the past decade or so. Given that the majority of the latter are still alive, I figure that it might just be a simple matter of contacting them (or people who know them) even though they are legion! On the other hand, there are those who got elected many years ago for one or two short terms in a legislature but who clearly must have been quiet backbenchers, for practically nothing seems to be known about them today. And, alas, along with life comes death: every day I check on those surviving former parliamentarians who are supposedly celebrating their birthday on that day to see if indeed they made it to another year. Even with that, during another routine check in early March 2022, I found that the sources on a member of parliament I thought was still alive had recently been updated to indicate that he had in fact been dead for more than 25 years! And it happened again in August 2023, when I found that two MLAs — one from British Columbia and another from New Brunswick — had in fact been dead since 1996.

That is why I will soon be inaugurating a members' area to the website — The Friends of PoliCan — to help me research those "elusives," to signal any mistake in the existing data, and to be on the lookout for sudden resignations or any changes of allegiance of sitting parliamentarians, or the passing of former ones. After all, I think I can safely say that I have already done (and will continue to do) more than my fair share, so I certainly wouldn't refuse some help at this point to make this resource even more complete and accurate. Plus I am already thinking of a succession plan should I no longer be able to maintain this site.

So, while my opening statement on the homepage of PoliCan may have come across as grandiloquent, you only need to start exploring this site to realize that it is more a simple statement of fact. A good place to start to whet your appetite might be the "Historic, unusual, or surprising" page in the People section, and following a few of the links presented on it...

© 2019, 2023 :: (Maurice Y. Michaud)
Pub.: 23 Dec 2021 00:01
Rev.:  5 Nov 2023 12:18 (but data presented dynamically)