In the 40th General Election on October 14, 2008, national voter turnout fell to an all-time low of 58.8% of registered electors. This was a decrease of 5.9 percentage points from the 39th General Election and consistent with the long-term trend. Likewise, voter turnout in the riding of Scarborough-Rouge River dropped to an unprecedented low of 47.5%, the lowest in Toronto and seventeenth lowest in the country. For the first time in the history of the riding more than 50 percent of registered electors did not vote. This paper discusses possible causes of poor voter turnout using ScarboroughRouge River as a case study. A number of suggestions are put forth to explain the decline.
The federal electoral district of Scarborough-Rouge River is an intriguing case given that it is the largest riding in population in the City of Toronto and has a growing list of registered electors Located in Toronto’s east end north of Highway 401, the riding has its own unique characteristics among ridings in the country’s large urban centres. The population is 89% visible minority, 68% immigrant, with a high youth-to-population ratio of 23.6%. It is on average younger than the national average and the South Asian and Chinese communities combined represent 61% of the total population.
The 2008 voter turnout percentage for ScarboroughRouge River was compared to each of the voter turnouts in Toronto, in Ontario and Canada. In doing so, it was found that the ridings turnout was not only lower than each of the three jurisdictions. In fact, the riding ranked lowest among each of Torontos 23 electoral districts, and 105th among Ontarios 106 electoral districts as well. Voter turnout in ScarboroughRouge River has been below both national and city turnout in each of the last seven federal elections. The data indicates that as turnout continues to drop, the gap in percentage points relative to these jurisdictions continues to widen with each successive election. To determine which polling districts within the riding displayed lowest turnouts, a poll-by-poll analysis was performed to compare individual poll turnout to the overall turnout in Toronto.
Voter Participation among Youth Aged 18-24
Canadas youth electorate, (registered voters aged 18-24 years) appears to be voting less over recent years. That young people vote at lower percentage rates than their older counterparts has been a common finding of electoral participation research and been confirmed by recent studies. Moreover, this tendency appears to hold true for virtually all established democracies in which the issue has been studied. In a survey of an equal number of voters and non-voters, it was found that turnout in the 37th General Election was only 22.4% among youth aged 18-20 and 25% among youth aged 18-24 years. In comparison to voters aged 58 and older who voted at a robust 81.5% youth turnout was 36.2 percentage points below the national turnout percentage (61.2%) in 2000. Thus, in Canadian federal elections when comparing voters under age 25 and voters over age 65, there is a gap in voter turnout of 10 to 25 percentage points.1
Why is turnout decreasing among youth? Social scientists attribute declining voter turnout among youth to the life cycle theory whereby propensity to vote was found to increase as the youth age. In fact, young voters appear to be voting less, especially when compared to previous generational cohorts. The decline may stem from a decline in civic engagement among youth. Civic engagement in this context refers to voters that, play an active role in the social and political life of their community.2 Voters are more likely to participate in elections when they are politically engaged. In other words, citizens who are active in their communities will have an interest in the political process, be knowledgeable about it and are therefore more inclined to participate at elections. Perhaps it is pertinent to ascertain whether youth play an active role in the social and political life of the riding before enquiring whether they participated in the federal election of 2008.
This generational effect is apparent in the Canadian electorate when voting behaviour in nine federal elections was observed between 1968 and 2000. It was found that though the average turnout for the six elections prior to 1990 was 74%, it dropped to 67% for the three elections between 1990 and 2000. The decrease of seven percentage points between 1990 and 2000 could be attributed to the generational effect. By examining a cohorts propensity to vote in successive generations, it was found that at the same age, turnout is 2 or 3 percentage points lower among baby-boomers (born between 1945 and 1959) than it was among pre-baby-boomers (born before 1945).3 This trend was consistent for subsequent generations. For the Generation X cohort (born during 1960s decade), voter turnout was 10 percentage points lower than the baby-boomers. For the cohort born since 1970, voter turnout was 10 percentage points lower than Generation X. With age being held constant, the propensity to vote decreases by more than 20 percentage points from the earliest to the most recent cohort. Declining voter turnout could also be attributed to generational replacement. Where post-baby-boomers made up roughly quarter (28%) of the electorate in 1988, they accounted for one half (49%) in 2000. Conversely, where the pre-baby-boomers made up 35% of the electorate in 1988, they represent only 22% in 2000. Therefore, post-baby-boomers are growing in population and voting less. Here it may be reasonably inferred that youth are not only voting less, but each successive generation of youth are voting less.
