Are Young Canadians Becoming Political Dropouts?

A Comparative Perspective

Henry Milner June 13th, 2005

Voter turnout in established democracies has been marked recently by a serious decline, but this has been especially acute in Canada. This country’s sharp, steady decline from 75 percent turnout in 1988 to 61 percent in 2004 has been matched only by that of the United Kingdom, among comparable countries. Moreover, converting the Canadian rate from registered voters into potential voters (the measure used in the US) shows that in 2004, for the first time in living memory, fewer Canadians than Americans turned out to vote. Of course, the US election that year was a polarized cliff- hanger; but the 2004 Canadian election – unlike the previous one – was also too close to call, at least until a few days prior.

We know that the decline in turnout was due almost entirely to the failure of those reaching voting age in the last decade to vote in numbers comparable to earlier generations. The implications of this phenomenon have been explored in recent comparative analyses, which show that since voting is habitual, turnout decline will accelerate as newly eligible-to-vote cohorts, set in their nonvoting ways, replace older cohorts with established voting habits.

In addressing this problem, Henry Milner insists upon drawing a fundamental but often-neglected distinction between informed citizens choosing not to vote (politi- cal protestors) and potential voters failing to vote because they lack the basic information needed to dis- tinguish among the choices (political dropouts).

When it comes to providing such information, Canada is found wanting. Young people do poorly in comparative tests of political knowledge. As the authors of the 2004 Canadian Election Study put it, “It is hard to cast an informed ballot if you do not know who the potential prime ministers are or what their parties are promising.” The challenge is thus to reduce the cost for people, especially young people, of being sufficiently informed to cast a vote. This is a matter not only of policies, especially those related to education, but also of institutions – specifically, the system through which elections are conducted.

Emphasis must be placed on civic-education-related measures to promote the habit of attentiveness to politi- cal information. Courses need to be taught in a realistic manner, introducing students to the often conflictual nature of politics. The best way to do this is to bring representatives of the political parties into the classroom – both physically and virtually – using the most up-to- date channels of communication, those that best fit the reading, listening and viewing habits of the emerging generation.

The key is for students to develop the habit of keeping up with political events, so that they will continue to do so after they leave school. Civic education courses are most likely to promote voting if students take them as close as possible to voting age, although given Canada’s elevated school dropout rate, many potential political dropouts will miss such classes. Thus, for optimal results, provinces should consider offering civic education at ages 15 and 16 and lowering the voting age to 16.

Adopting a proportional representation (PR) electoral system would increase interest by giving small parties with distinct principle-based positions on issues and some measure of popular support, such as the Green Party, a better chance of having democratically elected spokespersons to represent them in the classroom. This representativeness can also make the entire political sys- tem more legitimate in the eyes of young people, and the more principled partisanship fostered by a PR envi- ronment – compared to the volatility, ideological inco- herence and thus weak party identification under majoritarian systems – can be expected to have a posi- tive indirect effect.

What results can be expected from targeting those close to voting age with such a civic education program? There is no guarantee that it will significantly increase turnout. Nevertheless, we do know that by fostering political knowledge and the habit of participation in those whose home and external environments provide little in the way of political information, civic education programs will reduce the number of political dropouts.

Introduction

The recent decline in voter turnout in established democracies has been especially acute in Canada.1 Once average or better in this regard than comparable countries, Canada has plummeted. Its recent sharp, steady decline from 75 percent in 1988 to 61 percent in 2004 — the lowest ever, down from 64.1 percent in 2000 — has seen Canada join the traditionally low-turnout United States, Japan and Switzerland at the bottom of the list. Among compa- rable countries, only the United Kingdom experienced as precipitous a decline — from 78 percent in 1992 to 59 percent in 2001. Moreover, the Canadian 61 percent figure is itself somewhat misleading. It implies that turnout is still higher in Canada than in the United States, but, in reality, if the 2004 Canadian rate were to be converted from registered voters to potential voters (the measure used in the US) it would be about 53 percent.2 This puts us well below the unusually high US 2004 turnout rate of roughly 60 percent.3 Of course, that election was extremely polar- ized, and a cliff-hanger; but the 2004 Canadian elec- tion — unlike the previous one — was also too close to call, at least until just a few days prior.4

Moreover, it has become evident that in Canada, as elsewhere (perhaps even more than elsewhere), the key factor in the decline has been abstention among young people. Though people tend to vote more as they get older, the current decline largely reflects a generational phenomenon, since, if we compare by age groups, the largest — indeed the only significant — decline since the late 1980s has been among the under-30s (Gidengil et al. 2003). While in 2004 the turnout among potential first- time voters appears to have improved slightly over the shockingly low level of 2000, the overall trend remains highly worrisome. Clearly, given the further overall turnout decline from 2000 to 2004, the increased participation of first-time voters proved too small to offset the replacement of the older, higher- voting cohorts by those voting cohorts who arrived on the political scene in the 1990s.

Nonvoting by young people is especially acute in Canada, the UK and the US, but the phenomenon is an international one; even some traditionally high- turnout countries, like Finland and Norway,5 have not been spared.6 The implications of such a phenomenon are well set out in a recent comparative analysis of turnout trends in 22 democracies. Franklin argues that age groups (cohorts) are differently affected by the character of elections (2004). Since, as he shows, vot- ing is to an important degree habitual, the crucial group is the young, who have not yet developed habits of voting or nonvoting. These habits are developed, in particular, as a response to the perceived competitive- ness of the first elections for which one is eligible. The initial response is immediate, but the effect is a long- term one: turnout decline will accelerate as newly eligible-to-vote cohorts, set in their nonvoting ways, replace older cohorts with developed voting habits.

Hence, it becomes crucial to address, nowhere more than in Canada, this aspect of the democratic deficit. Otherwise, we face the prospect of a state of affairs in which only a minority of citizens exercises the demo- cratic franchise. Canadians must learn from the expe- rience of other countries that have faced — or managed to substantially avoid — the problem. This paper seeks to contribute to such an effort. In previous comparative work on political participation, I investi- gated the relationship between levels of political knowledge (civic literacy) and electoral turnout (Milner 2002). Here I apply the analytical framework and conclusions of that work, as well as more recent findings, to the problem of declining youth turnout both as a general phenomenon and as a manifestation specific to Canada. In setting out the basic facts as we know them, I insist upon a fundamental, if too often neglected in the literature, distinction between informed citizens who choose not to vote and poten- tial voters who fail to vote because they lack the basic information needed to distinguish among the choices — including the choice not to participate. I contend that failure to adequately differentiate the two phe- nomena has impeded progress in understanding — and thus addressing — this aspect of the democratic deficit. The real threat to democracy, I maintain, lies not in young citizens choosing not to vote, but in their lack of the basic knowledge and skills required to make that choice on an informed basis.

Therefore, in addressing the choices facing policy- makers, I maintain that we must reduce the cost for people, and especially young people, of being suffi- ciently informed to cast a vote. As I will argue, this is not only a matter of policies, especially those related to education, but also of institutions — specifically, the system through which elections are conducted — and the age of eligibility to vote. A key principle is that the institutions through which political leaders are chosen should be conducive to all legitimate political positions being represented and expressed and should reflect popular support for these positions at every level — from Parliament right down to the civics classroom.

Political Dropouts and Political Protestors

Let us start with the young people who reached voting age during the three years before the most recent elections. A study carried out by Elections Canada based on a sample of 95,000 voters drawn from electoral districts in every province and territory found that 38.7 percent of those identified as first-time elec- tors turned out to vote (Elections Canada 2005), com- pared to 22.4 percent for the same group, as estimated by Pammett and Leduc in 2000 (2003, 20). Since the latter conclusion is based on a survey of voters and nonvoters and subject to a much larger margin of error,7 we cannot conclude that 16 percent more voted in 2004. Clearly there was an increase due at least in part to the extra efforts made to register and mobilize this group in the intervening years (this will be dis- cussed further). The 38.7 percent figure still places young Canadians below young Americans,8 and just below youth in Britain — in 2001, only 40 percent of 18-to-25-year-old Britons went to the polls.9 Overall, young Canadians rank not only well below their older compatriots,10 but also below their peers in nearly all comparable countries (see table 1).

What do we know of the nonvoters? By “nonvoters” I mean potential voters who do not vote in elections as a matter of course; thus I exclude those who normally turn out but fail to do so due to factors relating to a specific election. In the case of young people, the dis- tinction is more difficult to draw, since they have had few opportunities to either vote or abstain. Hence, we need to get at the underlying difference between the two groups another way. For the purposes of this paper, young nonvoters are conceptualized as falling into two groups: one is termed “political dropouts”; the other, “political protestors.”

Table 1 Voter Turnout and Attitudes Toward Voting among the Youth

Political dropouts are young citizens so inattentive to the political world around them that they lack the minimal knowledge needed to distinguish, and thus to choose, among parties or candidates. Political dropouts are of special concern, because they constitute a grow- ing group among young people in established democ- racies who, despite being better educated on average, are less attentive to, and thus less informed about, available choices than were young people in earlier generations. Political protestors do not vote either, but, unlike the dropouts, they are sufficiently informed to deliberately forego traditional means of political par- ticipation — party membership and, especially, voting — and instead undertake unconventional forms of political engagement. While it may not always be easy to distinguish between the two groups on the basis of their actions, since the latter tend to act in unorgan- ized ways, we can — as I shall argue — use political knowledge as a convenient proxy.

Unfortunately, it is not yet standard practice to ask political-knowledge questions in surveys about politi- cal participation. When studies of political participation exclude the information dimension, they can, and sometimes do, classify as protestors young peo- ple who are inattentive and who abstain from partici- pating in traditional politics, assuming that they are practising a different kind of politics, one that is inaccessible, even incomprehensible, to older genera- tions. Canadians writing on the subject are not exempt from this tendency,11 but it seems to be espe- cially strong in Britain. For example, the British Electoral Commission found young nonvoters to be disproportionately inclined to state that they did not vote because it made little difference who won the election (UK Electoral Commission 2002, 18). Seizing on these responses, as well as others — such as “No one party stands for me” — a not untypical group of academic observers concluded that “young people are far from being apathetic,” since “politics is something that is done to them, not something they can influ- ence” (O’Toole et al. 2003, 359). Surely, conclusions of this sort would benefit from a test to determine whether the assertion “No one party stands for me” is based on at least a minimal knowledge of what the parties do stand for. In a similar vein, another British survey concluded — based on 71 percent of respondents agreeing with the statement “There aren’t enough opportunities for young people like me to influence political parties” — that young people are “sufficiently interested in political affairs to dispel the myth that they are apathetic and politically lazy” (Henn and Weinstein 2003).

