Generational Patterns in the Political Opinions and Behaviour of Canadians
Separating the Wheat from the ChaffOctober 24th, 2001
This paper tackles questions of generational differences among Canadians in their political attitudes and behaviours. Drawing on survey data collected by IRPP, it presents a detailed analysis of opinions toward many topical subjects, ranging from the influence of money in politics, to the role of minorities in elected bodies, to the general health of the Canadian democracy. Standard assumptions of inter-generational differences – heightened apathy among young voters, increasing disengagement from politics from one generation to the next – are put to test.
The principal conclusions include the following:
In keeping with expectations, younger Canadians are less likely to follow politics, to be politically knowledgeable, or to participate in politics at least by voting or joining a political party.
Somewhat surprisingly, younger Canadians reveal higher levels of satis- faction with a number of institutions, including Canadian democracy and elections generally and they hold healthier opinions of the federal gov- ernment, in that they are more likely to see it as fair and effective.
As anticipated, younger Canadians are more willing to reform traditional institutions and practices of Canadian democracy, showing higher levels of support for such changes as fixed election dates and measures to ensure better representation of women and visible minorities in elected bodies.
This study also seeks to uncover the underlying sources of these patterns by analyzing attitudinal and behavioural changes that have occurred within dif- ferent age groups over the 1990 to 2000 period. This analysis focuses on three distinct forces – life cycle, generational and period effects – that may account for age-related differences and change over time. At issue is whether today’s young Canadians are unique in their political attitudes and behavioural dispositions, or are simply repeating patterns established by older generations at the same age. The analysis suggests that a variety of forces are at work:
Today’s young Canadians are participating in the political system at lower levels than previous generations did at the same age, suggesting that recent declines in voting turnout and other measures of political partici- pation will not be reversed in future years.
As Canadians age, they tend to develop a greater interest in politics, but this is accompanied by a host of increasingly negative political evalua- tions, including a greater sense of political inefficacy, increased cynicism regarding political parties and the role of money in elections, and less sat- isfaction with democracy, elections and the electoral system in particular.
Some changes have affected the population as a whole, including an over- all increase in satisfaction with democracy, federal elections and in pop- ulist sentiment, a decrease in political cynicism and perceptions of politi- cal inefficacy, a decreased willingness to accept the electoral system’s dis- torting effects, increased cynicism regarding the influence of money in politics and heightened dissatisfaction with the effective choice provided by existing political parties.
A current marketing trend is to target advertising and goods toward particular generations of consumers. According to marketers, the tastes and buying pat- terns of Baby Boomers, Generation X and, most recently, Generation Y con- sumers (a.k.a. Echo Boomers) differ – distinct marketing strategies are neces- sary to target each group effectively. Marketers speak of cohorts, groups of con- sumers grouped by their birth years, sharing similar worldviews and behav- iours because of shared experiences in the formative years of youth and young adulthood. Such strategies have had some success. The iMac computer, avail- able in a trendy range of neon colours, was developed and marketed specifi- cally for Generation Y consumers who are said to be less brand loyal and more willing to purchase products simply because they are “cool” and reflect their particular culture.
The concept of generational cohorts has also been employed in the study of political opinion and behaviour. Popular pieces, including Boom, Bust and Echo and Sex in the Snow, have examined generational differences in social, political and other values.1 If the marketing concept provides a means for increasing the profits of companies that successfully target consumers, its equivalent in the study of public opinion might provide a means for assessing the degree to which political institutions and actors are considered legitimate in the eyes of particu- lar generations of citizens. From this, remedial measures can be developed that address issues of specific concern to different cohorts.
Two academic branches of research in the area of cohort analysis of pub- lic opinion and political behaviour suggest that democracy might very well be viewed differently across generations. The first and more optimistic branch suggests that recent periods of economic and physical security in advanced industrial societies and an increase in educational levels among younger gen- erations have had important consequences for democratic politics.2 Having come of age in periods quite different from those of previous generations, the post-war generations redefined their interests away from the material and instead adopted post-material interests; and they changed their political behaviour in response to this redefinition of values. Younger generations are more likely to engage in “new politics,” to be concerned with non-economic issues such as human rights and the environment, and to be involved with non-traditional institutions and processes such as grassroots social move- ments and protest behaviour. Moreover, while these studies suggest that inter- est in politics has increased among post-war generations, expressed confi- dence in governmental and non-governmental institutions is dropping. Younger generations, while perhaps more interested in the stuff of politics, are turned off by its traditional manifestations.
