Introduction

This project examines the extent to which Canadian governments reflect the will of the voters and the impact that abstaining voters could have if they choose to vote in future elections. I explore this in the context of Canada’s current party system and first-past-the-post (FPTP) electoral system, with comparison to how the House of Commons would have looked under a proportional representation (PR) electoral system.

The goal of this project is to illustrate how the entire electorate has been voting (or not voting) over the past two decades, that future election outcomes are not a foregone conclusion, and that abstaining eligible voters could have a significant impact on the results of future elections if they choose to vote.

Political Context

Canada uses a first-past-the-post (FPTP) electoral system, and is divided into 338 ridings, each with a similar population. Our federal elections are essentially a set of local elections in each riding (Carty and Young, 2012). In each riding, the candidate who receives the most votes becomes the Member of Parliament (MP). The party with the most MPs – or seats in the House of Commons – forms the government. The leader of that party becomes the Prime Minister. If that party secures more than fifty percent of the seats in parliament, it will form a majority government. In this first-past-the-post system, fifty percent of the overall national vote, often referred to as the popular vote, is not required to form a majority government. In fact, the two most recent majority governments, formed by former Prime Minister Stephen Harper and the Conservative Party in 2011, and Prime Minister Justin Trudeau and the Liberal Party in 2015, saw each respective party winning less than forty percent of the national popular vote. Arguments in favor of a FPTP system include its ability to produce a clear winner that can govern more effectively (Cairns, 1968). Canada’s FPTP system can produce majority governments, which some might argue leads to more stability and longer-lasting governments that can implement policies effectively. On the contrary, arguments in favor of PR include the artificial majorities that a FPTP system can produce and the idea of representation. Cairns (1968) defines an artificial majority as a governing party that secures a majority government despite winning less than 50% of the popular vote. PR would ensure that the seats in the House of Commons reflect the proportion of the popular vote that each party received

Political parties have always been at the center of Canadian elections and voting. Party affiliation is widely accepted as the single most important factor that Canadians consider when deciding which candidate to vote for (Cross, 2016). Furthermore, party leaders are the stars of Canadian politics and are often evaluated by voters as Prime Ministerial candidates. The only parties who have ever formed government are the Liberals and the Conservatives, however, Canada has three other parties; the New Democratic Party (NDP), the Green Party, and the Bloc Quebecois (BQ).

Voter turnout in Canadian elections, especially amongst young voters, sharply declined after 1988. Bittner, Johnston, and Matthews (2007) cite the decline in electoral competitiveness – specifically the back-to-back Progressive Conservative majority governments led by Prime Minister Brian Mulroney in 1984 and 1988, and the three consecutive Liberal majority governments lead by Prime Minister Jean Chretien in 1993, 1997, and 2000 – as the cause for this abrupt decline in the 1990s. They point to the return of competitiveness in Canadian elections as the reason why the turnout decline plateaued in 2004, and has remained relatively stable since. There have not been any major changes in Canada’s party system since the Progressive Conservative and Alliance Parties merged to form the Conservative Party of Canada in 2003. For those reasons, this project primarily focuses on data from Canadian general elections since 2004.

Project Outline

To tell a story about popular vote share, parliamentary representation, and the impact that abstaining voters can have at the constituency and national level in future elections if they choose to vote, I will present visualizations that illustrate my data analysis and compare:

• Voter turnout rates in Canadian general elections over the past three decades.

• The percentage of votes that each party has received among people did vote versus among all eligible voters, including those who didn’t vote since 2004.

• The percentage of the popular vote versus share of seats in parliament since 2004.

• The number of ridings in which a plurality of eligible voters voted for the elected MP versus the number of ridings in which a plurality of eligible voters abstained in the 44th Canadian General Election.

Importance

Voting is the simplest, easiest, and most common form of political participation for citizens of a democracy. But in recent elections, young people have had a substantially lower voter turnout rate than previous generations did when they were our age (Bittner, Matthews, and Johnston, 2007). The more citizens who vote, the better the government can reflect the will of the people and be accountable the electorate. The government cannot reflect the will of or be accountable to eligible voters do not participate in this essential democratic process.

• The Health of our Democracy

High voter turnout is a key indicator of a healthy democracy (Blais, Dassonneville, and Kostelka, 2020). It indicates high engagement among citizens, allows the will of the electorate to be represented in the government and in policy decisions, and it maintains the government’s democratic character by making it consistently accountable to the electorate. In our democratic system, the government is chosen by the people and is accountable to the people. Voting is the simplest and most direct way for us to participate in politics. It allows us to choose who represents us, to give the government a clear direction, and to hold the government to account every few years.