The data also indicate that older cohorts are heavily relied upon to keep voter turnout at more supportive levels. Invariably, the generational effect will decrease overall voter turnout, a trend which will likely continue. Certainly, ScarboroughRouge River is not immune from this trend, being a riding with a population three years younger than the national median.
Access or Proximity to Polling Stations
Access to polling stations refers to the actual or perceived level of convenience or ease involved in travel to the assigned polling station on election day. In the riding, access to polling stations requires three considerations: The distance voters must travel to polling stations, the availability of transportation if needed and the location or relocation (from prior federal, provincial or municipal elections) of polling stations.
We accept the following principle: The further the distance potential voters must travel to a polling station, then the less accessible the polling station. This principle may not apply to voters in rural areas where conditions vary considerably from the urban context. For instance, rural electors might be generally more willing to travel long distances to vote, have greater customary access to a vehicle and have less dependence on public transit. This principle would generally apply however to voters in large urban centres who may not have access to a vehicle, rely on public transportation or are reluctant to go out to vote during the evening darkness (more prevalent in the autumn months). Furthermore, if the distance to a polling station is too far to walk, is in an unfamiliar neighbourhood, or has a location that is not on a bus route, voters might also be less likely to vote. We believe that many of these factors could have impacted ScarboroughRouge River in 2008.
A disproportionate percentage of Rouge River constituents who are employed rely on public transportation as their primary mode of transportation to and from work, particularly women. In ScarboroughRouge River, 27.7% of the total labour force and 37% of employed women rely on public transit. On the other hand, only 11% of Canadas total labour force commutes to work by public transit. Given that a significant number of employed persons in ScarboroughRouge River rely on public transit, it is suggested that the electorate might have been dissuaded by polling stations that were: too far to access, located in unfamiliar neighbourhoods and not accessible by a nearby bus route. This could have posed an inconvenience to electors who would typically vote after working a standard daytime shift.
Access to polling stations is a factor which could also influence new immigrants and recent migrants to the riding. Migrants are movers who on census day were living in a different municipality or outside of Canada five years earlier. Both groups would be more likely to be unfamiliar with the locations of a polling station and may encounter difficulty finding its location in darkness after work. There are 15,575 immigrants since 2001 living in the riding which accounts for 8.4% of the ridings total population. Additionally, 48% of the riding occupied a different place of residence five years ago. This 48% of constituents accounts for 58,795 people 33% of whom are non-migrant movers or those who changed their place of residence within the municipality of Toronto. A number of these populations of recent immigrants and movers who are new to the riding would be less familiar with the riding in ways that would deter them from seeking or finding a polling station on Election Day, especially if the polling station was relocated and/or was outside the neighbourhood. Based on the data, from three locations in the riding, we can also suggest that the unfortunate placement of polling stations away from the residential neighbourhoods of constituents affects the likelihood that one would vote. What further exacerbates travelling a long relative distance to any outside or relocated polling station is that some constituents may not know exactly where it is located. Therefore, it is suggested that at least some of the repeat voters and newcomers to ScarboroughRouge River may not have known where to vote on election day, where polling stations were located or relocated outside the neighbourhood.