The second branch of work suggests that technological and social change over the last thirty years has led to a significant decline in social capital – con- sisting of levels of civic engagement, trust in traditional institutions and exchange among members of communities and neighbourhoods – which is most evident among younger generations in advanced industrial states.3 Increased cynicism toward political institutions, lack of trust in politicians, weaker ties to political parties and decreasing levels of voter turnout have been linked to this phenom- enon. This drop in social capital is problematic for democracy in part because it translates into lower levels of participation in traditional democratic institutions and processes. The success of democracy depends on its legitimacy in the eyes of citizens and on minimal levels of participation in its institutions and processes, most notably elections.
In keeping with these important veins of research, the objective of this report is to examine the degree to which the concept of generational cohorts helps in understanding the political beliefs and behavioural patterns of Canadians. The instrument that makes this examination possible is a public opinion survey (hereafter referred to as the Strengthening Canadian Democracy Survey or SCD Survey) of 1,278 Canadians, commissioned by the Institute for Research on Public Policy and carried out from February 16 to April 2 of 2000.4 The SCD Survey provides an opportunity to investigate gen- erational patterns in Canadians’ levels of trust in political institutions and par- ticipation in democratic processes.
The report consists of two sections. The first examines whether younger Canadians are distinguishable in their political activity and in their attitudes toward a range of political institutions and processes, information crucial for an assessment of the health of Canadian democracy. More specifically, the first section considers whether the decline in trust, and shifts in participation and in levels of support for political institutions are especially evident among younger Canadians. In order to do this, survey respondents are assigned to an age group. The determination of age groups – the breakdown of respondents by birth year into groups that have shared similar formative experiences – is less than straightforward.5 For the purposes of this report, respondents were grouped into one of five categories according to their year of birth: the youngest group’s birth years are after 1972;6 the second group’s between 1963 and 1972 inclusive; the third’s between 1953 and 1962; the fourth’s between 1943 and 1952; and the final group’s before 1943. This breakdown provides sufficiently small groupings to allow for a nuanced examination of younger Canadians without creating so many groups as to render the identification of patterns difficult. Throughout the first section of this report these groups will be identified by their ages at the time of the survey: 18-27 years, 28-37 years, 38-47 years, 48-57 years, and over 57.7
The second section of the report seeks to determine whether discovered patterns in opinion are due to generational, life cycle or period effects. Generational effects are those described above: differences that reflect the importance of shared formative experiences among generational cohorts. Alternatively, variation across age groups may reflect life cycle effects: as peo- ple progress through life their attitudes and behaviour may shift in light of the changing interests and experiences associated with different stages in the life cycle. Lastly, patterns of opinion across age groups can also reveal the impor- tance of some shared experience or event common to all Canadians – that is, period effects. In order to isolate these competing effects, it is necessary to examine how patterns of opinion change within different cohorts over time. Hence, in the second section of the report comparisons are made to the results found in a survey conducted in 1990 for the Royal Commission on Electoral Reform and Party Financing (the Lortie Commission) and another conducted in 1993 at the time of the federal election.8 Several of the questions from these earlier surveys were replicated in the SCD Survey, allowing for the direct com- parison of results between the two periods.
The importance of differentiating life cycle, period and generational effects stems from their differing consequences. As an example, evidence that today’s younger Canadians are less likely to vote may or may not be cause for concern. This result might simply reflect life cycle effects: young people are less likely to see the importance of voting but as they age their turnout will increase. The result is no substantial change in turnout level over time and lit- tle cause for concern.9 But if the source of such difference is generational, the result is more disturbing: the lower level of turnout among younger Canadians may reflect some distinct formative experience shared by the generation that is unlikely to be completely overturned as they age, resulting in an overall drop in voter turnout at the aggregate level as this generation moves through the life cycle. This type of generational effect would mean today’s younger Canadians are less likely to vote than were younger Canadians of the previous generation at the same age, and aging alone is unlikely to allow them to “catch up.” Finally, period effects are also a worrying possibility. Period effects would be said to exist if a drop in turnout were reported in every age group over a certain peri- od of time. Such a result would also be of obvious concern for the health of Canadian democracy.