• Our Civic Duty as Canadians

Voting is our civic duty. Canadian citizens ages 18 and older have the right to vote. We get a say in who represents us on the in the place where decisions about the present and future of Canada are being made – that is the beauty of Canada’s democracy. There are many people around the world who are fighting for this basic right. Over the past century, Canadians have also fought for their right to vote. In fact, not all Canadian citizens ages 18 and older have had this right throughout Canada’s history. As a young woman, I feel like it is my duty and responsibility to honor the women who fought to get women into the voting booth and to allow women to run for and be represented in the House of Commons just over a century ago during the suffrage. It is our responsibility to not take this right for granted and to participate in the process of choosing our government and holding our government to account.

• The Principle of Representation

High voter turnout is essential for equitable and equal reflection of citizens’ interests in public policy. There is a vicious cycle between low youth voter turnout and the under representation of young people in parliament and their interests, such as housing affordability and climate change, in public policy. While there is no evidence that voter turnout has any systemic effect on Canadian election results, it does have a substantial effect on policy. Politicians want to be re-elected and they know and care which demographics have high turnout rates and which ones do not. Therefore, they will prioritize the interests and policy preferences of citizens that they think are more likely to vote in the next election. Low voter turnout also exacerbates systemic inequalities; when overall voter turnout is low, there is a wider turnout gap between marginalized and non-margialized demographics (Blais, Dasonneville, and Kostelka, 2020).

Data Overview

Data Collection

The data collection for this project started with Elections Canada election data. The Elections Canada website has official records of voting and turnout available poll-by-poll, at the riding level, at the provincial level, and at the national level. General elections summary table data is only available on the Elections Canada Past Elections data page in CSV format from 2004-2021. I gathered data about voter turnout, vote share by party, and seat share by party at the constituency, provincial, and national levels from past elections data and official voting results. There was very limited summary results data available for the 1993, 1997, and 2000 general elections. In 2003, the Progressive Conservative and Canadian Alliance Parties merged to form the Conservative Party of Canada. For those two reasons, this project primarily focuses on data from Canadian general elections since 2004. For the 38th-44th Canadian General Elections inclusive (2004-2021), I the Official Voting Results (raw data) Summary Tables. The summary tables that I used for each of those general elections are as follows: Voter Turnout (Table 2), Distribution of Seats by Political Affiliation and Sex (Table 7), Number of valid votes by political affiliation (Table 8), Percentage of valid votes by political affiliation (Table 9), Voting results by electoral district (Table 11), and List of candidates by electoral district and individual result (Table 12).

Reliability

Elections Canada is an official, non-partisan, and trustworthy source. It is a regulated agency responsible for carrying out Canadian Federal Elections, and there is no reason to believe that anything unethical happened in the collection of this data. All of the data that I used for this project is available through both Elections Canada open data and Government of Canada records. For these reasons, we can be confident that this data is reliable and trustworthy. The website is monitored by the Office of the Chief Electoral Officer, and all information and data marked as archived is available on the web for research purposes and has not been updated since the date it was archived. Elections Canada outlines how results are validated and processed after polling night.

https://www.elections.ca/content.aspx?section=res&dir=pub/ecdocs/rom/vIII/ch_2&document=ch_2&lang=e

Ethical Considerations

Data Collection

As outlined above, the data used in this project was collected from Elections Canada, which is a reliable and trustworthy source. Ethical considerations include voter anonymity and privacy, transparency and openness, and research intent. Elections Canada keeps public records of how many ballots were cast, where those ballots were cast, and the candidate (and respective party affiliation) that was marked on each ballot. These voting records are anonymous and private; Elections Canada keeps records of election results but ensures that the votes remain anonymous and private, so no vote tallied in the publicly available records can be traced back to an individual person. Elections Canada is transparent in how votes are recorded and how the open data for researchers and summary tables are put together. Furthermore, elections Canada publishes official voting results data in a non-partisan way for public record and research intent. For these reasons, I have full confidence that the collection of this data was accurate and ethical.