If only. When respondents are given the costless choice of blaming others or, in effect, admitting to being apathetic and politically lazy, the result is cer- tain. Yet it is understandable that political scientists are reluctant to point fingers,12 seeking rather to cast the individual in a positive light. For example, we accept a response of “I’m interested in politics” at face value, seldom probing to see whether that inter- est was actually invested in any effort to gain politi- cal information. Yet political interest and political knowledge are not unrelated. A simple American experiment using both political-interest and political- knowledge questions showed that when asked the political-interest questions first, 75.9 percent of respondents reported following politics most or some of the time; but when they were asked political- knowledge questions first, this percentage dropped to 57.4 percent (Schwarz and Schuman 1997). Similarly, posing knowledge questions allows us to distinguish between political dropouts and political protestors, since both will agree that “All politicians are the same” and that “No one party stands for me” and blame politicians and parties for their lack of partici- pation. Making this distinction is important, since the protestors’ responses — unlike those of the unin- formed and inattentive political dropouts — reflect an informed choice to replace idiosyncratic conventional forms of participation by unconventional ones, a choice that can be revised when the situation changes either objectively or in terms of their own interests.13

Certainly, we should encourage electoral participa- tion on the part of political protestors through insti- tutional reforms (such as those discussed later), since, for one thing, the abstention of these protestors, stripped of its sophisticated rationale, can find its way into the wider generational culture. But we should not confuse such efforts with addressing the political dropout phenomenon, which must be our primary concern. It is, first of all, a simple question of numbers: protestors are a numerically small group. Numbers cited later in this paper suggest that in a few European countries the political protest phenom- enon can be a significant factor in youth turnout decline, but these are typically countries in which the overall decline has been far less precipitous. In Canada, as in the United States and the United Kingdom, bringing political protestors to the polls will have, at best, a marginal effect. The 2004 Canadian Election Study (CES) notes that “[e]vidence of particu- larly strong disaffection with government and politics on the part of young Canadians is…hard to find…[L]evels of [political] disaffection among the young are no more profound than they are among older Canadians” (Gidengil et al. 2005, 8). What does distinguish them is that

the under-30s are much less able to name a political party that would be best at dealing with their number one concern. This finding is not attributable to the fact that many of them see little to choose [from] among the contenders; people in this age group are actu- ally the least likely to think that there is not really a choice…Health may have been a pri- ority issue, but even in the closing days of the campaign, fewer than one in three knew which party was promising four billion dol- lars to reduce waiting times for surgery. Taxes were more important than the environment to young people. Even so, only 28% knew which party was promising to do away with the goods and services tax on family essentials. Most young people opposed increased spend- ing on defense, yet only 40% knew which party was promising to increase military spending by two billion dollars a year. Similarly, a majority of young people opposed scrapping the gun registry, but fewer than one in three knew which party was proposing to do this. (10)

The conclusion is obvious. “It’s not political cynicism that’s keeping young Canadians from voting… Respond- ents in their 20s turned out to be the most satisfied with the way democracy works in Canada” (Gidengil et al. 2004). In asking Canadian abstainers why they failed to vote in the 2000 election, Pammett and Leduc found 18-to-24-year-old respondents to have the lowest ten- dency (27.3 percent, versus 34.4 percent overall) to cite a flaw in the political process as a reason (2003, 17). Clearly, political inattentiveness is something entirely different from political alienation. Young people abstain out of protest less often than members of high-turnout generations. Their abstention is a reflection of lack of political interest and political knowledge,14 and the two are obviously related.

In systematically posing political-knowledge ques- tions, the CES has thus performed an important service for those concerned about the political participation decline in Canada. Unfortunately, national electoral surveys in other countries have generally not followed suit, though data is being accumulated as a result of the proliferation of surveys related to political partici- pation in response to the declining turnout. Increasingly, these surveys focus on young people, but, like the British surveys cited earlier, the most prominent publicly funded or foundation-financed studies seeking to explain declining youth political participation in the US (see, for example, Keeter et al. 2002) and Canada (Pammett and Leduc 2003) still do not ask political-knowledge questions.

For its part, the CES, despite containing more political- knowledge questions than most of its counterparts elsewhere, largely limits them to matters related to the election itself. Consequently, we cannot use the data to concretely compare the political knowledge of young Canadians with their peers in other countries. We need comparative data to better understand the phenomenon of political dropouts, its causes, and its possible consequences. In my own research, I have set out what we do know about comparative political knowledge, stressing the limitations imposed by the absence of a general set of political-knowledge ques- tions for use in international surveys (Milner 2002). In the next section of this paper, I will make a first effort to integrate into this analysis the findings of the emerging literature on youth political participation.

Political Knowledge and Voting Turnout among Young People

Before addressing the knowledge dimension directly, we should note that it is not only knowledge that brings people to the ballot box. Another obvious factor related to plummeting voter turnout is a decline in the sense of civic duty to vote. This declining sense of obligation, when set in the context of a wider generational culture given to political inattentiveness, can transform a provisional act of abstention into a habit of political dropping out. The 2004 CES team reported that

Seventy-five per cent of our respondents strongly agreed that ‘It is every citizen’s duty to vote in federal elections,’ and 32 per cent said that they’d feel very guilty if they didn’t vote in a federal election…However, young Canadians are much less likely to share these sentiments: Only 55 per cent strongly agreed with the statement about duty, and only 18 per cent said that not voting would make them feel very guilty.” (Gidengil et al. 2004)

The comparative data on civic duty reflect the gen- erational character of the decline, though more in some countries than others. For example, an American survey of 3,246 adults 15 years and older found that only 38 percent of 15- to 25-year-olds say that citizenship entails special obligations, while 58 percent say simply being a good person is enough. This is markedly different from the responses provided by older generations; between one-half and two-thirds choose special obligations (Andolina et al. 2002). Another study found that just 20 percent of young people described voting as a responsibility, and only 9 percent as a duty.15 Differences in Europe seem less marked. The European Social Survey (ESS) of 2002 asked how important it is for a good citizen to vote in elections in 20 new and old European democ- racies, using an 11-point scale from “extremely unim- portant” to “extremely important.” As set out in the last two columns of table 1, the overall average was 7.61, while for those under 25 the average was 6.89.16

The data show that far more people in earlier gener- ations with marginal levels of political knowledge voted out of a sense of civic duty than is the case with young people today. This is illustrated by the remark- able generational difference in the UK, where 63 percent of those who claimed they were “not at all interested in news about the election” nevertheless cast a vote, but among the 18- to 24-year-olds, this was the case for only 16 percent (UK Electoral Commission 2002, 29). A similar phenomenon has been observed in Canada. Howe compared data from 1956 Gallup Polls testing political knowledge with those from the political- knowledge items in the 2000 Canadian Election Study (Howe 2003).17 Age differences turned out to be signifi- cantly more important in 2000, especially among those with no more than a high-school education. The young were both less informed about politics in 2000 than they were 45 years before and more likely to have this condition influence their decision to vote or not to vote. In 1956, the difference in reported turnout level between the groups at the lower and upper ends of the knowledge scale was 17 percentage points; moreover, for the youngest age group (21 to 29 years), the differ- ence was actually lower — only 12 points. The 2000 election study showed that the overall gap in turnout between the knowledgeable and the ignorant had risen to 32 points; but the relationship to age was reversed. The 43-point gap that separated the least and most knowledgeable respondents aged 18 to 29 declined with age to 13 percent among those 50 and older. “Nowadays…it is only older Canadians who will vote simply out of duty,” Howe concluded quite pessimisti- cally. “[Y]ounger Canadians think differently; without some knowledge to make the voting decision compre- hensible and meaningful, they prefer to abstain…They know less about politics and…their impoverished knowledge is more likely to affect whether or not they vote” (2003, 81).

We can thus understand that for young people, not casting a vote can easily become a habit that in turn diminishes their already limited interest in politics. Lacking a sense of civic duty to vote, young people are less inclined to seek the information they need to vote meaningfully, and their declining sense of civic duty makes turning out to vote increasingly dependent on an adequate level of political knowledge. It becomes evident that, more than ever, addressing the decline in turnout means enhancing political knowledge.

But what do we know of the differences in levels of political knowledge among young people and their relationship to the phenomenon of political dropouts? In my recent work, using a variety of mainly indirect indicators, I identify the northern European, and espe- cially Scandinavian, countries as high-civic-literacy countries — that is, countries where the proportion of citizens sufficiently informed to vote meaningfully is relatively high (Milner 2002). In contrast, the English- speaking countries tend to fall into the low-civic- literacy category. Unfortunately, in the research that led to those findings, young citizens were not singled out to determine if their relative levels of political knowledge corresponded to those of the country as a whole. However, subsequent cross-national research has begun to address this question. A useful contribu- tion is made by Grönlund, who assembled the responses to the three political-knowledge questions in recent election surveys in 23 countries participating in the Comparative Studies of Electoral Systems (CSES). He found that at all levels of education 18- to 35-year-olds are less knowledgeable on political mat- ters than their elders (Grönlund 2003).18

Grönlund confirms what has been shown by many single-country studies.19 Yet the difference appears especially acute in North America today. For exam- ple, the Times Mirror Center analyzed survey results from the 1940s through the 1970s, revealing that pre- vious generations of young people knew as much as, if not more than, their elders (1990). This is in com- parison to Parker and Deane, who reported that, on average, only 36 percent of Americans under 30 answered the information questions correctly, com- pared to 45 percent of those aged 30 to 49 and and 49 percent of those aged 50 and over. Only 26 per- cent of young people answered campaign-related questions correctly, compared to 38 percent of those 30 to 49 and 42 percent of those 50 and over (1997). Responding to questions related to national politics, young people averaged 32 percent correct answers, compared to 44 percent of middle-aged Americans and 48 percent of those 50 and older.20

Similarly, in Canada, in a 1990 survey carried out for the Royal Commission on Electoral Reform and Party Financing, 56 percent of 18- to 29-year-olds were able to answer correctly, at most, one of three political-knowledge questions, compared to 40 percent for the sample as a whole. A 2000 survey showed the younger group to be falling further behind: fully 67 percent of 18- to 29-year-olds correctly answered no more than one out of three questions,21 compared to 46 percent for the sample as a whole (Howe 2001). Even greater disparities were reported by the authors of the 2004 Canadian Election Study.

Even in the campaign’s final days, only 60 per cent of respondents in their 20s could name Paul Martin as Liberal Party leader…During the first 10 days of the campaign, a mere 38 per cent knew this basic fact. Gilles Duceppe fared little better, despite the fact that the Bloc Québécois has traditionally held more appeal for young voters: Only 64 per cent of young Quebeckers interviewed in the final 10 days of the campaign could come up with his name. Across Canada, in the campaign’s last days, only 47 per cent of young Canadians could name Stephen Harper as leader of the Conservative Party. Just 34 per cent got the name of NDP leader Jack Layton right (and only 50 per cent of young Canadians could name their provincial premier). (Gidengil et al. 2004)

The only international survey that allows us to place such findings about the political knowledge of young North Americans in a comparative context is a recent study of the political geography knowledge of young people.22 In 2002, the National Geographic-Roper Global Geographic Literacy Survey assessed 3,250 young adults in nine countries. Respondents were asked to identify countries on a world map, and there was another series of questions testing knowledge related to international politics.23 As we see in table 2, of the 56 questions asked in the countries surveyed, young Americans, on average, answered 23 questions correctly (just ahead of the last-place Mexicans); young people in Canada (27) and Great Britain (28) fared almost as poorly. Sweden led (with 40), followed by Germany and Italy (both 38), then came France (34) and Japan (31).

Of course, this is but one survey. Yet its results corre- spond reasonably closely to those we would expect from levels of overall civic literacy (Milner 2002). Moreover, they also correspond, as hypothesized, to lev- els of turnout. As we see in table 1, of those countries also participating in the European Social Survey, young Swedish respondents reported having voted at 81.4 percent, Germans at 72.8 percent, and Italians at 76.4 percent, compared to 41 percent for those in the UK.

