Electoral Insight – Youth Participation in Elections
Electoral Participation and the Knowledge Deficit
Assistant Professor, Department of Political Science, University of New Brunswick
Since the federal election of November 2000, considerable attention has been directed to the declining participation of younger voters in Canadian elections. One recent estimate for the 2000 campaign, which corrects for the tendency of surveys to produce inflated estimates of voting, puts turnout among those aged 18 to 24 at about 25 percent, in an election that saw overall participation drop to 64 percent, Footnote 1 the lowest level since the Second World War. In trying to identify the reasons for voter withdrawal and strategies to win them back, the question of flagging youth participation must figure prominently.
Researchers looking at the phenomenon in greater depth have discovered that it is not as novel as the recent flurry of attention would suggest. Looking at surveys conducted during federal elections from 1968 to 2000, they have identified a long-standing tendency for Canadians to vote less in early adulthood than in later stages of life. Between the ages of 20 and 50, there is roughly a 15-point increase in voter turnout. This – the life-cycle effect in electoral participation – is not, however, especially critical to current concerns, since it is not implicated in the overall decline in voter turnout. Footnote 2 More important in that regard are significant generational differences that have emerged in the past 20 years or so, between those born in the 1960s and 1970s and older Canadians. Turnout among those born in the 1970s, for example, who mostly joined the electorate in the 1990s, is about 20 points lower than turnout was among pre-baby boomers (those born before 1945) when they were young adults. As these younger birth cohorts have come to account for a greater proportion of the electorate, their failure to turn up on election day has started to pinch, accounting for much of the aggregate decline in voter turnout over the past several elections. Footnote 3
A pressing task for further research is to sift through the various factors associated with age differences in voter turnout to determine which underlie these important cohort effects and which are responsible for the less critical life-cycle pattern. One cohort-related factor that stands out is a significant gap in levels of political knowledge between older and younger cohorts. In surveys conducted at the time of the 1984, 1993, 1997 and 2000 general elections, as well as a separate study conducted in 1990, Canadians were asked a wide variety of factual questions that tapped into their knowledge of Canadian politics – questions such as the names of political leaders and the campaign promises of different parties. Each survey included a different bundle of questions, so in order to draw comparisons, it is necessary to standardize in some fashion. A simple method is to assign respondents on each survey a ranking between 0 and 100 (a percentile score, to be precise), based on how their knowledge level compared to other respondents in the same survey. Table 1 uses these standardized scores to capture patterns of political knowledge over time, reporting the gap in knowledge between various birth cohorts and an older comparison group (those born between 1926 and 1938). Two points stand out: first, there are large differences in political knowledge between the younger cohorts and the comparison group, with the gaps tending to grow larger with each successive cohort; and second, those gaps have been closing only marginally over time, even as the younger cohorts have aged (what improvement there is seems to come when cohorts are relatively young, after which the knowledge gap more or less stabilizes). The pattern is more suggestive of cohort effects – sizable and persistent gaps between those born at different points – than a life-cycle pattern of growing knowledge with advancing age.
Entries represent the difference between the mean knowledge level for each cohort and a comparison cohort (1926-1938), where political knowledge is measured on a 0 to 100 scale.
Sources: 1984, 1993, 1997 and 2000 Canadian Election Studies and The Survey of Attitudes About Electoral Reform (1990). Further information on these studies is provided in the Acknowledgements.
This is only one way, however, in which political knowledge is implicated in the cohort effects that have depressed voter turnout in the past several elections. A second lies in the heightened impact of political knowledge on electoral participation among those who have more recently joined the electorate. Breaking down the data for various birth cohorts, Footnote 4 Table 2 reports the gap in voting turnout across a series of elections between those with high levels of political knowledge and those with low levels. The turnout gaps, it would appear, are substantially larger among younger cohorts than older ones. For example, looking at the 1976 to 1982 cohort in the 2000 election, turnout was 46.9 points lower among the least knowledgeable respondents than the most knowledgeable (based on reported turnout of 41.3 percent in the former group and 88.2 percent in the latter). In short, there are two dynamics working together to drive turnout down among younger cohorts: lower levels of knowledge, the effects of which are magnified by the escalating impact of knowledge on participation. The net result is that political knowledge is a critical factor – perhaps the critical factor – underlying cohort differences in voter turnout.
* N < 20 for high knowledge category
Entries represent the gap in voting turnout between those with low political knowledge and those with high knowledge, each representing roughly one third of total respondents. Some cells are empty because there were no respondents in those cohorts for those years.