Data Cleaning

I used data cleaning to make the large data files that I downloaded from the Elections Canada website usable for my research and data visualizations. My data cleaning primarily consisted of extracting columns from the tables that are relevant to vote totals, turnout rates, seat share, and vote share at the constituency and national level. Ethical concerns of data cleaning might include data manipulation, computational or coding errors, excluding important data to draw a conclusion that supports a particular argument, and transparency. To ensure that my data cleaning was done in an ethical manner, I compared my SQL tables to the original tables to ensure that all the information was accurate and included, tested my code to ensure that it was computing metrics such as turnout rate and proportions of the overall electorate who voted for each party’s candidate and did not vote accurately, and ensured that complete data from all 338 ridings in the 2021 election was included.

To ensure the transparency of my data cleaning, I have listed some of the steps I took below:

• I used SQL (Strucured Query Language) to efficiently extract the columns I needed and to join data tables for use in my data visualizations. SQL and the relational database model, which I have studied, help to ensure that the data tables are reliable, efficient, and have minimal or no redundancies. Many of the tables included data from just one election or from a few of the elections, with some overlap between which elections were included in the tables.

• Aggregation, grouping, and row operations in SQL on the raw data to calculate vote totals, vote share, and turnout at the constituency level for the 2021 election and at the national level for every election included in this project.

• Including data from every province, constituency, and poll where relevant: I ensured that all available data was used and no outiers were excluded.

• Omitted data about demographics, riding locations and geography, etc. but included all vote totals, numbers of eligible voters, numbers of votes that went to each candidate relevant columns only were included.

Analytical Strategy

I analyzed the raw turnout, national popular vote, and seat share data provided by Elections Canada to tell a story about voter turnout, popular vote share, and parliamentary representation in Canada since 2004 through data visualizations. Certain aspects of my project, including finding the percentages of the entire electorate who voted for each party and didn’t vote in each election required further data analysis and computations. By analyzing and illustrating how many eligible voters have been voting for the party that formed government versus abstaining, the percentages of eligible voters who actually vote for each party, and the number of ridings in which a plurality of eligible voters actually did not vote, we can gain valuable insights into the impact that the abstaining voters can have in future elections if they decide to vote.

These are some of the steps that I took to create my visualizations that include data about the percentage of eligible voters who didn’t vote and how many ridings saw a plurality of eligible voters abstain in the most recent election:

• Proportions of the popular vote were calculated directly from the Elections Canada official voting results data. It was shown as the total number of votes that each party received in each province in each election in separate tables, so I took the sum of the votes that each party received across all provinces divided by the total number of votes cast for all parties across all provinces to get

• The percentage of eligible voters who did not vote was calculated as (100 - turnout percent)%, and each party’s share of the popular vote was multiplied by the turnout proportion provided in the turnout data tables (the number of ballots cast / the number of eligible voters).

• To illustrate how the 44th Parliament would have looked if Canada used a proportional representation system, I calculated the number of seats that each party would have won by multiplying their proportion of the popular vote by 338. I then used the GGParliament package to plot it.

• To illustrate what 44th Parliament would have looked under Canada’s current first-past-the-post system if the Did Not Vote Party existed, I downloaded the 2021 election results which included how many votes each candidate received, the size of the electorate, and turnout rate for each riding. To calculate how many seats the imaginary ‘Did Not Vote’ party would have won, I wrote code to count how many rows had more eligible voters (total electorate size*(1-turnout proportion)) abstaining than voting for the elected Member of Parliament. I also wrote code to find out how many of these seats was actually won by each party to get a final calculation of how the House of Commons would’ve looked if the Did Not Vote Party existed under FPTP.

• To illustrate how the 44th Parliament would have looked if the Did Not Vote Party existed and if Canada used a proportional representation system, I multiplied 338 by the proportion of the electorate that voted for each respective party, other, and did not vote to get the seat shares in this scenario and plotted that using the GGParliament package.

Visualizations

The visualizations presented below cover a range of topics related to voter turnout, election results, and representation in Canadian parliament. Each visualization includes a brief description and an an interpretation of what it can tell us about one or more of these topics. Some of the visualizations included are interactive - you can hover over the certain parts of the graph such as data points and bars to see more details of the data that is presented.

Turnout

Voter turnout in Canadian elections, especially amongst young voters, sharply declined after 1988. Bittner, Johnston, and Matthews (2007) cite the decline in electoral competitiveness – specifically the back-to-back Progressive Conservative majority governments led by Prime Minister Brian Mulroney in 1984 and 1988, and the three consecutive Liberal majority governments lead by Prime Minister Jean Chretien in 1993, 1997, and 2000 – as the cause for this abrupt decline in the 1990s. They point to the return of competitiveness in Canadian elections as the reason why the turnout decline plateaued in 2004, and has remained relatively stable since.