Age Patterns in Political Attitudes and Behaviour
Are younger Canadians disengaged from traditional political institutions? The survey provides several measures that suggest political engagement does vary by age group, and very often in the anticipated direction. The young appear at best to be politically passive, and at worst, politically apathetic.
The survey asked respondents whether or not they “followed” politics. The pattern revealed among the responses in Table 1 is clear: younger respondents are significantly less likely to pay attention to politics and political news than are older Canadians.10 In fact, in all age groups but the youngest, a majority reveal that they follow politics very or fairly closely. Among the 18-27 group, 59 per- cent of respondents said that they do not follow politics at any level. Among the most attentive group, respondents over 57, 68 percent said they follow politics at some level. This most basic political act, requiring a minimum degree of effort, exhibits significant variation across age groups.
Following politics closely should result in greater political knowledge, and the results suggest that older Canadians are indeed significantly more politically knowledgeable than younger Canadians. When asked to name the Prime Minister, the federal Minister of Finance and the official opposition in Ottawa, older Canadians are more likely to answer correctly (Table 2). Although the dif- ferences in knowledge are relatively small for a well-known figure such as the Prime Minister, they become significant as the difficulty of the question increas- es. For example, only 22 percent of respondents aged 18-27 were able to identi- fy Paul Martin as the Minister of Finance while the corresponding figure among those in the over-57 group is 65 percent. Similarly, while only 20 percent of the youngest respondents knew that the Reform Party was the official opposition, this share consistently increases with age to 48 percent of respondents in the old- est age group. At least at the aggregate level, the greater attention paid to politics by older Canadians appears to translate directly into greater political knowledge.
Respondents were then asked about their voting behaviour, a key political act for citizens in liberal democracies. Voting requires little effort, so low voter turnouts may suggest that the health of a democracy is in jeopardy. The pattern across age groups revealed in levels of political interest is reproduced in voting participation rates (Table 3). Although a majority of respondents in every age group reports having voted in the 1997 federal election, this share increases directly with age. Among respondents in the 18-27 group who were of voting age in 1997, 66 percent report having cast a ballot in the 1997 federal election. The corresponding figures in the two oldest age groups are 92 percent and 91 percent.
This decreased willingness to exercise voting rights stems in part from a stronger belief among younger respondents that voting is not very important (Table 3). Respondents were asked “In your view, how important is it that people vote in elections?” While 75 percent of the youngest respondents believe voting is essential or very important, about 90 percent of respondents in the three oldest age groups provided this response. This suggests the weaker participation among younger Canadians may be due to the lesser significance they accord to voting.11
If fewer younger Canadians are voting, it is also the case that almost none of them have joined a political party (Table 4). In the 18-27 group, virtually all (98 percent) respondents report that they have never been a member of a polit- ical party. The participation rate improves as one moves across the age groups: among the over-57 group, 33 percent report having been a member of a politi- cal party at some point in their lives. (It should be noted that question wording is partly responsible for this result. The question asks whether respondents have ever been a member of a political party, so the higher rates of participation among older respondents are to be expected.)
The low rate of party membership among younger Canadians is not likely due to a lack of access since joining a political party often involves little more than a phone call to a constituency office or a visit to a website. Instead, it lies at least partly in the fact that few younger Canadians have even thought about joining a political party. Respondents who had never been a member of a party were asked, “Have you thought about joining one?” Among these non-party members, the vast majority of respondents in all age groups have never thought of joining a party, but there is some evidence that party membership is less often considered among younger Canadians, ranging from 97 percent of non-members in the 18-27 group to 85 percent in the 48-57 group (data not shown).
Respondents were also asked if they had ever been a member of an inter- est group. And as in the case of political party membership the rate of participa- tion increases with age. While 9 percent of respondents in the youngest group identified themselves as having been an interest group member, this share increases to 19 percent among the 48-57 group. The trend is not consistent, however, in that interest group membership among the oldest age group drops slightly to 11 percent.