Table 2 Young Peoples Knowledge of Political Geography

Clearly, the low and declining electoral turnout in North America reflects low and declining levels of political knowledge even more among the young than among citizens as a whole. We must then ask if the causes, and the possible intervention measures, are the same and also how they differ. An obvious case of the latter is in the area of education, and we shall espec- ially look at educational initiatives to address the deficit in political knowledge and attentiveness. But we first need to consider the institutional context — specifically, the effects of electoral institutions. Despite wide discussion about the effect of electoral institu- tions on turnout, there is still much we do not know. But we do know one important thing when it comes to explaining this relationship. As I argue in the section to follow, political knowledge is a key intervening variable in explaining the higher average voter turnout in countries using proportional electoral systems.

Institutions and Policies Associated with Political Knowledge and Turnout among Young People

Electoral institutions

We can begin our exploration of political knowledge as an intervening variable in the relationship between proportional electoral systems and voter turnout with the results of the National Geographic-Roper survey, displayed in table 2. These suggest a relationship between electoral insti- tutions and the civic literacy of young people. Among the eight countries (excluding Mexico), the three with first-past-the-post (FPTP) electoral systems — the United States, Canada and Great Britain — scored low- est in terms of civic literacy of young people, while the three proportional representation (PR) countries, Sweden, Germany and Italy (which changed recently from proportional to semi-proportional), scored high- est. In between was Japan, with its mixed system, and France, with its second-ballot system. Why should these seemingly unrelated variables be connected?

The most recent assessment for the last election in 39 democracies where voting is not compulsory found turnout averaged 68.2 percent in non- proportional systems compared to 70.8 percent in proportional systems (Farrell 2001).24 Much of the dis- cussion about higher turnout under proportional elec- toral systems focuses on votes counting for more under PR than under FPTP, where votes in uncompet- itive districts are “wasted.” However, this explanation is insufficient, since even under PR a single vote almost never changes the outcome. It is only when we incorporate the incentive that parties have under PR to mobilize all potential supporters, and not just those in winnable districts, that we begin to approach a full explanation (see Aldrich 1993). And in doing so, we introduce the factor of political knowledge. This is because fundamental to mobilizing support — especially when the electoral rules encourage this, as they tend to do in PR countries (Bowler, Carter, and Farrell 2000) — is the task of informing potential vot- ers. To put it simply, all things being equal, more vot- ers receive information from more political parties under PR than under majoritarian systems.

My contention, therefore, is this: All things being equal, voters are likely to be better informed under PR than under FPTP. This goes against conventional thinking, which assumes that voting under FPTP is a simpler proposition since it is typically a choice between “keeping the bums in, or kicking them out.” But such conventional thinking views voters one- dimensionally. Parties under PR are not subject to the volatility of FPTP, which blows up a party’s strength when it does well and shrivels it when it does poorly, thereby discouraging it from operating at levels — national, regional and local — other than the one at which it is best organized. In other words, because PR systems are more conducive to the formation and durability of ideologically coherent parties that con- test elections at more than one level, they provide potential voters with a political map that is relatively clearly drawn and stable across time and space. They make it easier for the potential voter to identify with a given political party and to use that identification as a guide in dealing with complex issues and actors over time at various levels of political activity. In this way, I maintain, PR fosters political participation, especially at the lower end of the education and income ladders, where information is at a premium.

The data for directly testing this claim compara- tively using political knowledge as the dependent variable are inadequate, since there is as yet no set of political-knowledge questions used cross-nationally. Nevertheless, it is possible to derive insights from the responses to the CSES’s political-knowledge questions. In a recent paper, a colleague and I, using the CSES data, examined the dispersion of political knowledge among educational categories by calculating the com- parative variation from the mean of the political- knowledge score of the group with the lowest education (Milner and Grönlund 2004). If the analysis presented here is correct, average dispersion would be lower under PR because it reduces the cost of the political knowledge needed to make an informed vote for those for whom the cost is highest — that is, those lacking in educational resources. The results, confirming this hypothesis, are presented in figure 1, which corre- lates the data derived from Lijphart’s application of the Gallagher Index of Disproportionality to general elec- tions from 1945 to 1996 (on the X axis) (1999, 162), with education-related dispersion of political knowledge (on the Y axis). As hypothesized, figure 1 reveals a strong linear association: as electoral outcomes become more proportional to the popular support attained by political parties, political knowledge becomes less dependent on formal education.25

Figure 1 The Relationship between the Effects of Education upon Political Knowledge and Electoral Disproportionality

If the same logic applies to young people, and there is every reason to believe it does, then we can better understand the even stronger relationship between pro- portional electoral systems and youth turnout. In examining differences in turnout level for voters between the ages of 18 and 29 in 15 Western European countries in the late 1990s, International IDEA esti- mated that in countries using PR systems, the average youth turnout rate was almost 12 percentage points higher than in non-PR countries (IDEA 1999, 30).

IDEA’s interpretation of the difference stressed that PR electoral systems facilitate access to representation in Parliament for small parties by making the propor- tion of seats correspond to the proportion of votes. This is surely true, but the observation applies to informed young people poorly represented under majoritarian systems, those we could term potential political pro- testors, rather than to the uninformed political dropouts. The wider effect of PR electoral systems, I contend, is on what might be termed potential politi- cal dropouts. Here the key factor is the electoral sys- tem’s effect on political knowledge, as set out earlier. Therefore, part of the explanation — and possibly of the remedy — for low and declining turnout in the UK, the US and Canada lies in the electoral system.

Of course, electoral system reforms cannot in them- selves address the purely generational aspects of the phenomenon. Though the drop witnessed was not as great as in Britain and Canada, certain traditionally high-turnout PR states have also experienced in recent years a real decline in turnout for legislative elections; examples include Finland (77.3 to 65.2 percent between 1987 and 2003)26 and Norway (81.5 to 73.1 percent between 1989 and 2001). Moreover, in New Zealand, which adopted the mixed-member proportional (MMP) form of PR in 1996, we observe a kind of spike: turnout rose by about 3 percent in 1996, and the decline that marked the 1980s resumed in 1999. Similarly, in Scotland, which also uses an MMP form of PR, the turnout of voters casting ballots in the new Assembly elections fell in the second election (to 49.4 from the 58.8 percent recorded in 1999), held in 2003. We should note in this context that PR is far from rooted in these countries. In New Zealand, PR has only made it in fits and starts to local elections;27 and Scotland, which is only now moving to introduce single transferable vote (STV) for local elections, operates in the context of Westminster’s FPTP environment. Neither has yet built proportionality into the wider political landscape, sim- plifying the political map by rendering citizens’ experi- ence consistent over time and space — as have Germany, Sweden and other high-turnout European countries that have used PR for many years and at all levels.

What this means is that in and of itself, the effect of changing Canada’s electoral system, now a practi- cal possibility in several provinces (Milner 2004a, 2004b), will only marginally improve turnout, specifi- cally by allowing into the legislatures smaller parties,28 in particular the Green Party, which won 4 percent of the vote in 2004 and seems to be especially popular among informed young people. Nevertheless, given that Canadian young people are, overall, more main- stream in their political attitudes than their elders,29 the number of young supporters of excluded parties brought to the polls by PR would be small.

Greater competitiveness could perhaps bring a slightly larger proportion of young abstainers to the ballot box in Britain, given young voters’ widely expressed and active dislike of the main party candi- dates (MORI 2001, 21).30 A wider range of real alterna- tives under PR could bring some additional young people to the polls, many of whom, one may presume, would be political protesters who would be voting ear- lier in their lives than they would otherwise have done.

As far as the US is concerned, the issue is moot, due to the impossibility of meaningfully changing the electoral system. But there is no doubt that the virtual disappearance of competitive congressional districts as the result of widespread gerrymandering and the decline in the number of states that are competitive in presidential and senatorial elections has created a steady decline in turnout in years where congressional elections are not accompanied by presidential ones.

Moreover, apart from giving smaller parties an opportunity to win seats, PR adds an element of uncertainty to overall outcomes, even when one party dominates. And, as noted, it removes the disincen- tives that FPTP places upon parties and voters in uncompetitive districts. The combined effect, even if marginal, is likely to be especially strong in Canada, as there is a comparatively large number of safe seats due to regional voting patterns.

Finally, however limited the immediate effect of more competitive elections on young people, an important recent analysis reminds us of the long- term importance of that effect. According to Franklin, the main factor explaining turnout is the “character of elections,” as defined by the type of electoral sys- tem, the fractionalization of the party system, the time elapsed since the previous election, the closeness of the outcome and other factors related to competi- tiveness (2004). Changes in the character of elections, he shows, largely account for the average 7 percent turnout drop in the past 30 years in the 22 countries studied — in large part by affecting the habits of young (non)voters. More than anything else, Franklin argues, it is the character of the first election they encounter that will influence whether they vote over the long term. This could explain both higher partici- pation by first-time voters in 2004 and the fact that, despite this, turnout in Canada has continued to decline. If Franklin is right, increased competitiveness in 2004 over 2000 attracted more first-time voters,31 but it had less of an effect on those now in their 20s who developed habits of abstaining in the three pre- ceding uncompetitive elections. The more uncertainty exists about the outcomes of upcoming elections, the greater the competitiveness effect. But history teaches us that such competitiveness is far from assured, since under FPTP Canada has alternated between periods of single-party domination and two-party competition. The only way to assure such competi- tiveness is through electoral-system reform.

Having considered all this, we should bear in mind that PR in itself provides no protection against the other factors that account for turnout decline. For example, a shift to a culture that measures the value of an activity according only to its meaningfulness to the individual carrying it out will dampen turnout under any institutional arrangement. To sum up, we might say that although PR is ultimately no protec- tion against turnout decline, its installation in combi- nation with complementary reforms (discussed in the last part of this paper) can be expected to at least slow, and perhaps even reverse, the decline.

The media

In my work on civic literacy, I find an extremely strong negative relationship at the aggregate level between television-watching (especially commercial television) and political knowledge, and a positive relationship between newspaper-reading and political knowledge. Table 3, based on data from the World Association of Newspapers, shows that Canada, like the US, is worrisomely close to the bottom of the list of comparable countries when it comes to the reported daily newspaper readership of young people compared to that of adults as a whole. American surveys have clearly shown that part of the explanation for youth abstention is the decline in attentiveness to politics as reported in the media. For example, the spring 1998 Pew Research Center Biennial News Use Survey (of 4,002 adults) revealed that only 33 percent of Americans aged 18 to 29 made an effort to keep up with the news, compared to 68 percent of those over 50; the latter group was almost twice as likely to follow national politics and domestic policy closely, and 10 percentage points more likely to follow election cam- paigns and international politics (Bennett 1998). An early-1998 Pew survey of first-year college students found that only 26.7 percent thought that keeping up to date with political affairs was very important or essen- tial, compared to 57.8 percent in 1966 (Bronner 1998).

Table 3 Age Structure of Daily Newspaper Readership

Of course, a significant boost in young people’s political interest took place around the time of the 2004 US elec- tion. Yet a similar boost in political interest in the after- math of September 11, 2001, proved ephemeral, as reflected in a 2002 national telephone survey of 3,246 Americans 15 years and older (Andolina et al. 2002), which found that only 24 percent of those 25 and under reported following government and public affairs “very often” (compared to 60 percent of the oldest group, 50 per- cent of baby boomers and 37 percent of those 25 to 35).