The visualization below shows turnout in Canadian General Elections from 1993-2021. The visualizations that follow will focus on 2004-2021, as those are the elections that have taken place in Canada’s current party system.

Starting in 1993, until 2000, we can see that turnout continued to decline. This is a continuation of the post-1988 decline in voter turnout. Between 2000 and 2004, we can see that there was a very slight decrease in voter turnout. Since 2004, however, turnout has only fluctuated. There is a less than two percentage point difference between the 2004 and 2021 voter turnout rates, which aligns with what Bittner, Johnston, and Matthews (2007) suggested. With the exception of 2015, at least one-third of eligible voters have abstained from each election since the turnout decline plateaued in 2004. In 2008, more than 40% of eligible voters abstained. Since 2004, Canada has only seen two majority governments: a Conservative majority led by Stephen Harper in 2011 and a Liberal majority led by Justin Trudeau in 2015. However, further research, beyond the scope of this project is required to explain voter turnout trends.

Tracking the Vote Choices (or lack thereof) of Eligible Canadian Voters

Similarly to the visualization below, the election results presented in the media usually only offer insights into how individuals who voted marked their ballots. While this perspective provides an accurate representation of how individuals who participated in the democratic process voted, it does not fully capture the entire electorate or show these results relative to the proportion of eligible voters who did not vote.

• In every Canadian general election in the twenty-first century, with the exception of the 42nd Canadian General Election in 2015, at least one-third of eligible voters have abstained from the democratic process.

• In most cases, with the exceptions of the 43rd and 44th Canadian General Elections in 2019 and 2021, respectively, the party that wins the popular vote, defined as having the most votes for its candidates cumulatively, across the country, has formed government. In 2019 and 2021, the Conservative Party won the popular vote, but the Liberal Party won the most seats (but not a majority), and Prime Minister Trudeau went on to form minority governments.

Starting in 2004, we can see that in each of these elections, except for 2011, the Liberals and Conservatives had the two highest proportions of the popular vote. In 2011, the Liberal Party became a third-party and the New Democratic Party formed the official oppoition for the first time in Canada’s history.

However, there are a few things to take note of before we get to the next visualizations:

• No party won more than 50% (a simple majority) of the popular vote in any of these elections.

• The New Democratic Party (NDP) won more than double the share of the popular vote that the Bloc Quebecois (BQ) won.

• The Green Party won a slightly higher share of the popular vote in 2008 than it did in 2019.

• The majority of Canadians who voted in each of these elections did not vote for the Prime Minister’s party; in each of these elections the party that formed government won significantly less than 50% of the popular vote, therefore, the majority of Canadian voters mark their ballots for an opposition party.

The visualization presented above is simple: its sole purpose is to illustrate the percentages of votes across the country that went to each party in every election since 2004.

But this snapshot doesn’t capture the entire picture. The approximately one-third of eligible voters who didn’t vote are not included. This raises a couple key questions:

• What percentage of eligible voters across Canada actually voted for each party?

• What percentage of eligible voters have actually voted for the party that formed government in recent elections?

From this visualization, it becomes clear that when we account for how the entire Canadian electorate voted (or didn’t vote), the simple plurality of eligible voters have not been voting for either of the two parties that have ever formed government. In fact, a plurality of eligible voters have simply not voted in every single Canadian general election included in this project.

Starting with the 38th General Election (2004), we can see that in every election since then, the percentage of eligible voters who didn’t vote has been at least ten points higher than the percentage of eligible voters who voted for the popular vote winner (not necessarily the Prime Minister’s Party), with the exception of 2015. The 43rd General Election (2015) had the highest turnout rate in the period between 2004 and 2021. This election, resulting in a Liberal majority government, also had the highest percentage of the electorate voting for a party during this period. While national popular vote does not determine who forms government in Canada’s FPTP electoral system, it is clear that recent elections have consistently seen a significant plurality of eligible voters abstaining from essential democratic process.

The House of Commons and the will of the Electorate

How well does the government represent what the voters want? The next few visualizations will illustrate how winning votes across the country has translated into parliamentary representation for each party from 2004-2021. These next three visualizations will analyze the results of Canada’s first-past-the-post electoral system from 2004-2021, only taking into account how the eligible voters who actually participated in the democratic process voted to explore how parliamentary representation compares to popular vote.