One final related consideration concerning access to polling stations in ScarboroughRouge River in the 40th General Election specifically concerns the relocation of 23 polling stations from their previous locations in the 39th General Election. The data is quite compelling given that low turnout could be attributed to reduced access to polls. For example, the polling station for poll 402, which displayed the poorest turnout (26%), was moved from a school at 380 Goldhawk Trail, to a school outside the neighbourhood at 136 Ingleton Blvd. The increased distance an elector must travel is 1.4 km. The turnout in 2006 for poll 402 was 10.8 points higher. It is believed that those unfamiliar with the new location and/or without a readily available mode of transportation did not believe that they had easy and convenient access to the polls. However, Elections Canada reports that:
During the 40th general election, Canadians had a range of opportunities to vote in various ways, in accessible venues and in more locations than ever before electors expressed high levels of satisfaction with the administration of the election. As such, administrative barriers do not appear to be a prominent factor in dissuading voting behaviour, although we note that more young electors indicated these as factors than the general population did.4
Contrary to this observation, on Election Day during the 40th General Election some 23 venues in ScarboroughRouge River were not so accessible. In fact, roughly 9,441 registered electors may not have had a fair opportunity to vote. Of the 23 polling stations that were relocated in the 40th General Election, 21 polls were moved 1 km away from their previous location in the 39th General Election
The twenty-three relocations described above may have drastically impacted voter turnout and could account for a significant portion of the 9.5 percentage point decrease from the previous election. Moreover, there were 16 relocated polling stations with extremely poor turnouts of 15 percentage points or greater below Torontos voter turnout percentage (56.4%). Of the relocated polling stations with extremely poor turnout (below 41.4%) 14 polling stations displayed decreases in turnout of at least 5 percentage points from the 2006 election. In other ridings, polling place relocations may or may not have been relevant. However, this was not the case in ScarboroughRouge River. Of the 16 polls with extremely poor turnout (and where polling stations had been relocated), 12 were shown to have dropped 10 percentage points or more from the previous election. Therefore, it can be safely inferred that relocation of various polling stations to locations over 1 km away and requiring electors to travel further, was a material factor in the extremely low turnout percentage.5
Reduced Voter Attention to Contest
Reduced voter attention to the contest refers to the changes in the level of interest of an elector in the election and/or its outcome. It is not the aim of this report to quantify such levels of interest. It is simply suggested that reduced voter attention during the 40th General Election campaign may have contributed to lower voter turnouts in some ridings. In an election where voter turnout dropped to an all-time recorded low for Canadian general elections, it is suggested that some of the drop could be attributed to reduced voter attention. Feedback following the election posits that policy issues during the election were not enough to incite enthusiasm in voters. This was especially evident in a widespread disenchantment among Liberal voters, who turned their back on an uninspiring leader and the contentious Green Shift carbon tax policy.6
Voter attention to the 2008 federal election among ScarboroughRouge River voters may have been reduced for that and/or other reasons. At this point it is important to make a distinction between voter attention and voter apathy. The latter refers to a voters lack of interest in all forms of political participation. It is also suggested that reduced voter attention may have more than one cause. Though reduced voter attention was not necessarily a factor specific to ScarboroughRouge River in 2008, two separate causes of reduced attention might be proposed as having impacted this riding. In addition to the reduced enthusiasm factor described by Heard above, voter attention may have also been affected by the presence on a ballot of a long-time (20 year) incumbent winning by large margins. Admittedly, both these factors could have impacted any given riding; however, few ridings are home to long-time incumbents who have appeared on many consecutive ballots. Take for example the riding of Crowfoot in Alberta, whose winning candidate, a Conservative (Sorensen), also was a long-time incumbent winning by large margins and where the turnout also dropped by a similar 10 percentage points between 2006 and 2008 (64.8% down to 54.9%).
It is not necessary to review the major election issues of the 2008 election given that these apply to all ridings. However it would be interesting to determine whether the presence of a long-time incumbent who wins with large majorities could contribute to decreased voter turnouts over time in a given riding. We propose that the effect of a long-time incumbent on voter turnout could manifest as, the incumbent has won the past many consecutive elections, and therefore the incumbent does not need my vote to win again. This logic may or may not be faulty, but if there is a candidate who usually wins by substantial margins then a voter may not feel like their vote is as important. ScarboroughRouge River and Crowfoot ridings have had incumbent candidates of that nature and the 2006 and 2008 results do suggest that such a pattern.
Disinterest in Party Platform or Leadership
Disinterest in Party platform or leadership may also have dissuaded potential voters in ScarboroughRouge River from voting in October 2008. Based on election results, it is suggested that decreased voter turnout in the riding could be attributed in part to decreased turnout among Liberal supporters. This difference accounted for almost the entire reduction in total ballots cast between 2006 and 2008. Between 2006 and 2008, the total number of ballots cast decreased by 5749 and the number of votes received by the Liberal candidate decreased by 6,567; a difference of only 818 ballots. This trend (in reverse) was evident again in 2004: when the number of votes for the Liberal candidate increased, the result was a roughly equivalent increased number of total ballots cast.