There is thus a relationship between inattentiveness to media and the phenomenon of political dropouts. A clue to its workings is provided by Howe’s finding (noted earlier) that the effect of age in Canada is sig- nificantly more important today than it was in the 1950s, especially among young males with no more than a high-school education: those under 30 average 30 percent lower levels of political knowledge than those over 50 with high-school education or less (Howe 2003). Of course, the link between social class and political attentiveness and participation is a long- established one, and it is even more manifest in the US than in Canada.32 But there were periods in US his- tory characterized by “life experiences…dampening the biases in patterns of political participation attrib- utable to socioeconomic status” (Strate et al. 1989, 456). Specifically, during the high-mobilization period in the United States in the 1960s and 1970s, political participation increased from age 18 to 65 only mar- ginally among the best educated, but significantly (from 20 percent to over 50 percent) among the least well educated (Wolfinger and Rosenstone 1980). Unfortunately, in the current era we cannot count on this kind of mobilization over the life cycle to awaken attentiveness to media.

Clearly — in the North American context, at least — a media-focused approach can only indirectly affect young political dropouts. To have any real effect, such an approach needs to be integrated into the lives of potential young dropouts. Practically speaking, this places the focus of media-oriented measures on education, and particularly civic educa- tion (explored in the next section). There is some international experience in this area, since, under- standably, newspaper publishers have been keen to encourage the newspaper-reading habit among stu- dents. A series of efforts and experiments supported by the World Association of Newspapers’ Newspapers- in-Education program and the European Newspaper Publishers’ Association have shown some promising results33 — but, in this age of electronic communica- tion, we must not set our hopes too high.

One avenue to be explored thus is the new media, especially the forms favoured by young people. Yet, given the anarchic nature of the Internet, it will be difficult to get young people to use it in ways other than those to which they are accustomed. It is a path more readily accessible to high-turnout, high-civic- literacy societies, like those of the Nordic countries, which have traditionally led the way in newspaper consumption among young people as well as among adults (see table 3). A recent study comparing the political participation of youth in Scandinavia and Europe as a whole, using data from the recent European Social Survey (ESS), found that although Nordic respondents of all age groups stand out in agreeing that it is important for a good citizen to vote in elections, the youngest group of Scandinavians, unlike middle-aged and older Scandinavians, are no longer significantly more likely to watch TV news and current affairs shows and read newspapers than their non-Nordic peers. But they do stand out in their use of the Internet and e- mail (Ersson and Milner forthcoming).

This brings us, finally, to an unexplored dimension of turnout — namely, access to the ballot box. In the mature democracies, much has been learned and accomplished with regard to mechanisms for enhanc- ing voter access. Use of the new information tech- nologies (“e-democracy”) raises new possibilities in this regard. But it is not a panacea, or even a simple matter.34 Electronic information can help bring people to the ballot box by making information about when, where and how to vote more readily available. It is interesting to note in this context the findings of a study of the effects of state efforts to simplify the registration and voting process with the intent of boost- ing youth turnout.35 The main effect of postal voting, for example, appears to be to facilitate voting for those who normally vote (Karp and Banducci 2001). If e-democracy has any effect on turnout, it is likely to be of the same nature, and thus cannot be expected to bring into the electoral arena those otherwise ex- cluded. Moreover, e-democracy — like postal voting, but to a potentially far greater degree — carries the risk of distancing people from the human exchange that has been a key dimension of political life.36

Finally, in the next section I turn to the key area of policy intervention, namely civic education. To overstate what follows, I shall argue that we need to bring politicians into the classroom. Part of this argu- ment is linked to an aspect of the media’s role that has so far not been raised. It is commonly agreed that many potential young voters are “turned off” by what they judge to be inauthenticity in politicians, and this judgment is based on what they see on TV and hear on the radio — especially during election campaigns. The media are often blamed in this for practicing “gotcha journalism.” But to some extent the news media are simply doing and will continue to do their jobs by putting politicians on the defensive. Thus it is unrealistic to expect young potential voters to see politicians seeking office as authentic when their only contact is through the prism of an adversarial and ratings-driven media. In effect, by abstaining at the ballot box young people are voting all politicians “out of the apartment,” in much the same way that young viewers of reality shows vote in large numbers to expel residents they find inauthentic from commu- nal apartments or desert islands. If, as I suggest below, it was standard for politicians to visit civic education classes, large numbers of young people would be exposed to another, potentially more authentic side of those seeking their votes than that provided through the media.

Civic education in the classroom

Where knowledge is concerned, an obvious sphere of policy intervention is education. I have shown that widespread use of various forms of adult education, directed especially at those with comparatively low levels of educational attainment, distinguish the high-civic-literacy countries, Sweden in particular — a country I have studied in depth (Milner 2002).37 When it comes to young people, the focus is evident- ly on courses taken during the years of schooling. Indeed, it is nowadays regularly claimed that civics education is the solution to the problem of youth dis- engagement. While I, too, shall argue that civics edu- cation is an indispensable element of any approach aimed at addressing political dropouts, one cannot assume from the existing literature that such courses are certain to have an appreciable, lasting effect.38 Much depends on specific factors, such as the age of the students and the content, methodology and con- text of the courses given.

The general situation was not long ago described by Dekker and Portengen as follows: “[S]ocial studies is a low status area of the school curriculum in many countries. Politics is only one of the subjects in social studies and receives attention for only a small part of the few school hours reserved for the subject. Many social studies teachers do not give priority to political topics [and] have limited political knowledge them- selves” (2000, 467). Moreover, much of it is targeted at adolescents; yet there is good reason to believe that adolescence is a stage of life not especially conducive to the kind of learning provided by civics courses.39 Given these factors, we should not be surprised that a recent American study found practically no positive effects on later voting from exposure to various forms of civics-related high-school courses (Lopez 2003).

Different results, however, begin to emerge with regard to civics courses given at the end of the period of secondary education, when students are on the threshold of adulthood (Niemi and Junn 1998). There is some American, and especially Swedish, evidence that civics courses taken in very late adolescence enhance the political knowledge of students (see Westholm, Lindquist, and Niemi 1989). Still, we do not know to what extent the information acquired in such courses is retained into adulthood and affects long-term political participation. One thing to remem- ber is that Swedish secondary students are older than their Canadian counterparts and the dropout rate is far lower in Sweden than it is in Canada,40 so that given to 17- and 18-year-olds, the civics courses would address a significantly larger fraction of the cohort than they would in North America.

Of course, the advantages of young people remain- ing in school longer transcend the positive effects of civics courses. Since school dropouts are probable political dropouts, the first step toward decreasing the number of political dropouts is to keep young people in school long enough to acquire the skills and habits of literacy that will help them to take their place in society as adults. Though beyond the scope of this paper, one priority must be to identify and counteract the societal factors that lead young people to drop out of school in their mid-teens (especially young men, who drop out of school at much higher rates than young women), only to face a life of functional illiter- acy and marginality. It means that schools must make even greater efforts to foster appropriate habits of liter- acy, reading, writing, library use and the like.

When it comes to civics courses, we do not yet have good, systematic data about the age of the students and the duration and content of the courses. In the US, 41 states’ statutes specifically provide for the teaching of social studies, which can include government, civics and/or citizenship. While 39 states require course cred- its in government or civics for high-school graduation, only 5 of those states require students to pass an exit exam that includes social studies: Georgia, Louisiana, Mississippi, New Mexico and New York (CIRCLE “High School”). As far as Canada is concerned, there is little systematic information on the measures used in each province, but, overall, efforts appear to be relatively limited (Hébert 2002) — certainly as compared to those of unitary states, like Sweden, where the level of gov- ernment concerned with citizenship is also responsible for education. Four provinces have some sort of com- pulsory program. Ontario, for example, has a relatively new civics curriculum that emphasizes participatory learning, but it is voluntary, left up to the local school boards to implement. My own province of Quebec, despite its identity as a “distinct society,” is passive in its approach. It seeks to encourage civic education as a dimension of other high-school courses, such as histo- ry; the junior colleges (or CEGEPs) — where 16- to 18- year-olds still in school are found — are a wasteland as far as civics education is concerned.

Without assuming that we could attain the same results, given the different cultural context, there is good reason to take Sweden as a model when it comes to citizenship education. It is reasonable to presume that part of the explanation for the low (4.3 percent) turnout-level gap between 18- to-29-year-old Swedes and all others (IDEA 1999), and the relatively positive attitude of young Swedes toward voting (see table 1), is that they take civics courses at 18 (or close to it), when they are poised to apply the course information con- cretely as new voters.41 Thus, we should not be surprised to learn that the relatively modest turnout decline in Sweden over the past 30 years is not synonymous with a declining interest in politics — quite the contrary (Holmberg and Oscarsson 2004). However, for Canada to even aspire to results of this kind in light of its school dropout rate, it would have to offer civics courses to students at age 16 (if not 15). But this is well before the voting age — the point at which young people can put such learning into practice at the ballot box. Clearly, any effort to bring Canadian political dropout levels closer to Swedish levels through civic education would require lowering the voting age.

Turning to course content, we can again learn from Swedish practice. In designing civics courses targeted at young people about to become citizens and voters, importance is accorded to the positions taken by the different parties on relevant local, regional and national issues. Party spokespeople are regularly invited into the classroom42 — an opposite approach to the American one of keeping politics out of the classroom.43 Having parties present their own positions would also serve to allay fears in certain Canadian provinces — Quebec, for example — that teachers would be partisan in their presentation of the alternative positions. In this context, I should add, e-democracy could be effective. Given the increased use and sophistication of Internet-based information provided by the political parties, the school visits could be virtual as well as physical.

This approach can be contrasted with the American one, which, to the virtual exclusion of poli- tics, emphasizes the nation’s history and constitution in the civics classroom and encourages community- based volunteer activities outside it. As far as the lat- ter is concerned, we have good reason to doubt whether such activities positively and directly affect attentiveness to, or interest in, politics.44 Indeed, there is mounting evidence that the great stress on youth volunteering that has characterized the American response to declining civic engagement has, if any- thing, depoliticized participants. One representative study found that service experiences did not change “the students’ assessments of the value of elections” or their “definitions of what civic responsibility is and should be” (Hunter and Brisbin 2000, 625).

Instead of stressing constitution, history and insti- tutions to the effective exclusion of different party positions on policies and issues, the opposite approach, as a recent paper shows, can have a posi- tive impact on political knowledge. Stroupe and Sabato compared classes that used the National Youth Leadership Initiative (YLI) curriculum and a control group of similar classes that did not (2004). They found that YLI programs have substantial positive effects on students’ levels of political knowledge and, to a lesser degree, on their likelihood of future politi- cal participation.45 Similarly, another study found that students who participate in open class discussions and who learn to communicate their opinions through letter-writing and debate are much more active than those who don’t have these experiences (Andolina et al. 2002). A North Carolina study found that “young people who reported having to stay cur- rent on political events showed higher levels of polit- ical knowledge…and interest in voting” (Henzey 2003). A recent study takes this even further, finding that the degree to which political and social issues are discussed has a greater impact on civic profic- iency than the frequency of social studies classes (Campbell 2005).46

Civic education outside the classroom

This approach to civic education transcends the class- room. The cultivation of habits of literacy in the earl- ier years of education should therefore be complemented by a set of activities designed to estab- lish habits of citizenship among young people as they approach voting age. A useful list of suggestions along these lines was advanced in the International IDEA report. They included various efforts to register young citizens, mock elections, and specially targeted artistic and cultural events. One recommendation was to make first-time voting a rite of passage by sending, for example, congratulatory birthday cards to new voters explaining how and when to register for elec- tions, or by adding an element of public spectacle through a national youth voter registration day. In 2004, Elections Canada was especially active in pro- moting voting by the youth using such methods;47 a glance at its Web site demonstrates this.48 One page provides links to relevant youth organizations, such as Rush the Vote and Youth Vote 2004, which it sup- ported in cooperation with the Dominion Institute.49 A useful set of broad recommendations relating to this issue was recently put forward by the New Brunswick Commission on Legislative Democracy (2004, chap. 5).