The visualization below simply shows the percent of seats that each party has had in the 38th-44th Canadian parliaments. Due to the nature of Canada’s electoral system, the seat shares shown below are not necessarily reflective of the national popular vote breakdowns shown in previous visualizations.

Starting in 2004, we can see that the Liberals and Conservatives tend to be rewarded with a higher share of seats than votes. The have only been two majority governments elected during this time: a Conservative majority in 2011 and a Liberal majority in 2015. It is a very simple illustration: its sole purpose is to directly illustrate the share of seats that each party has had in the House of Commons between 2004 and present.

The visualization below compares the percentage of the popular vote to the percentage of seats that each party won in Canadian federal elections from 2004-2021. You can hover over each point to see which party and election it represents, and that party’s share of the popular vote in comparison to its share of seats in parliament.

Starting at the lower leftmost corner of the graph, we can see that the cluster of green and gray dots (representing the Green Party and other parties/no party affiliation) stays pretty flat as the percentage of popular vote increases. Focusing on the Green Party, this visualization shows that increases in the Green Party’s national popular vote share do translate into more seats in the House of Commons. Looking at the teal dots, representing the Bloc Quebecois, we can see that they are all gathered around the same part of the graph, and they appear to form a steep line. This shows that increases in the BQ’s national popular vote typically lead to increases in its seat share in parliament. Furthermore, their seat share in parliament has often been higher than the share of the popular vote that they received. This makes sense, as they are a small regional party, that only exists in Quebec. There have been instances where a very similar popular vote share has resulted in significantly more seats for the NDP than for the BQ.

Hovering over the top of the graph, we can see that the majority governments won by the Conservatives and the Liberals in 2011 and 2015, respectively, fit the Cairns (1968) definition of artificial majorities. In 2011, the Conservative Party won approximately 39.63% of the national popular vote and a parliamentary majority with approximately 54.5% of the seats. In 2015, the Liberal Party won approximately 39.47% of the national popular vote and a parliamentary majority with approximately 54.4% if the seats. These examples show the Harper-led Conservatives and the Trudeau-led Liberals winning a similar vote share of below 40% and around 54% of the seats in parliament in two separate elections.

The visualization below compares the popular vote share to the seat share of the government party following the 2004-2021 general elections. Its sole purpose is to examine the extent to which Canada’s FPTP electoral system rewards the government party with a share of seats that is substantially higher than the share of the popular vote that they received. Hover over each bar to see the exact percent of the popular vote and percent of seats in parliament that the government party received in each of these elections.

Cairns (1968) pointed out that Canada’s FPTP electoral system has a tendency of rewarding the government party with a greater percentage of seats in parliament than of the popular vote. These visualizations clearly show that that has held true, even with minority governments. We can see that this gap was smaller following the 2004, 2006, and 2008 elections. In each of these instances a minority government was elected and the government party won the popular vote. We see this gap widen in 2011 and 2015, when the Conservatives and Liberals each won a majority government. However, the gap remains wider than the 2004-2008 elections in 2019 and 2021, even though both of these elections resulted in minority governments. It is important to once again note that the government (Liberal party) had a slightly lower share of the national popular vote than the official opposition (Conservative party) in the 2019 and 2021 elections. Overall, this visualization shows that the FPTP system does tend to reward the government party with a percentage of seats in parliament that is higher than the percentage of the popular vote that they received and that gap was smallest when there was a minority government that also won a plurality of the national popular vote.

Examining the Popularity of Canadian Prime Ministers (and governments) Amongst Eligible Voters

These visualizations compare the percentage of eligible voters who actually voted for the Prime Minister’s party to the percentage of eligible voters who didn’t vote and the percentage of eligible voters who voted for any of the opposition parties. In this case, the entire electorate has been taken into account, and popular vote percentages have been adjusted to account for and include the eligible voters who didn’t vote.

Hover over each bar to see the exact percentages.