From 2004 to 2006 the number of votes for the Liberals increased by 7,721 and the total ballots cast also increased by 7,166. In a riding which has for 20 years supported the Liberal Party in a city which predominantly follows suit, variation in support for Party leadership or platform appear to manifest at the polls as turnout. In essence, the people express a proclivity to vote with their feet (rather than switch party support) when they are moved or unmoved by Party leadership or platform. It is not the aim of this report to discuss precisely why Liberal support might have reacted to Liberal leadership or policy.
Decreased votes for the Liberal candidate were not redistributed among other parties. Instead, the data indicates that decreased votes for the Liberal candidate led to decreased voter turnout overall. In the 40th General Election, somewhere between 600,000 and 700,000 Liberal supporters did not vote. In fact, 851,525 more Liberals across Canada voted in 2006 than in 2008. Thus, across Canada, the majority of Liberal supporters who did not vote Liberal in 2008 apparently did not vote at all; thus, decreasing the overall voter turnout percentage. This is a trend which also proves true for ScarboroughRouge River.
In looking at the 2008 election results in the riding of Wellington-Halton Hills in the Greater Toronto Area, there is a trend in voter turnout over the past two elections that may similarly reflect this trend. The number of ballots cast for the Conservative incumbent in both 2006 and 2008 was generally the same amount; whereas the Liberal runner-up in 2008 received 4,753 votes less than was received by the Liberal runner-up in 2006. The total ballots cast in 2008 were 4,409 votes less than were cast in 2006. Again, it is evident that the number of votes not received by the Liberal candidate is closely equivalent to the drop in the number of total ballots cast. Thus, the votes not received by the Liberal candidate may have been the result of electors showing their disinterest in Party platform or leader by not voting at all.
This same trend holds true for the riding of Scarborough-Agincourt. A riding in which there is a long-standing incumbent candidate with a history of very large majorities, it is clear that the drop in total ballots cast between 2006 and 2008 is nearly equivalent to the decrease in ballots cast for the Liberal candidate. The total ballots cast in the 40th General Election were 4,589 votes less than were cast in the 39th General Election, while the Liberal incumbent received 5270 votes less in 2008 than in 2006. Though some 681 votes may have been redistributed among other parties, again it is evident that some Liberal supporters displayed disinterest in the Partys platform or leadership by voting with their feet.
Social Science evidence suggests that residential relocation and varying housing arrangements can impact voter turnout. That is, voters who move have lower turnout than those who stay put.7 This might occur for two reasons: Residential mobility and community mobility. One who changes residences within the community is likely to vote for their network of social and political ties so that it is not disrupted. Those who change residences and communities find their social and political network disrupted and in turn are less likely to vote:
The act of relocating disrupts ones social networks of political information and support, thereby reducing movers interest and motivation to vote and ultimately their turnout. Simple interaction with others on political matters increases both information and support, but newcomers have fewer opportunities for interaction than do more stable citizens8.
In Hightons study of voter turnout among movers, residential mobility was compared to community mobility to test the impact of each variable on voter turnout. It was found that those who lived at their residences for ten years or more had a turnout rate 14 percentage points higher than those who lived in their residences for two years or less. Furthermore, people who move must re-establish themselves politically, most importantly by registering to vote. However, Highton observes that whereas all moves impress upon movers the requirement to register to vote, only some voters [those who change communities] experience the disruption of social ties.
According to Statistics Canada, between 2001 and 2006, 48 percent of the ScarboroughRouge River ridings population consisted of movers. Exactly 69% f the ridings movers were non-migrants, ?movers who were living at a different address, but in the same municipality as the one they lived in one year earlier. Though some non-migrants may have changed place of residence within the riding, it is more likely that the majority of the 40,565 non-migrant movers were from outside the riding (but by definition within the same municipality). If this is the case, community mobility may certainly have impacted voter turnout given the number of newcomers to the riding and number of those who have changed place of residence.