Perhaps the most interesting of the various activi- ties that complement civic education is the mock elec- tion. Mock elections were initiated in North America in 1998 with Kids Voting USA, a nonprofit, non- partisan voter-education program. In schools in 39 states, teachers help students gather information about the candidates and issues in local, state and national elections. On election day, older students cast their ballots in special booths; the younger ones go to the polls with their parents. Research has shown that the effect of this initiative has been positive, especial- ly on the parents.50 This aspect has been absent from Canadian spinoffs. In the first of these, in Ontario, students in grades 9 to 12 in about three-quarters of the public high schools cast ballots supplied by Elections Ontario identical to those used in the October 2003 provincial election. The ballots of over 43 percent of Ontario high-school students were col- lected and tabulated, and the results were presented live on CBC Television alongside the results of the official vote. On the whole, the students took the exer- cise seriously and used it to inform themselves. Civics or history classes often took the lead in engaging and informing the student body during the campaign. This was followed by Student Vote 2004, organized around the 2004 federal election, which was held on June 28. Unfortunately, this date was too late to allow for a simultaneous vote, so each school selected an election day. Results were tabulated for 1,168 schools in 267 ridings across Canada, with over 265,000 students casting ballots. The numbers were clearly kept down by the lateness and uncertainty of the election date.51

To summarize, a main focus of civic-education- related measures should be placed on promoting the habit of attentiveness to political information. Courses, as Delli Carpini and Keeter insist, should be “taught in a realistic manner, introducing students to the conflictual, often unsettling nature of politics” (1996, 279). They must develop opportunities to engage in political realities, including “partisanship, without advancing one side or the other” (Beem 2005, 7). In so doing, of course, full use must be made of the most up-to-date channels of communication, electronic and otherwise — those that best fit the reading, listening and viewing habits of the emerging generation. The key is for the students to develop the habit of keeping up with political events so that they will continue to do so after they leave school.

Lowering the voting age and fixing the voting date

Mock elections cannot replace real elections. Bringing more high-school students to real ballot boxes entails lowering the voting age, a controversial reform that should be placed in the appropriate institutional con- text. We have already discussed the electoral system. With regard to civic education, PR elections are con- ducive to fostering political attentiveness, since they give small parties that have distinct principle-based positions and that carry some measure of popular support, such as the Green Party, a better chance of having democratically elected — and therefore legiti- mate — representatives. This representativeness can also give the entire political system more legitimacy in the eyes of young people, while the partisanship fos- tered by a PR environment — compared to the vola- tility, ideological incoherence and thus weak party identification under majoritarian systems — can be expected to have a positive indirect effect. For one thing, we know that parental partisanship boosts the political participation of young people, especially those still under their parents’ roofs (Plutzer 2002).

It is in this context that we address the question of lowering the voting age. The IDEA report evoked two controversial possibilities. The first is making voting compulsory, which I do not address here, since it raises ethical issues that require a lengthy treatment beyond the scope of this paper.52 The other is reducing the age of eligibility to 16,53 an idea that tends to be dismissed out of hand, since we instinctively assume that if young people don’t vote at 18, they are hardly likely to vote at 16. Yet the idea merits further reflection. As noted, paying attention to the political world and thus being sufficiently informed to vote when an election is called is mainly a matter of habit. As shown by Plutzer (2002), the costs of learning to vote are higher if one’s first election occurs during early adulthood, a time when one is only starting to establish the social net- works that will frame future choices, including political choices. If, in the first few years they are eligible to vote, potential voters are preoccupied with things other than politics and public affairs, they are more likely to develop the habit of not voting. People aged 18 to 20 are typically in a period of transition; they are with- drawing from their home and school environments, but they have not yet settled into a new environment.

Franklin provides evidence of a secular decline in turnout after the minimum voting age was reduced, typically to 18, in different countries during the 1960s and 1970s (2004). He maintains that this is due to a cer- tain number of individuals becoming socialized into nonvoting behaviour. Most became voters later in life, but some did not. Of those who did not, some would have done so had their first opportunity to vote occurred later, when they were in a better position to develop the habit. Given that raising the voting age to 21 again is politically unfeasible, Franklin proposes reducing it to 16, a less unsettled age. Since parents can more easily set an example for 16 year olds than for 18 year olds, reducing the age of eligibility to 16 should get more young people to the polls. While this con- tention is intriguing, it is far from proven. In my view, while increased parental influence cannot be dis- counted, to have a real potential of fulfilling Franklin’s expectations, lowering the voting age would need to be complemented and reinforced through civic educa- tion in the school setting. In Canada, this would entail taking advantage of the fact that many potential school and political dropouts would still be in school when their first opportunity to vote arose.

Finally, a related institutional change would be a fixed voting day, already instituted in British Columbia and on its way to being instituted in Ontario and, probably, New Brunswick. While minor, the effect would clearly be positive, since it would allow those initiating civic education courses, mock elections and other activities that encourage youth voting to plan their programs well in advance.54 The fixed voting date clearly facilitated the latest election simulation, Student Vote BC, which took place in 350 BC schools in the spring of 2005.

Beyond the Political

Before concluding, we need to be reminded of the modesty of our objectives. The phenome- non of political dropouts transcends the politi- cal; it has roots in socio-economic conditions and policies — matters well beyond the scope of this paper. Specifically, socio-economic changes have contributed significantly to transforming nonvoting from a life- cycle to a generational phenomenon. The labour mar- ket in contemporary societies has effectively excluded from secure employment a large number of young people, especially young men, who lack the necessary levels of literacy and numeracy. To put it baldly, these people fail to act as political citizens (vote, or pay any attention to politics) because they have been excluded from social citizenship.55 They lack what their counter- parts in the 1950s and 1960s had — namely, the eco- nomic and educational wherewithal to be full citizens, secure in their capacity to support their families and communities. Though this is a phenomenon that tran- scends national borders, the countries least affected are those that have developed the institutional and policy- adjustment mechanisms to overcome entrenched inequality. This is reflected in the data from the European Social Survey. Looking at the bottom line of table 1, we note that the average reported overall turnout was 80.3 percent, dropping to a worrisome 52.7 percent for those under 25.56 Limiting ourselves to Europe, and leaving aside the new democracies of Eastern Europe and countries with compulsory or quasi-compulsory voting, we find that both absolutely and in comparison to voting among older citizens, voting among young Danes, Swedes, Dutch, Germans and Austrians is high, while among British, Irish, Swiss, Spanish and Portuguese youth it is low. A sys- tematic analysis of these and similar figures would show,57 I contend, that differences in the level of social and economic security emerging generations can expect help explain variations in the number of politi- cal dropouts.

I have earlier argued that high-civic-literacy soci- eties can be distinguished from low-civic-literacy societies by policies aimed at the redistribution of resources that are both material and “nonmaterial,” the latter taking the form of measures enhancing access to knowledge, including political knowledge (Milner 2002, 13). The more laissez-faire policy stance of the main English-speaking mature democ- racies with regard to both material and nonmaterial redistribution — and, consequently, their relatively low level of civic literacy — stands in contrast to the more active approach of the high-civic-literacy northern European/Scandinavian countries. More recent work shows political knowledge to be more dependent on formal education in countries where income is less equally distributed (Milner and Grönlund 2004). Clearly, the costs to societies that do not meet the challenge of declining turnout will be heaviest for those least able to pay. In a democratic society, “les absents ont toujours tort.” By excluding those with low resources from informed political par- ticipation, we make it less likely that the policies that can help them attain access to those resources will be implemented, and the result of this will be further abstention on their part, and so on — a classic vicious circle (Milner 2002, part IV). Conversely, nonmaterial redistribution — achieved by fostering informed par- ticipation among those low in resources — promotes policy choices leading to more material redistribu- tion, giving rise to a “virtuous circle,” since such out- comes encourage citizens to keep well informed of governmental decisions. Thus, the cycle begins again.

Conclusion: Where Do We Go from Here?

Since there is much we still do not know in order to explain adequately the phenomenon of political drop-outs, our most immediate conclusion is an insistent call for further comparative research using surveys that include a battery of political- knowledge questions, which would enable us to dis- tinguish the politically uninformed young nonvoters from the political protestors and thus select and test measures aimed at the former.58

In selecting specific measures, we must be tentative. Indeed, the application of these measures should be viewed as a pilot project subject to further development and refinement. The Canadian provinces are an ideal venue for such projects. And it is the provinces, rather than the federal government, that have primary consti- tutional jurisdiction over actions in this field. The rec- ommendations that follow are offered with a view to their being undertaken by the provinces, supported, where appropriate, by the relevant federal bodies. At the core of these recommendations is the intention of targeting those close to voting age with a civic educa- tion program focused on bringing the issues — and the political parties59 — into the classroom and giving these young people, and their parents, the chance to vote under a new electoral system in which political parties and their supporters can expect to be represented fairly.

Recommendations

Political institutional reforms

The general rule is that electoral systems and comple- mentary rules and regulations concerning media access, party financing, information dissemination, (fixed) election dates and access to the ballot box must be designed to ensure, and to allow citizens and actors to expect, that legitimate political positions are given expression and representation in the various democratic institutions — from the local to the nation- al and beyond — at a level approximating their sup- port in the population.

Functional literacy

The acquisition of the skills and habits of literacy (and numeracy), including media literacy, must be prioritized throughout the educational system (including adult education programs). This should be reinforced by appropriate programs of access to libraries, publica- tions, high-speed Internet and so on.60 The goal is for all those intellectually capable to acquire the skills and knowledge to take their places as full citizens and to contribute to family and community. One aspect of this is the objective of having practically all young people at school when they reach the age of citizenship.61 Fewer school dropouts means fewer political dropouts.

Civic education stressing informed choice among political options and a lower voting age

Civic education should be compulsory in the year or two before the age of citizenship is attained. Since 16- and 17- year-olds are about to be called upon to vote, this is appropriate, and likely more effective than the current arrangement. A major goal is to instill the habit of being attentive to politics and public policy, new and old, as reported through the media. A crucial component of civics courses is the information attained through contact with the relevant political actors. Thus, importance should be accorded to the positions on relevant issues taken by the different parties at the local, regional and national levels. Party representatives should be invited into the classroom both physically and electronically. (This is a more natural and applicable option within the PR framework, where the various parties have a legitimate and relatively stable political presence at each level.) In this way, the wall between political life and “real” life that serves to justify political abstention is removed — a wall that is especially high and strong in the United States.

But, as noted, given school dropout rates, few if any Canadian provinces will reach a large enough propor- tion of young people in this manner. Hence, I propose that one of them undertake as a pilot project a combi- nation of the recommendations made here with lower- ing the voting age to 16 (the courses would thus be offered to 15- and 16-year-olds). The change in voting age could be temporary, subject to renewal after an assessment of its effects — after, say, two elections. A possible transitional measure would be to tie the right to vote at age 16 to successful completion of the civics course, an approach somewhat similar to that applied to the acquisition of a driver’s license. While such an approach would likely raise questions — and charter challenges — related to age discrimination if applied generally, it might be acceptable as a pilot project to test the effects of a wider application of the measures.