This visualization clearly shows that when we examine how the entire electorate voted (or didn’t vote), there has been a significantly higher share of eligible voters not voting than eligible voters voting for the party that formed government in every election from 2004-2021. To put that into perspective, in every election in Canada’s current political party system, the percentage of eligible voters who voted for the Prime Minister’s party has trailed the percentage of eligible voters who didn’t vote at all. This gap was the largest in 2008, at around 19 percentage points; approximately 21.15% of eligible voters across the country voted for the Conservative Party, and approximately 41.17% did not vote at all. This gap is the smallest in the 2015 election at around 5 percentage points; approximately 26.949% of eligible voters across the country voted for the Liberal Party, and approximately 31.72% did not vote at all. This gap tends to be smaller when overall voter turnout is higher.

The visualization below puts eligible voters into three categories: didn’t vote, voted for the Prime Minister’s party, and voted for an opposition party. It is important to note that in Canada’s multi-party system, there is always more than one opposition party, with the party with the second highest share of seats in parliament forming the Official Opposition. It is also important to note that the opposition parties are not necessarily similar in terms of ideology or policy stances. The purpose of this next visualization is to simply compare the percentages of eligible voters who have abstained, voted for the government party, and voted for a party that did not form government in every Canadian election from 2004-2021. Hover over each bar to see the exact percentages.

This visualization clearly shows that in every election from 2004-2021, the percentage of eligible voters who voted for the government party has trailed both the percentage of eligible voters who voted for one of the opposition parties and the percentage of eligible voters who did not vote at all. We can see that because at least one-third of the electorate usually abstains from voting, the presence of multiple parties, and the nature of the FPTP electoral system, political parties have only needed around 20-25% of eligible voters across the country to vote for them in order to form government in recent elections.

The 2021 Election & the 44th Canadian Parliament

The most recent Canadian general election took place on September 20, 2021. It had the lowest turnout rate since 2011 and turnout declined 5.7 points between the 2015 and 2021 general elections. However, it’s important to note that the turnout rate of 62.6% in 2021 is comparable to the other elections have taken place since 2004 and it is possible that the slight drop between 2019 and 2021 might be due to the Covid-19 pandemic. The visualizations in this section will show how our current parliament would have looked if Canada used a first-past-the-post system and the number of ridings in which a plurality of eligible voters abstained.

The visualization below illustrates the share of seats that each party has in Canada’s current parliament (since 2021).

As shown above, the Liberal Party, led by Prime Minister Justin Trudeau, has a minority government with 160 seats, and the Conservative Party is the official opposition with 119 seats. The Bloc Quebecois, New Democratic Party, and Green Party won 32 seats, 25 seats, and 2 seats, respectively. The number of seats that each party has in parliament is reflective of the number of ridings in which its candidate won a plurality of votes; for example, the Liberal candidate won the most votes in 160 ridings, and the Conservative candidate won the most votes in 119 ridings.

In comparison, the Conservative party would have had a minority government (significantly smaller than the one that the Liberals currently have) if the 2021 election had been conducted under a proportional representation system. As previously shown Conservative Party got a slightly higher share of the popular vote than the Liberals did, and the New Democratic Party won around double the share of the popular vote that the Bloc Quebecois won.

In this particular example, the Conservatives would have had a small minority government of only 114 seats. This raises questions about whether a minority government that small would be stable, have the ability to govern and implement policies, or if coalition governments, which have never existed in Canada before, would emerge in a PR system. Regardless of one’s stance on FPTP vs PR, it is clear that if Canada used a proportional representation system, parliament, and possibly election and political dynamics, would be very different.

Other notable differences between Canada’s actual 44th Parliament and how it would have looked if Canada used a proportional representation is the significant changes in the number of seats that the NDP and BQ would have had. Under FPTP, the NDP won 17.8% of the popular vote and 7.4% of seats in the House of Commons. The BQ won 7.6% of te popular vote and 9.5% of the seats in the House of Commons. Despite having more than double the votes that the BQ had nationally, the NDP won fewer seats in the House of Commons. Similarly to the NDP, the BQ is a smaller party. However, unlike the NDP, which is a national party, the BQ is a regional party, so all of its support is concentrated in Quebec ridings. This key difference is what allows the BQ to win a pluraity of votes in more ridings than the NDP, which translates into more seats in the FPTP system.

Ridings with a plurality of eligible voters abstaining

These next visualizations will illustrate the effects of low turnout at the constituency level in the most recent federal election. If, hypothetically, the votes (or lack thereof) of the entire electorate were considered, and the votes of eligible voters who did not vote were transferred to the imaginary “Did Not Vote Party,” who runs mystery candidates in each riding across Canada, how would the House of Commons look?