The results of a study involving U.S. voters showed that 48% of respondents who had lived in a residence for 2 years or less voted, whereas, 60% of respondents who lived in a residence for 6 years or more, voted. Squire and his colleagues observed that, moving has a substantial and statistically significant impact on turnout when all other variables are controlled.9
Concerning short-term living arrangements such as month-to-month rental agreements and rooming houses, it is believed that those in houses who pay rent, as a class, will vote less than those who own their home. In a US National Election Study that identified the characteristics of movers, the majority (54%) of movers who lived at their current addresses for less than two years were renters. This explained the result of lower voting rates among movers, since it was observed that people who own their own homes turn out at a higher rate than do renters. It is also believed that those who live in apartment buildings and pay rent are less likely to turn out than those who live in houses. There is considerable evidence from within the riding to suggest that there is a significant percentage of constituents living in circumstances in which houses and basement apartments are converted into rooming houses houses where there is an increased number of rooms and increased number of tenants per room. As a percentage of total occupied private dwellings, the number of dwellings in ScarboroughRouge River with more than one person per room is 7.4, six points higher than Canada nationally. In such cases, it is suggested that the owners of these dwelling are registered as voters whereas the tenants manifest much higher community and residential mobility.
Employed and Self-Employed Voters Who Work Long Hours
Like many ridings, ScarboroughRouge River contains a class of employed and self-employed voters who are faced with the choice of voting and taking time away from work or their business. In this sense, it is suggested that voting, for such electors, is an opportunity-cost scenario. Dyck and Gimpel (2005) discuss declining voter participation and the nature of voter abstention as a cost-benefit relationship.10 Are there costs associated with whether an elector chooses to vote or not? Rational choice theory claims that the choice to vote is indeed a basic cost-benefit decision. A voter calculates his or her benefit by: multiplying the perceived benefit that the voter will receive from his or her desired outcome (Party will win) by the probability that the voter will cast the marginally important vote (the vote that will make a difference). From this value, the voter subtracts his or her cost. If the value is positive, the voter votes; if it is negative, he or she does not.
Let us consider some possible costs and benefits an elector in ScarboroughRouge River might face, based on a working knowledge of this riding. Suppose an electors ideal outcome is that the Party they support wins the election. Also suppose they vote after work during the final hours that polls are open and in doing so, they feel that their vote may be one of the decisive votes in achieving their ideal outcome. If these two factors are of a lesser cost than not voting, then an elector is likely to vote. However, there are some cases where the cost is negative and an elector is less likely to vote. It is believed that even if an elector feels that their Party will win and they will cast a decisive ballot, they may not turn out if they are employed, or self employed and work late on Election Day. In such cases the cost of voting is too high, as an elector is faced with having to forego wages or close his/her business early in order to vote. This turnout factor is proposed as material in view of the large numbers of immigrants (67% of the riding) in the Asian and South Asian communities, where many commercial work days run mid-morning to mid-evening (the same as polling hours) both as employees or entrepreneurs. We believe that this variable has some effect on voter turnout, though it is unclear precisely how much.
New Voter Identification Requirements
This was the first general election with voter identification rules that were rigorously imposed upon the electorate. A number of electors were turned away at the polls for not presenting correct ID. Unfortunately, the number of those turned away from polls was not recorded. There were three ways an elector could prove their identity: Provision of an original piece of identification issued by a government agency which contains photo, name and address; in the event that a voter could not meet this requirement, provision of two original pieces of identification, one with the voters name, and one with the voters address. Otherwise, voters were vouched for by an elector whose name appears on the list of electors in the same polling division and who has an acceptable piece or pieces of identification, both of whom were required to make a sworn statement before the returning officer. According to the Report on the Evaluations of the 40th General Election these new requirements represented a significant change for Canadians.
The impact of the new voter identification requirements on voter turnout are certainly open to conjecture. ScarboroughRouge River is a riding that has an extremely large proportion of immigrants (68%), a percentage of whom may have been first time voters. Being a riding also with a high youth population, certainly a number of first-time voters were also youth voters. Presumably, the unknown number of voters turned away from polls on Election Day was more likely to have been first-time voters. Furthermore, the majority of first-time voters would likely fall into the categories of youth and/or recent immigrants. It is not necessarily implied that the voter identification rules deterred a large number of potential voters. It is suggested that if a large number of potential voters were turned away, then they would have fallen into groups that are significant components of the electorate in ScarboroughRouge River. The number of new immigrants (15,575) and youth population (13,700) make up 22% of the ridings total population (130,555) and 29% of the ridings adult voting age population (99,765). Hence, though an unknown number of electors were turned away from the polls for not fulfilling the voter ID requirements, if a significant number of rejected electors who were turned away were either youth or immigrants, both of whom would also likely be first-time voters and unfamiliar with the electoral process, then it would have had a negative impact on voter turnout.