Would these measures have any significant effect? The evidence suggests that, if well carried out, they might indeed. But there are no guarantees. Still, I am not convinced that youth abstention is an expression of the “good judgment of young people on the failings of political elites,” any more than I accept that “democ- racy is best served if the ignorant abstain.” Hence, I conclude quite simply that the time to act is now.

  1. An analysis of turnout in 20 countries found an aver- age decline of 5 percent, from 83 percent in the 1950s to 78 percent in the 1990s (Dalton 1996, 44-5; see also Wattenberg 1998).

  2. International IDEA, the Institute for Democracy and Electoral Assistance, based in Stockholm, calculates the turnout based on both methods (IDEA 2004, 131). For Canada, IDEA estimates an 8 percent difference between the two due to unregistered potential voters. Applied to the 2000 election turnout figure (the most recent calculated by IDEA), turnout among potential voters was 54.6 percent.

  3. For an analysis of turnout in the 2004 US election, see McDonald (2004).

  4. This is partly explained by the late date of the 2004 election; it was held on June 28, when many Canadians were getting into a vacation mood.

  5. In Finland, nonvoting has increased most markedly among the young: the turnout gap between those 19 to 24 years of age and the overall average rose to 17 percentage points in 1999 (Martikainen 2000). For Norway, Bjorklund finds particular effects of age upon turnout in local elections, in whose level there has been an especially worrying decline (2000). In 1999, only 31 percent of those born after 1975 voted, a num- ber rising steadily by age cohort to 72 percent for those born between 1930 and 1945.

  6. In 1999, an International IDEA report examining polit- ical participation among young people in 15 Western European countries found that while the young have tended to vote less than their elders, by the early 1990s the turnout gap between citizens 18 to 29 and those over 30 had grown to 12 percent.

  7. Moreover, since it was presumed that, as in all survey- based studies, the rate of voter turnout was over- reported, the researchers used statistical corrections on the rate of turnout for the different age groups. It may very well be, however, that young people, given the low sense of civic duty to vote, do not overreport.

  8. The turnout rate of 18-to-24-year-old voters rose from 36.6 percent in 2000 to 42.3 in 2004: the 5.7 percent difference corresponds to the overall rise from about 54 to 60 percent. Judging by the increased support among young people for the Democrat candidate, that increase is attributable to intense mobilization efforts, especially in the 15 swing states. Many of these newly mobilized young citizens were disillusioned by the result (Center for Information and Research on Civic Learning and Engagement 2004), so it remains to be seen whether it signals an upward trend or just a blip. One indication that the latter is the case is the finding in a study conducted in the month before the election that voter registration rose only in the battleground states (McDonald 2004).

  9. In the 2001 UK election, overall turnout sank to a postwar low of just 59 percent, with only 39 percent of young people casting a vote (UK Electoral Commission 2001).

  10. The same study by Elections Canada found that the 57-plus age group’s turnout was 35 points higher than that of the 18 to 21 year olds.

  11. A Canadian example of this making-a-virtue-of-necessity approach can be found in a study carried out with the support of Canadian Heritage that suggested, among other things, that participation should be widely defined to include media-consumer choices, such as casting votes for Canadian Idol (Smith and Barnard 2003).

  12. Political scientists tend to underestimate the impor- tance and effect of low political knowledge, since, as individuals, they are by definition knowledgeable about politics.

  13. More generally, as cogently argued by Althaus, by not incorporating the political-knowledge dimension into attitudinal surveys on politically related issues we are failing to take into account the quality of the opinions expressed, and thus failing to distinguish meaningful opinions held by respondents from artifacts of the interview process (2003). An example of the need for political-knowledge data to distinguish between the informed and uninformed abstainer lies in the emerg- ing literature testing whether higher turnout would have altered the outcome of elections and finding, typ- ically, that it would not have done so (see, for exam- ple, Marsh and Bernhagen 2004). The problem is that such analyses include knowledgeable abstainers who do not bother to vote because they can see that their vote cannot affect the outcome. Proving that they are right is of little interest. It is only when we are able to separately analyze what would happen if those who abstained out of lack of information actually voted that the question becomes interesting — but in order to do so, we must incorporate the political-knowledge dimension in our data.

  14. “Young people rated their interest in politics at only 4.5 on a 0 to 10 scale (where zero indicated no interest at all), compared to 7.5 for those in their sixties and up…It is hard to cast an informed ballot if you do not know who the potential prime ministers are or what their parties are promising” (Gidengil et al. 2003, 11).

  15. This is from a survey of 1,500 Americans between the ages of 15 and 25 commissioned by the Council for Excellence in Government’s Center for Democracy and Citizenship and the Partnership for Trust in Government in cooperation with the Center for Information and Research on Civic Learning and Engagement (CIRCLE 2004).

  16. I wish to thank Svante Ersson of Umeå University for these calculations. There was significantly less varia- tion between age groups in the new ESS democracies than in the old ones. An example of the latter is Norway, where Bjorklund signals a “dwindling support for voting as a form of civic virtue…The difference between cohorts is pronounced. It is the youngest cohort that most often sticks to the [voting as] self- interest alternative” (Bjorklund 2000, 19).

  17. In the 1956 Gallup Poll, respondents were shown a list of 10 prominent political figures, of which 2 were Canadian, and asked to identify the country and posi- tion of each; they were also shown a list of Canada’s 10 provincial premiers and asked to identify their provinces. The 2000 CES included an unprecedented number of knowledge items: the names of the leaders of the Liberal, PC, Alliance and New Democratic Parties; the name of the federal finance minister; and the name of the respondent’s provincial premier.

  18. Holding education constant, Grönlund found that for those with less than a complete secondary education, the average score on the three or more CSES political- knowledge questions was .40 for the 18- to 35-year- olds, compared to just under .50 for the 34- to 55- year-olds, and .53 for those 55 and over. For those who completed secondary or vocational school, the disparity was essentially the same (with the score of the youngest group rising to .53). Only when we get to those who completed university is the disparity reduced — by roughly half — with the youngest group averaging .65 right answers (Grönlund 2003).

  19. For example, Chiche and Haegel show that 18- to-29- year-old French men and women are over 10 percent less politically knowledgeable than those over 30 (2002, 280). Rose tested the knowledge of local politi- cal actors, institutions and policies in Denmark and Norway, finding that, “Given differences in education- al levels that exist among younger and older age cohorts in both countries, however, it is every bit as remarkable to note that age is consistently related to political knowledge, even after educational differences are held constant” (Rose 2002, 6).

  20. In response to three questions posed by the NASS Millennium Project, 79 percent of respondents were able to give the name of the US vice-president, 67 per- cent could name their governor, and only 37 percent knew the term length for a member of the US House of Representatives (Parker and Deane 1997).

  21. In 1990 they were asked, “Who is the prime minister?” “Who is the Liberal leader?” “Who is the NDP leader?” In 2000 their task was to identify the prime minister, finance minister and official opposition party.

  22. I exclude the results of the Civic Education Study, which tested nearly 90,000 14-year-old students in 28 coun- tries and 50,000 17 to 19 year olds in 16 countries on political knowledge, skills and attitudes (Torney-Purta et al. 2001). Not only was Canada not included, but the questions are problematic since they do not allow for a clear comparative assessment of political knowledge. Instead, they test understanding of the logic of democ- racy and of the functions of institutions in democracy.

  23. Examples of such questions include: “The Taliban and al-Qaeda movements were both based in which coun- try?” “Which of the following organizations endorses the euro as the common currency for its members?” “Which two countries have had a long-standing con- flict over the region of Kashmir?” (National Geographic Foundation 2002).

  24. Estimates based on earlier data were higher: for Lijphart, it was about 9 percent (1997), a difference similar to that found by International IDEA in its report Voter Turnout from 1945-1997 (1997), which used vot- ing-age populations rather than registered voters (see also Ladner and Milner 1999).

  25. Note that there are a number of CSES countries missing from figure 1. This is because they held no, or very few, democratic general elections during the years in ques- tion. The countries are scattered along a more or less lin- ear pattern, with Belgium the most significant outlier due to the significantly higher dispersal effects of educa- tion there than in countries with similar degrees of pro- portionality. The complex binational character of Belgium’s consociational political arrangements may have something to do with this. (The relationship is highly significant statistically when Belgium is excluded, but far less so when it is included.)

  26. For presidential elections, the drop was lower: from 85.2 percent in 1988 to 76.8 percent in 2000.

  27. A new law allows local authorities to run elections under single transferable vote (STV) as well as FPTP. For the 2004 municipal elections, only 10 chose the former (see New Zealand Department of Internal Affairs 2003).

  28. Karp and Banducci show that this factor helps explain the turnout spike in the first New Zealand MMP election in 1996 (1998).

  29. Canadian data shows that young people, though less attentive and informed, are in fact more supportive of “politics as usual” than older Canadians. O’Neill found 18 to 27 year olds to be roughly 10 percent more satis- fied with Canadian democracy and elections than mem- bers of other age groups and, comparatively, even more willing to view the federal government as fair and effec- tive. Nor are they less distrustful of multinational corpo- rations than older Canadians (O’Neill 2001).

  30. In Britain, young people who did not vote in 2001 were more likely than all other nonvoters to believe that the act of voting was a meaningless or insignificant one, and that nothing would change, whatever the outcome (UK Electoral Commission 2002, 27).

  31. Pammett and Leduc report that 6.5 percent of respon- dents 18 to 24 gave the absence of a contest as their rea- son for not voting (2003, 17). While this proportion is small — even smaller than the overall 9 percent of abstainers who so responded — it is meaningful, given that abstainers constituted more than 75 percent of those 18 to 24 in their sample.

  32. In 2000, the voting rate of persons below the poverty line was estimated at about 25 percent, compared to 65 percent for those above it (Leighley and Nagler 2000, 1). In the period 1996-97, the turnout ratio between those who had completed university and those who hadn’t fin- ished high school was 7 to 5 in Canada and better than 9 to 4 in the United States (Martinez 2000, 219).

  33. Usually, a teacher places bulk orders for newspapers to be delivered to the school on a schedule that accords with the lesson plan. The teacher distributes the newspapers to the students and uses them in class as a teaching tool. For a study based on the application of such a program in Argentina, see Morduchowitz and Galperin (1998).

  34. A useful series of articles on the subject is found in a 2001 special issue of the National Civic Review titled “Making Citizen Democracy Work” (90, no. 3).

  35. While the overall effect is small, the authors note that “the biggest effect comes from mailing sample ballots, which most influences young people with less access to information from other sources” (Wolfinger et al. 2004, 13). Another study provided teenagers with an interactive CD campaign handbook with encouraging results (Iyengar and Jackman 2004).

  36. It is now possible to hire a private Internet-based com- pany that, using your electronic responses to a ques- tionnaire, will determine which candidate/party for every possible election most suits you.

  37. Finkel finds a positive effect of adult civic education in South Africa and the Dominican Republic (2002).

  38. The civics course practically all German students take one hour per week from grades 7 or 8 seems to have had little effect (Händle et al. 1999); in the Dutch case, there was a correlation only for the less than 10 per- cent of students who took the civics course (called “society”) as part of the formal program leading to the final examination (Hahn 1998, 15). Dekker suggests that its effects are likely to prove short-lived (1999). See also Dekker and Portengen (2000).