Sources: 1984, 1993, 1997 and 2000 Canadian Election Studies and The Survey of Attitudes About Electoral Reform (1990)
To this conclusion, the objection might be raised that the problem surely runs deeper, that the knowledge deficit is but a symptom of a more pervasive malaise, namely the wholesale disengagement of young Canadians from politics. In this alternative view, the problem is first and foremost motivational: young Canadians do not vote (and do not know much about politics) simply because they are not interested in politics. However, a similar analysis to the above, looking at levels of political interest across birth cohorts over time, undermines this contention. When it comes to political interest, cohort effects are relatively weak. Instead, political interest exhibits a stronger life-cycle pattern: low levels of interest in early adulthood that pick up substantially as cohorts age. When the various pieces of information are pulled together and plugged into a model of electoral participation over time, the conclusions are reinforced: political interest has its greatest impact on the life-cycle pattern in electoral participation, whereas political knowledge principally explains cohort differences. Footnote 5
Some reflection on the causal linkages between knowledge and interest also undermines the notion that motivation must be the prime mover in matters of political disengagement. If it seems reasonable to suggest that interest provides the incentive to learn about politics, it seems equally reasonable to suppose that knowledge renders politics more intelligible and hence more interesting. Likely the two are linked in a virtuous circle, interest generating knowledge and knowledge piquing interest, with some momentum on either side potentially serving to start the circle spinning. Common sense would suggest that an injection of knowledge could be especially effective at an early age – adolescence, say – when interests are still relatively fluid and malleable. Practical considerations also suggest that emphasizing political knowledge is the sounder strategy: it seems a less daunting task to teach teenagers something about politics than to cajole them into caring about a subject they find categorically boring.
The reasoning above leads to the question of civics education. Are we doing enough in our schools to educate young people about politics? And if we started doing more, would it have positive effects on political knowledge and electoral participation? The literature in this area is somewhat ambiguous. There are suggestions that the long-term impact of civics education is less than overwhelming, especially if it comes too early in the educational cycle. Footnote 6 But broad conclusions can obscure finer differences. More needs to be known about the potential impact of full-throttled civics education in a country that has never emphasized the subject and appears to be trailing other nations in its political knowledge capacities. Footnote 7
Fortunately, we have a ready-made case study at hand: the new civics curriculum introduced in Ontario high schools. Since 2000, all Grade 10 students must take a course simply entitled "Civics" as part of the Canadian and World Studies program. Footnote 8 What makes this initiative especially valuable, in more ways than one, is its compulsory nature. This means, first, that all students, including those who might otherwise avoid the civics program, will be given a basic grounding in the subject. It also offers an advantage to researchers interested in assessing the effects of the new curriculum. Where civics courses are optional, the methodological problem of self-selection arises: students with pre-existing knowledge of and interest in politics are more likely to take the courses than others, making it difficult to know what effect the courses actually have. With self-selection removed from the equation, more definitive assessment of civics education – its impact on levels of knowledge, political engagement and participation – should be possible. And soon too, as at least some of the first graduates should be eligible to vote in the next federal or provincial election. Investigating these matters and disseminating the results so that other provinces might learn from the Ontario experience would be one concrete step to help address the problem of declining electoral participation among young Canadians.
Newspapers, television and Internet
|Age||Attention to politics on television||Political knowledge (0 to 100 scale)|
Attention to politics measured by attention to federal election campaign on television over last few days (a 0 to 10 scale, with 0 to 3 coded as "low", 4 to 6 coded as "moderate", and 7 to 10 coded as "high").
Source: 2000 Canadian Election Study
In reflecting on the potential role of civics education in enhancing levels of knowledge and participation, it is important to bear in mind its multiple channels of influence. As others have noted, there is more to civics education than simply stuffing students full of facts and sending them on their way; equally important are the skills and predilections acquired in school that underwrite a process of continual learning after graduation. The objective should be, as Henry Milner puts it, to instill habits of "civic literacy" so that citizens naturally and effortlessly keep themselves abreast of politics. Footnote 9
One important habit of civic literacy often highlighted is newspaper reading. Newspapers are an important source of political information, but younger generations are less inclined to pick them up. There are, however, other strategies that deserve at least equal emphasis. In particular, television appears to be an especially important medium for acquiring political knowledge, among younger Canadians. Table 3, based on the 2000 Canadian Election Study, reports levels of political knowledge in different age groups as a function of people's attention to the federal election campaign on television; political knowledge, as above, is measured on a 0 to 100 scale. For those under age 30, the impact of television viewing is quite dramatic: a 33-point difference in political knowledge separates those in the low attention category from those in the high attention group. The effect of television viewing is also considerable in the adjacent category, the 30- to 39-year-olds, but diminishes considerably in the older age groups. Reading newspapers also has a considerable impact on political knowledge in all groups, but for the younger respondents – those 18 to 39 – following politics on television makes a bigger difference, particularly when the variables are subjected to controls in multivariate analysis. Footnote 10
One part of a general strategy to raise levels of political knowledge among young Canadians, then, should be to shift the viewing patterns of young people so that they pay greater attention to politics on television. The best place to start is probably in school. In the spirit of building habits of civic literacy, watching broadcasts of politically oriented programming might be woven into the civics curriculum. Just as students are sometimes required to read the newspaper each day for their civics class, so they might be told (and probably to greater effect) to watch a national newscast each night. Or the events of the day might be shown in class and serve as the basis for debate and discussion. To those wedded to the printed word, this may seem like a strategy of capitulation, but it may prove to be the most effective way of sustaining a basic level of political knowledge among generations raised primarily on electronic media.