According to the Elections Canada Report on the Evaluations of the 40th General Election, 94 percent of electors were aware of the proof-of-identity requirement, as were 92 percent of those who did not vote. Awareness was lower among those with only high-school education (91 percent) or whose household incomes were $20,000 or less (91 percent). Awareness of the proof-of-address requirement was lower (85 percent) than of the proof-of-identity requirement. Regionally, residents of the Atlantic Provinces (78 percent) and Manitoba and Saskatchewan (79 percent), as well as those living in rural areas (80 percent), were below average in awareness. Among Aboriginal electors, awareness of the proof-of-identity requirement was 84 percent overall, 82 percent among those living in rural areas, and 78 percent among non-voters. This recent data implies that the majority of Canadians living in urban centres would have largely been aware of the changes. Nonetheless, it does not indicate the extent to which recent immigrants or youth and first-time voters would have been aware.
1. See Jon Pammett and Lawrence LeDuc, Explaining the Turnout Decline in Canadian Federal Elections: A New Survey of Non-voters (Elections Canada: 2003) p. 1 http://www.elections.ca/content.asp?section=loi&document=index&dir=tur/tud&lang=e&textonly=false. Viewed August 27, 2009-11-23.
2. Deitland Stolle and Cesi Cruz, Youth Civic Engagement in Canada: Implications for Public Policy in Social Capital in Action – Thematic Policy Studies ( 2005), p. 82 Viewed 28 Aug 2009 http://www.citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.114.2289&rep=rep1&type=pdf#page=90.
3. André Blais in Jon Pammett and Lawrence Leduc, op. cit.
4. Chief Electoral Officer of Canada, Report on the Evaluations of the 40th General Election of October 14, 2008 (Library and Archives Canada Cataloguing in Publication: Ottawa 2009), p. 9., Viewed 28 Aug 2009 http://www.elections.ca/loi/res/40eval/evaluation_e.pdf.
5. A study which took place in Clark County, Nevada (Greater Las Vegas, USA, Area) might be comparable to ScarboroughRouge River in showing that distance can place a levy on voter turnout. See Joshua Dyck and James Gimpel, Distance, turnout, and the convenience of voting, in Social Science Quarterly 86.3, (2005), p. 531. http://www.polsci.buffalo.edu/contrib/faculty_staff/documents/dyck/dyckDistanceTurnoutVoting.pdf Viewed 27 Aug 2009.
6. Andrew Heard, Historical Voter Turnout in Canadian Federal Elections – 1867-2008, (Simon Fraser University: 2008) p. 1, Site last updated 2008, Viewed 27 Aug 2009 http://www.sfu.ca/~aheard/elections/historical-turnout.html.
7. Benjamin Highton, Residential Mobility, Community Mobility and Electoral Participation (2000), p. 116.
8. Craig L. Brians, Residential Mobility, Voter Registration, and Electoral Participation in Canada in Political Research Quarterly: 50.1 (Web: 1997) p. 217.
9. Peverill Squire, Raymond Wolfinger and David Glass, Residential Mobility and Voter Turnout in American Political Science Review 81.1 (1987) p. 52.10. See Joshua Dyck and James Gimpel, Distance, turnout, and the convenience of voting, in Social Science Quarterly, 86.3, (2005), p. 531.
10. See Joshua Dyck and James Gimpel, Distance, turnout, and the convenience of voting, in Social Science Quarterly, 86.3, (2005), p. 531.
Source: Elections Canada Past Election Results, ScarboroughRouge River, Official Voting Results Table 4 & 11.
Existing social science evidence suggests that declining voter turnout among youth is impacting turnout at federal elections and that low turnout among youth contributes to declining turnout in ScarboroughRouge River.
Our conclusion is that the further away a polling station is, the less likely that an elector will vote.