  39. In Australia, Hugh MacKay concludes, “typically, teenagers find little to interest or inspire them in the political process, and they often report that politics is the most boring subject discussed at home” (quoted in Civics Expert Group 1994, 182).

  40. In Sweden, only 2 percent leave school at the end of compulsory schooling, at age 16 (Skolverket, 1998).

  41. Westholm, Lindquist and Niemi found that upper-sec- ondary students taking civics courses were more likely to retain knowledge about international organizations (11 percent more) and international events (6 percent more) when retested two years later than those in a control group (1989).

  42. One example is a program instituted in the 1990s in civics classes in the upper-secondary schools of the northern Swedish city of Umeå, where I teach. In order to provide a bridge from the classroom to political organization and activity outside the school, thus encouraging political participation, representatives of the local units of political parties are invited to explain their programs and describe their actions. Another example is how, in advance of the 2003 monetary ref- erendum, spokespeople for both sides were systemati- cally invited to civics classes to present their cases on adopting or not adopting the euro.

  43. “Schools [in Wisconsin] feared being charged with being partisan. Having students deliver food baskets was safe; having students work to oust a politician who cut food-stamp programs was not” (Beem 2005, 10). And they were right in their fears. The Corporation for National Service, a major funder of service learning, explained its refusal to allow participants in the youth service program Americorps to attend the “Stand for Children” rally in Washington, DC: “National Service has to be non-partisan…it should be about bringing communities together by getting things done. Strikes, demonstrations and political activities can have the opposite effect” (quoted in Walker 2000; emphasis added). Hence it comes as no surprise that in a recent article comparing the scores of 14-year-old American students on the IEA Civic Education study with the mean of the 28 participating countries, the American students did worst on the question about the function of political parties (Torney-Purta and Barber 2004).

  44. Typically, as college students complained to researchers in one recent study, “they had received much more encouragement and opportunities to get involved in service, but hardly any [to go] into politics.” This was the conclusion of a weekend exchange sponsored by the Johnson Foundation between 20 representative Wisconsin college students and politicians. The author of a report on the conference concludes that for stu- dents, “while there are ample and readily accessible opportunities for community service, they do not know how to find out who their assemblyman [is], or how to get involved in a campaign, or even how to register to vote…it was as if a light was supposed to go off when someone turned eighteen” (Beem 2005, 10).

  45. An impressive election-simulation game based on the presidential primaries was run at Townsend Harris High School and sponsored by the Taft Institute for Government and Queens College.

  46. Campbell’s study was based on data from the IEA Civic Education Study. Ironically — if not surprisingly — he found that students attending racially diverse schools were less likely to report open classrooms, suggesting that discussions of diverse or controversial opinions are more likely to be encouraged in racially homoge- neous classrooms (2005).

  47. The chief electoral officer reported, “We also developed a series of outreach initiatives for young people…Community relations officers for youth identi- fied neighbourhoods with high concentrations of stu- dents for special registration drives, assisted in locating polls in places easily accessible to youth, and informed the community and youth leaders about reg- istration and voting. The redesigned ‘Young Voters’ section of the Elections Canada Web site, which offered information on the electoral process, was visit- ed more than 103,000 times during the election period” (Elections Canada 2003).

  48. Go to http://www.elections.ca/content_youth.asp?sec- tion=yth&document=index&lang=e&textonly=false

  49. The Dominion Institute also sponsored Youth Text 2004, which enabled people to engage in a dialogue about political participation. Those registering to receive election updates were eligible to enter a draw for one of about 60 Nokia handsets.

  50. The program appears to enhance the attentiveness of the students to politics in the media and at home. It also makes parents better informed about politics, through their children, leading them to vote more often (Golston 1997). Other research found that it sharpens students’ critical thinking and narrows the gender and socio-economic gap in civic education (Kids Voting USA).

  51. Quebec schools were underrepresented, which explains why the Bloc Québécois received a low 3.1 percent and the Liberals received 29.2 instead of 36.7 percent (Verboczy and Giguère 2004). Student Vote 2004 con- ducted surveys across Canada to test the impact of the simulation. The presurvey, completed prior to partici- pation in the program, had 14,344 responses, but only 2,841 responded to the post-election survey. The change is nonetheless great enough to suggest that the program had some effect: while in the presurvey 71 percent said they would vote if they had the oppor- tunity, this rose to 88 percent among those who had participated in the program (Student Vote 2004).

  52. To favour it, I would need to be persuaded that com- pulsory voting does more than make inattentive and uninterested young people commit an act they find meaningless.

  53. The voting age has been lowered in six German states (Aarts and van Hees 2003).

  54. I shall address this question in a future IRPP paper.

  55. A useful analysis of this group’s abstention from voting is found in Lyons and Sinnott (2003).

  56. Since many were too young to vote in the last election, the is quite small.

  57. The 1999 International IDEA youth voter participation study, which was based on figures from the mid-1990s, identifies the same patterns.

  58. I have been working to include such a battery in the third round (2006) of the European Social Survey. If it is included, it will then be crucial to conduct a similar survey in Canada.

  59. There is much that political parties could do to encour- age the involvement of young people, facilitated by a PR environment. But that is beyond the scope of this paper, which focuses on choices of policy-makers.

  60. A comprehensive survey conducted in 20 countries in the 1990s revealed that less than 10 percent of adults in the four Nordic countries had literacy skills below those needed to function in today’s world, compared to over 20 percent in the US and the UK, and 18 percent in Canada (Organisation for Economic Co-operation and Development 1997; 2000).

  61. One useful approach developed especially in Germany entails dual programs linking educational institutions with on-the-job training.

Aarts, Kees, and Charlotte van Hees. 2003. “Lowering the Voting Age: European Debates and Experiences: Youth Participation in Canada.” Elections Canada, Electoral Insight, July.

Aldrich, John. 1993. “Rational Choice and Turnout.” American Journal of Political Science 37: 46-78.

Althaus, Scott. 2003. Collective Preferences in Democratic Politics. Cambridge: Cambridge University Press.

Andolina, Molly, Scott Keeter, Cliff Zukin, Krista Jenkins. 2002. The Civic and Political Health of the Nation: A Generational Portrait. College Park, MD: National Alliance for Civic Education.

Beem, Christopher. 2005. “From the Horse’s Mouth: A Dialogue between Politicians and College Students.” CIRCLE Working Paper 27. Accessed April 28, 2005. http://www.civicyouth.org/PopUps/WorkingPapers/WP 27Beem.pdf

Bennett, Stephen E. 1998. “Young Americans’ Indifference to Media Coverage of Public Affairs.” PS: Political Science and Politics 31, no. 3: 535-41.

Bjorklund, Tor. 2000. “The Steadily Declining Voter Turnout: Norwegian Local Elections 1963-1999.” Paper presented at the conference of Della Societa Italiana di Studi Elettorali, October 19-21, Naples, Italy.

Blais, André. 2000. To Vote or Not to Vote: The Merits and Limits of Rational Choice Theory. Pittsburgh: University of Pittsburgh Press.

Blais, André, Elizabeth Gidengil, Richard Nadeau, and Neil Nevitte. 2002. Anatomy of a Liberal Victory: Making Sense of the Vote in the 2000 Canadian Election. Peterborough, ON: Broadview.

Bowler, Shaun, Elisabeth Carter, and David M. Farrell. 2000. “Studying Electoral Institutions and Their Consequences: Electoral Systems and Electoral Laws.” Paper presented at the conference of the International Political Science Association, August,Quebec City.

Bronner, Ethan. 1998. “College Freshmen Aiming for High Marks in Income: Developing a Philosophy of Life Can Wait.” The New York Times. January 12, A14.

Campbell, David. 2005. “Voice in the Classroom.” CIRCLE Working Paper 28. Accessed April 28, 2005. http://www.civicyouth.org/PopUps/WorkingPapers/WP 28campbell.pdf

Center for Information and Research on Civic Learning and Engagement (CIRCLE). 2004. CIRCLE Update, November 9.

——. “High School Civic Education.” Accessed April 28, 2005. http://www.civicyouth.org/research/areas/hs_civic.htm

Centre for Comparative Social Surveys. 2003. European Social Survey. Accessed May 27, 2005. www.euro- peansocialsurvey.org

Chiche, Jean, and Florence Haegel. 2002. “Les connaissance politiques.” In La démocratie à l’épreuve, edited by G. Grunberg, N. Mayer and P. Sniderman. Paris: Presses de Sciences Po.

Civics Expert Group. 1994. Whereas the People…Civics and Citizenship Education. Canberra: Australian Government Publishing Service.

Dalton, Russell J. 1996. Citizen Politics: Public Opinion and Political Parties in Advanced Western Democracies. Chatham, NJ: Chatham House.

Dekker, Henk. 1999. “Citizenship Conceptions and Competencies in the Subject Matter ‘Society’ in Dutch Schools.” In Civic Education across Countries: Twenty- four National Case Studies from the IEA Civic Education Project, edited by J. Torney-Purta, J. Schwille, and J.A. Amadeo. Amsterdam: Eburon-IEA.

Dekker, Henk, and R. Portengen. 2000. “Political Knowledge: Theories and Empirical Research.” In Democracies in Transition: Political Culture and Socialization Transformed in West and East, edited by Russell F. Farnen, Henk Dekker, Daniel B. German, and Rüdiger Meyenberg. Oldenburg: BIS-Verlag.

Delli Carpini, Michael, and Scott Keeter. 1996. What Americans Know about Politics and Why It Matters. New Haven: Yale University Press.

Elections Canada. 2005. Electoral Insight, February.

——. 2003. Electoral Insight, July.

Ersson, Svante, and Henry Milner. Forthcoming. “Political Participation in Scandinavia.” In The Nordic Countries in the New Europe: Institutions, Politics and Policy, edited by Gunnel Gustafsson, Cynthia Kite, and Kjell Lundmark. Oslo: Universitetsforlaget.

Farrell, David. 2001. Electoral Systems: A Comparative Introduction. London: Palgrave.

Finkel, Steven E. 2002. “Civic Education and the Mobilization of Political Participation in Developing Democracies.” Journal of Politics 64, no. 4: 994-1020.

Franklin, Mark N. 2004. Voter Turnout and the Dynamics of Electoral Competition in Established Democracies since 1945. Cambridge: Cambridge University Press.

Gidengil, Elisabeth, André Blais, Joanna Everitt, Patrick Fournier, and Neil Nevitte. 2004. “Why Johnny Won’t Vote.” The Globe and Mail, August 4.

——. 2005. “Missing the Message: Young Adults and the Election Issues.” Elections Canada, Electoral Insight, February 2005.

Gidengil, Elisabeth, André Blais, Neil Nevitte, and Richard Nadeau. 2003. “Turned off or Tuned out: Youth Participation in Canada.” Elections Canada, Electoral Insight, July.

Golston, Syd. 1997. “Never Too Young: Kids Voting USA.” Civnet Journal 1, no. 2.

Grönlund, Kimmo. 2003. “Knowledge and Turnout: A Comparative Analysis.” Paper presented at the confer- ence of the European Consortium on Political Research, September, Marburg, Germany.

Hahn Carole L., ed. 1998. Becoming Political: Comparative Perspectives on Citizenship Education. Albany: State University of New York Press.

Händle Christa, Detlef Oesterreich, and Luitgard Trommer. 1999. “Concepts of Civic Education in Germany Based on a Survey of Expert Opinion.” In Civic Education across Countries: Twenty-four National Case Studies from the IEA Civic Education Project, edited by J. Torney-Purta, J. Schwille, and J.A. Amadeo. Amsterdam: Eburon-IEA.