Another strategy is to encourage use of the Internet as a source of political information. Though not a great deal is currently known about its impact on levels of political knowledge, Footnote 11 consulting reliable news Web sites on a consistent basis could become as effective as reading newspapers regularly. Again, instilling the appropriate habits early, as part of an enhanced civics curriculum, would likely prove an effective way of piggy-backing on a trend – extensive Internet use – that is already well entrenched among younger generations.
These suggestions for ways to raise voter turnout address one dimension of the issue – the lower levels of knowledge exhibited by younger Canadians. But the second component, the heightened impact of knowledge on voting among younger generations, is equally critical. This part of the problem may be a tougher nut to crack.
It seems perfectly reasonable that political knowledge would have some influence on electoral participation. If people are drawn to cast a vote to exercise influence over public policy and those who formulate it, it only stands to reason that people unfamiliar with the issues and political players will be less inclined to vote. Indeed, what is probably most striking about the pattern of voting and not voting across cohorts is not that knowledge makes such a big difference to participation among younger cohorts; it is that knowledge makes such a small difference among older generations. The gap in turnout between more and less knowledgeable individuals in the older age categories has often been around 10 percentage points in recent elections (see Table 2), which means that roughly 80 to 85 percent of politically ill-informed, older citizens choose to vote despite this evident debility.
Lack of knowledge about the issues and the contestants in politics is much less likely to discourage older Canadians from voting than young persons.
We can only surmise that something else – something other than the desire to register one's views on the issues of the day and the leaders who will manage them – sustains electoral participation among older generations. That something else, others have suggested, is a strong sense of civic duty. Footnote 12 It is the responsibility of every citizen, immersed in the issues or not, to cast a ballot on election day. For voters who think this way, going to the polls is more a reflexive instinct than a conscious decision.
Those who highlight the importance of civic duty to voting have at the same time identified a diminished sense of duty among younger age groups. Footnote 13 This is entirely consistent with broader trends highlighted in various studies. Younger generations are less inclined to take cues and directions from those around them; they are more self-directing, more autonomous in their decision making, less likely to defer to outside authority. Footnote 14 When it comes to voting or not voting, younger people are more likely to be guided by their own lights than directed by social pressures.
It is in this context that political knowledge comes to the fore. Self-directed citizens are only likely to vote if they feel it will represent a personally meaningful act. For those who know little about politics, this is unlikely to be the case – ticking one box or another with little information to guide them would be an empty, even counter-productive, gesture. Learning something at election time is always an option, but not a very feasible one for someone starting from scratch; the learning curve would be very steep indeed. It is in this way that knowledge, in the absence of a strong sense of civic duty, comes to assume such influence over the voting decision.
The analysis in this article is admittedly speculative, but it seems a plausible account of the rising importance of knowledge to political participation. It also carries an important implication: the decline in electoral participation among younger Canadians is partly rooted in a pervasive culture shift that has altered the basis for social and political action. Self-directed behaviour is the norm nowadays, and this will not easily be undone. A change in the motivational underpinnings of voting and not voting is probably something that simply has to be accepted.
But if this part of the equation cannot be altered, it can at least be turned to our advantage. The fact that knowledge strongly influences electoral participation in younger cohorts means that efforts to raise levels of political knowledge could have a very sizable impact on turnout levels. Sorting through the various factors that have contributed to declining participation among young voters can help pinpoint where policy leverage exists; it can also help identify forces that might be harnessed to put that leverage to maximum effect.