Under Canada’s first-past-the-post electoral system, the candidate with the most votes in each riding is elected as the Member of Parliament. This visualization shows how many ridings across Canada saw a plurality of eligible voters abstain in the 2021 general election.

In the 2021 election, the Did Not Vote Party would have won a super majority of 289 out of 338 seats in the House of Commons. That would have been a larger majority government than those of Justin Trudeau in 2015, Stephen Harper in 2011, Jean Chretien in the 1990s, Brian Mulroney in the 1980s, and any other majority government in Canadian history. This example is exaggerated; the Did Not Vote Party is not real, and this entire scenario is unrealistic. But, to put that into perspective, in 289 out of 338 ridings across Canada, there were more eligible voters who abstained than eligible voters who voted for the Member of Parliament representing them in the 44th parliament. This hypothetical scenario would have produced a massive majority, as only 37.40% of eligible voters abstained in the 2021 election, but the Did Not Vote Party would have won over 85% of the seats in parliament. The few seats that the Liberal Party, Conservative Party, Green Party, and Bloc Quebecois still have in this scenario represent the ridings where there were more eligible voters who voted for the elected Member of Parliament than eligible voters who didn’t vote.

This is an extremely exaggerated and imaginary example of a first-past-the-post system producing an majority if its vote is spread out efficiently enough. majorities often used to argue in favor of a proportional representation system, so I’ve included a visual representation of what the 44th Canadian Parliament would have looked like in a proportional represenation system, where the imaginary Did Not Vote Party existed.

A plurality of eligible voters across Canada did not cast a ballot in the 44th General Election, so in this proportional representation scenario, the Did Not Vote Party would still have had the most seats in the House of Commons, but not a majority. Roughly 37% of eligible voters abstained in 2021, so the Did Not Vote Party would have had 126 out of 338 seats in the House of Commons.

Conclusions

Some interesting conclusions be drawn from the data visualizations presented:
• Low voter turnout is bad for representation.

I initially assumed that left-leaning parties, such as the Liberals and the NDP would perform better when turnout was higher. However, existing literature and my data analysis and visualizations show no evidence of that. When we view representation in government as the extent to which the government reflects the will of the electorate and the extent to which the electorate can hold the government to account, we see that this decreases with turnout as well. While we observed that the government party is consistently over-represented in parliament relative to the share of the popular vote it received and that the percentage of eligible voters who voted for the government party has trailed the percentage of eligible voters who voted for an opposition party or didn’t vote, these gaps become wider when voter turnout is lower. High voter turnout is not beneficial to any particular political party. It is beneficial to our democracy as a whole. An engaged electorate that consistently participates in the democratic process indicates that the government is chosen by as many eligible voters as possible, representative of the will of the voters, and can be held to account regularly.

• Abstaining voters can have a significant impact in future elections if they choose to vote.

Lastly, the data analyzed and visualizations presented show that the percentage of eligible voters abstaining has significantly outnumbered the percentage of eligible voters who voted for the government party in recent Canadian federal elections. The same question has lingered every time: what would have been different if more eligible voters actually voted? But as Carty and Young (2012) put it, federal elections are actually contested at the local level in each of Canada’s 338 ridings. This data shows that in the most recent election, a plurality of eligible voters abstained in 289 out of 338 ridings. Increasing citizen engagement in this important democratic process quite literally starts at home. Approximately 85% of current Members of Parliament had fewer eligible constituents voting for them than eligible constituents who simply chose not to vote.

References

Bittner, A. (2018). Leaders always mattered: The persistence of personality in Canada. Electoral Studies 54. 297-302. www.elsevier.com/locate/electstud

Bittner, A., Johnston, R, and Matthews, J.S. (2007). Turnout and the Party System in Canada, 1988-2004. Electoral Studies, 26(4) 734-745.

Blais, A., Dassonneville, R., and Kostelka, F. (2020). Political Equality and Turnout. Oxford Handbook of Political Representation in Liberal Democracies. 396-412. https://doi.org/10.1093/oxfordhb/9780198825081.013.20

Cairns, A. C. (1968). The Electoral System and the Party System in Canada, 1921-1965. Canadian Journal of Political Science / Revue Canadienne de Science Politique, 1(1), 55–80. http://www.jstor.org/stable/3231695

Carty, R. K. and Young, L (2012). The Local Underpinnings of Electoral Competition in Canada, 1979-2008. Canadian Political Science Review, 6(3), 227-236.