Hébert, Yvonne. 2002. Citizenship Transformation in Canada. Toronto: University of Toronto Press.

Henn, Matt, and Mark Weinstein. 2003. “First-Time Voters’ Attitudes towards Party Politics in Britain.” Nottingham Trent University, Economic and Social Research Council, March. Accessed April 28, 2005. http://www.research-tv.com/stories/society/ virginvoters/briefing.doc

Henzey, Debra. 2003. “North Carolina Civic Index: A Benchmark Study Impacts State-Level Policy.” National Civic Review 92, no. 4: 27-36.

Holmberg, Soren, and Henrik Oscarsson. 2004. Swedish Voting Behaviour. Gothenberg: Swedish Election Studies Programme.

Howe, Paul. 2001. “The Sources of Campaign Intemperance.” IRPP Policy Options, January- February: 21-4. Montreal: IRPP.

——. 2003. “Where Have All the Voters Gone.” Inroads 12: 75-82.

Hunter, Susan, and R.A Brisbin Jr. 2000. “The Impact of Service Learning on Democratic and Civic Values.” PS: Political Science and Politics 33, no. 3: 623-6.

International Institute for Democracy and Electoral Assistance (IDEA). 1997. Voter Turnout from 1945- 1997: A Global Report on Political Participation. Stockholm: IDEA.

——. 1999. Youth Voter Participation: Involving Today’s Young in Tomorrow’s Democracy. Stockholm: IDEA.

——. 2004. Voter Turnout since 1945: A Global Report. Stockholm: IDEA.

Iyengar, Shato, and Simon Jackman. 2004. “Technology and Politics: Incentives for Youth Participation.” Center for Information and Research on Civic Learning and Engagement (CIRCLE). Accessed April 28, 2005. http://www.civicyouth.org/

Karp, Jeffrey A., and Susan A. Banducci. 1998. “The Impact of Proportional Representation on Turnout: Evidence from New Zealand.” Paper presented at the annual meeting of the Australasian Political Science Association, Christchurch.

——. 2001. “Absentee Voting, Participation, and Mobilization.” American Politics Research 29, no. 2: 183-95.

Keeter, Scott, Cliff Zukin, Molly Andolina, and Krista Jenkins. 2002. The Civic and Political Health of the Nation: A Generational Portrait. College Park, MD: Center for Information and Research on Civic Learning and Engagement (CIRCLE); Pew Charitable Trusts.

Kids Voting USA. “Research Summary.” Accessed April 28, 2005. http://www.kidsvotingusa.org/shared/Research Summary6-04.pdf

Ladner, Andreas, and Henry Milner. 1999. “Politicization, Electoral Institutions, and Voting Turnout: The Evidence from Swiss Communal Elections in Comparative Context.” Electoral Studies 18, no. 2: 235-50.

Leighley, Jan, and Jonathan Nagler. 2000. “Socioeconomic Class Bias in Turnout: Evidence from Aggregate Data.” Paper presented at the annual meeting of the American Political Science Association, Washington, DC.

Lijphart, Arend. 1997. “Unequal Participation: Democracy’s Unresolved Dilemma.” American Political Science Review 91, no. 1: 1-14.

——. 1999. Patterns of Democracy: Government Forms and Performances in Thirty-Six Countries. New Haven: Yale University Press.

Lopez, Mark H. 2003. “Civic and Social Studies Course Taking and Civic Engagement.” Paper presented at the International Conference on Civic Education Research, New Orleans, November.

Lyons, Pat, and Richard Sinnott. 2003. “Voter Turnout in 2002 and Beyond.” In How Ireland Voted 2002, edited by Michael Gallagher, Michael Marsh, and Paul Mitchell. Basingstoke: Palgrave.

“Making Citizen Democracy Work.” 2001. Special issue of National Civic Review 90, no. 3.

Market and Opinion Research International (MORI). 2001. “How Britain Voted in 2001.” British Public Opinion 24, nos. 3-4: 20-26.

Marsh, Michael, and Patrick Bernhagen. 2004. “Turnout Matters _ Sometimes.” Paper presented at the European Consortium on Political Research joint sessions, April, Uppsala.

Martikainen, Tuomo. 2000. “Disengagement from Politics. Part 1: Electoral Participation.” Paper presented at the Oslo workshop “Towards Spectator Democracy,” October 28-29, Helsinki.

Martinez, Michael. 2000. “Turning In or Tuning Out: Electoral Participation in Canada and the United States.” In Canada and the United States: Differences that Count, edited by David M. Thomas. Peterborough, ON: Broadview.

McDonald, Michael P. 2004. “Is Voter Registration up Everywhere in America?” Brookings Institution, Governance Studies, October 5. Accessed April 28, 2005. http://www.brookings.edu/views/papers/mcdon- ald/20041005.pdf

Milner, Henry. 2002. Civic Literacy: How Informed Citizens Make Democracy Work. Hanover, NH: University Press of New England.

Milner, Henry. 2004a. “First past the Post? Progress Report on Electoral Reform Initiatives in the Canadian Provinces.” IRPP Policy Matters 5, no. 9.

Milner, Henry, ed. 2004b. Steps toward Making Every Vote Count: Electoral System Reform in Canada and Its Provinces. Peterborough: Broadview.

Milner, Henry, and Kimmo Grönlund. 2004. “The Determinants of Political Knowledge in a Comparative Perspective.” Presented at the 2004 annual meeting of the American Political Science Association.

Morduchowitz, Roxana, and Hernan Galperin. 1998. “Education for Democracy in Argentina: Effects of a Newspaper In-School Program.” In Citizenship Education in a Changing World: International Perspectives, edited by Ort Ichilov. Essex: Frank Cass.

MORI. 2001. “How Britain Voted in 2001.” British Public Opinion 24, nos. 3-4: 3. Accessed May 19, 2005 http://www.mori.com/election2005/bpoge2001.pdf

National Geographic Education Foundation. 2002. National Geographic-Roper Global Geographic Literacy Survey. Accessed April 28, 2005. www.nationalgeographic.com/eosurvey/

New Brunswick Commission on Legislative Democracy. 2004. Final Report and Recommendations. Accessed April 28, 2005. http://www.gnb.ca/0100/FinalReport-e.pdf

New Zealand. Department of Internal Affairs. 2003. “STV Information.” Accessed April 28, 2005. http://www.dia.govt.nz/diawebsite.nsf/wpg_URL/Resource -material-STV-Information-Index?Open Document#four

Niemi Richard, and Jane Junn. 1998. Civic Education: What Makes Students Learn? New Haven: Yale University Press.

O’Neill, Brenda. 2001. “Generational Patterns in the Political Opinions and Behaviour of Canadians: Separating the Wheat from the Chaff.” IRPP Policy Matters 2, no. 5.

Organisation for Economic Co-operation and Development (OECD). 1997. Literacy Skills for the Knowledge Society: Further Results from the International Adult Literacy Survey. Paris: OECD.

——. 2000. Literacy in the Information Age: Final Report of the International Adult Literacy Survey. Paris: OECD.

O’Toole, Therese, David Marsh, and Su Jones. 2003. “Political Literacy Cuts Both Ways: The Politics of Non-participation among Young People.” Political Quarterly 74, no. 3: 349-60.

Pammett, Jon, and Larry Leduc. 2003. “Explaining the Turnout Decline in Canadian Federal Elections: A New Survey of Non-voters.” Ottawa: Elections Canada.

Parker, Kimberly, and Claudia Deane. 1997. “Ten Years of the Pew News Interest Index.” Paper presented at the annual meeting of the American Association for Public Opinion Research, Norfolk, VA.

Plutzer, Eric. 2002. “Becoming a Habitual Voter: Inertia, Resources, and Growth in Young Adulthood.” American Political Sciences Review 96, no. 1: 41-56.

Rose, Lawrence E. 2002. “Local Political Participation in Denmark and Norway: Does Information Make a Difference?” Paper presented at Workshop 22, “Political Participation and Information,” ECPR Turin, April.

Rosenstone, Steven, and J.M. Hansen. 1993. Mobilization, Participation and Democracy in America. New York: Macmillan.

Schwarz, Norbert, and Howard Schuman. 1997. “Political Knowledge, Attribution, and Inferred Interest in Politics: The Operation of Buffer Items.” International Journal of Public Opinion Research 9: 191-5.

Skolverket. 1998. Gymnasieskola för alla – andra. En studie om marginalisering och utslagning i gymnasieskolan. Stockholm: Skolverket.

Smith, Shelley, and Robert Barnard. 2003. “Understanding Active Citizen Engagement among Canada’s Information Age Generations.” Accessed April 28, 2005. http://www.d-code.com/citizenregen.html

Strate, John M, Charles J Parrish, Charles D Elder, and Coit Ford III. 1989. “Life Span Civic Development and Voting Participation. The American Political Science Review. 83: 443-64.

Stroupe, Kenneth S., and Larry J. Sabato. 2004. “Politics: The Missing Link of Responsible Civic Education.” CIR-CLE Working Paper 18. Accessed April 28, 2005. http://www.civicyouth.org/PopUps/WorkingPapers/WP 18Stroupe.pdf

Student Vote 2004. “Post-election Report.” Accessed April 28, 2005. http://www.studentvote.ca/ uploads/PostelectionReportSV2004.pdf

Times Mirror Center for the People and the Press. 1990. “The Age of Indifference: A Study of Young Americans and How They View the News.” Washington, DC: Times Mirror Center.

Torney-Purta, Judith, and Carolyn Henry Barber. 2004. “Strengths and Weaknesses in U.S. Students’ Knowledge and Skills: Analysis from the IEA Civic Education Study.” Fact Sheet, The Center for Information and Research on Civic Learning and Engagement (CIRCLE), University of Maryland. June. Accessed May 27, 2005. http://www.civicyouth.org/ PopUps/ FactSheets/FS_CivicKnowledge.pdf

Torney-Purta, Judith, Rainer Lehmann, Hans Oswald, and Wolfram Schulz. 2001. Citizenship and Education in Twenty-eight Countries: Civic Knowledge at Age Fourteen. Amsterdam: Eburon-IEA.

United Kingdom. Electoral Commission. 2002. Voter Engagement and Young People. London: Electoral Commission.

Verboczy, Akos, and Christian Giguère. 2004. “Électeurs en Herbe: Building Citizenship Step by Step. A Quebec Approach to Civic Education.” Inroads 16: 84-9.

Walker, Tobi. 2000. “The Service/Politics Split: Rethinking Service to Teach Political Engagement.” PS: Political Science and Politics 33, no. 3: 646-9.

Wattenberg, Martin P. 1998. “Turnout Decline in the US and Other Advanced Industrial Democracies.” University of California, Irvine: Center for the Study of Democracy.

Westholm, Anders, Arne Lindquist, and Richard Niemi. 1989. “Education and the Making of the Informed Citizen: Political Literacy and the Outside World.” In Political Socialization, Education and Democracy, edit- ed by Ort Ichilov. New York: Teachers’ College Press.

World Association of Newspapers. 2004. World Press Trends. Paris.

Wolfinger, Raymond E., Benjamin Highton, and Megan Mullin. 2004. “How Postregistration Laws Affect the Turnout of Registrants.” College Park, MD: National Alliance for Civic Education.

Wolfinger, Raymond E., and Steven J. Rosenstone. 1980. Who Votes? New Haven: Yale University Press.