All datasets were made available by the Institute for Social Research (ISR) at York University. The 1984, 1993, 1997 and 2000 Canadian National Election Studies were funded by the Social Sciences and Humanities Research Council of Canada (SSHRCC). The Survey of Attitudes About Electoral Reform (1990) was conducted for the Royal Commission on Electoral Reform and Party Financing. Principal investigators for the studies are as follows: 1984 Canadian Election Study, R. D. Lambert, S. D. Brown, J. E. Curtis, B. J. Kay and J. M. Wilson; the Survey of Attitudes About Electoral Reform (1990), André Blais and Elisabeth Gidengil; the 1993 Canadian Election Study, Richard Johnston, André Blais, Henry Brady, Elisabeth Gidengil and Neil Nevitte; the 1997 Canadian Election Study, André Blais, Elisabeth Gidengil, Richard Nadeau and Neil Nevitte; and the 2000 Canadian Election Study, André Blais, Elisabeth Gidengil, Richard Nadeau and Neil Nevitte. The original investigators, the study sponsors and the ISR bear no responsibility for the analyses and interpretations presented here.
Return to source of Footnote 1 Jon H. Pammett and Lawrence LeDuc, "Explaining the Turnout Decline in Canadian Federal Elections: A New Survey of Non-voters" [on-line research report], Elections Canada (March 2003), p. 20, available at www.elections.ca under Electoral Law & Policy. The 64 percent figure for overall turnout is the corrected figure produced by Elections Canada after the 2000 election, once the National Register of Electors had been purged of duplicates (the original figure was just over 61 percent).
Return to source of Footnote 2 The logic here is that the life-cycle effect could only diminish overall turnout if young people somehow came to account for a larger portion of the electorate. But, of course, the age distribution of the electorate is more or less constant over time.
Return to source of Footnote 3 André Blais, Elisabeth Gidengil, Neil Nevitte and Richard Nadeau, "The Evolving Nature of Non-Voting: Evidence from Canada," paper presented at the annual meeting of the American Political Science Association, San Francisco, August 30 – September 2, 2001.
Return to source of Footnote 4 The cohorts from Table 1 are collapsed into larger bands of birth years in Table 2 to increase sample sizes and produce more reliable estimates.
Return to source of Footnote 5 This analysis has been conducted using election study data from 1984, 1988, 1993, 1997 and 2000. These results, along with the cohort analysis of political interest, will be detailed in a forthcoming paper by the author.
Return to source of Footnote 6 On this point, see Henry Milner, Civic Literacy: How Informed Citizens Make Democracy Work (Hanover, New Hampshire: University Press of New England, 2002), pp. 118–121.
Return to source of Footnote 7 Results from an international survey of geopolitical knowledge among 18- to 24-year-olds place Canada third from the bottom of a group of nine countries, ahead only of Mexico and the United States. See "Geography Quiz Stumps College-Age Canadians," The Globe and Mail, November 21, 2002, p. A14.
Return to source of Footnote 8 Details are provided in Ontario, Ministry of Education, "The Ontario Curriculum, Grades 9 and 10: Canadian and World Studies", pp. 46–53, available at http://www.edu.gov.on.ca/eng/document/curricul/secondary/canadian/canasc.pdf
Return to source of Footnote 9 Milner, Civic Literacy, especially chapters 7, 8 and 9.
Return to source of Footnote 10 These results are presented in greater detail in Paul Howe, "Name That Premier: The Political Knowledge of Canadians Past and Present," paper presented at the annual meeting of the Canadian Political Science Association, Toronto, May 29–31, 2002, pp. 17–19.
Return to source of Footnote 11 Internet use is included in the multivariate analysis included in Paul Howe, "Name That Premier," in Table 4, p. 31. It has a positive impact, though its effect on political knowledge is smaller than that of either newspaper reading or TV viewing. The Internet measure in the survey used for that analysis, however, is relatively imprecise, simply asking people if they ever use the Internet to become informed about politics. More precise measurement – which Web sites are consulted and how often – might produce stronger results.
Return to source of Footnote 12 André Blais, To Vote or Not to Vote: The Merits and Limits of Rational Choice Theory (Pittsburgh: University of Pittsburgh Press, 2000), pp. 92–114.
Return to source of Footnote 13 Blais, To Vote or Not to Vote, p. 97.
Return to source of Footnote 14 See Neil Nevitte, The Decline of Deference: Canadian Value Change in Cross-National Perspective (Peterborough, Ontario: Broadview Press, 1996) and Russell Dalton, Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies (London: Chatham House Publishers, 2002), especially chapters 1 to 6.
The opinions expressed are those of the authors; they do not necessarily reflect those of the Chief Electoral Officer of Canada.