Most interesting, however, is a comparison of rates of participation in parties and interest groups. This reveals that younger Canadians are far more likely to choose the latter over the former than older Canadians. The ratio of interest group participation to political party participation (see Table 4) is 4.5 to 1 for the youngest age group but only 0.3 to 1 among the oldest age group. Interest groups are more likely to be the organization of choice for younger Canadians who wish to participate politically. Among older Canadians, political parties come out ahead.
This conclusion is reinforced by responses to an additional survey ques- tion asking respondents which of the two, interest groups or political parties, was the more effective instrument for bringing about political change. The results by age group appear in Figure 1. Although there is agreement across the age groups that interest groups are more effective than political parties as mechanisms of political change, this belief is strongest among younger Canadians. While 73 per- cent of respondents in the youngest age group choose interest groups as the more effective mechanism, only 56 percent of respondents in the oldest age group respond similarly. Older Canadians are more likely to select the “both” and “nei- ther” responses and to consider political parties more effective than are younger Canadians.
Thus, when compared to older Canadians, younger Canadians
are less likely to follow politics closely;
are less politically knowledgeable;
are significantly less likely to have voted in the 1997 federal election;
are less likely to see voting as an “essential” democratic act;
are less likely to be or have been a member of a political party or interest group; and
are more likely to believe interest groups are more effective than political parties for bringing about change.
These preliminary findings point to significant political withdrawal among younger Canadians that may signal deeper problems ahead for Canadian democracy.
Political Attitudes and Beliefs
The SCD Survey allows us to examine the degree to which the generational participation patterns outlined above are mirrored in levels of support for the polit- ical system, its institutions and processes and in expressed levels of political effica- cy, trust and cynicism. If younger Canadians are disengaged, is it linked to increased levels of political apathy or, alternatively, to a lack of support and trust in the sys- tem? Both explanations have implications for the health of Canadian democracy.
We begin by examining the level of satisfaction with democracy exhibited across the age groups. If younger Canadians have stopped engaging democrati- cally, it may be due to their unhappiness with the system’s general political struc- tures. The results, however, reveal quite the opposite (Figure 2). Respondents were asked, “On the whole, are you very satisfied, fairly satisfied, not very satis- fied or not satisfied at all with the way democracy works in Canada?” Contrary to expectations, younger Canadians are more likely to feel satisfied with democ- racy than are other Canadians. While 82 percent of the 18-27 group are very or fairly satisfied with democracy, only 65 percent of the over-57 group respond similarly. The relationship is clear: the older the respondent, the lower the level of satisfaction with democracy.12
This pattern persists for satisfaction with federal elections. The share expressing some measure of satisfaction ranges from 81 percent of the 18-27 group to 66 percent of the 48-57 group (Figure 2). On the election question, however, significant variation exists across the age groups in the share of respon- dents responding “don’t know”: 43 percent of the 18-27 group but only 21 per- cent of the 48-57 group. Younger Canadians, it appears, haven’t thought much about federal elections, reflecting perhaps their more limited participation over- all. But among those who do offer an opinion, they are far more likely to be sat- isfied with federal elections than older Canadians.
The Federal Government: Fair and Effective?
The SCD Survey also asked respondents their opinions about the fairness and effectiveness of the federal government. Younger Canadians may be less politi- cally engaged because of their views of the federal government. But as shown in Table 5 the generational patterns found in satisfaction with democracy and federal elections are repeated for opinion on the federal government’s fairness and effec- tiveness. When read the statement, “The federal government generally treats all Canadians fairly,” younger Canadians are more likely to agree than older Canadians. Among the youngest age group a full 64 percent agreed with the statement, while this share drops to 47 and 53 percent respectively among the two oldest groups.
The pattern is the same for the question concerning the federal govern- ment’s effectiveness. Younger Canadians are more likely to agree that “the feder- al government usually does a good job of getting things done” given all the demands made on it. Among the two youngest age groups, roughly 60 percent of respondents either strongly or somewhat agree that the government is effec- tive in this sense, while the equivalent share in the three remaining age groups drops to between 48 and 52 percent. These results underscore a current of dis- satisfaction among older Canadians.
Political Efficacy, Confidence and Trust
This section examines whether the lower levels of satisfaction and more negative evaluation of government performance expressed by older age groups are accompanied by a sense of powerlessness vis-à-vis the government. That is, do those groups who are less politically satisfied overall also believe government does not respond to them or reveal skepticism regarding politicians and people in general? Moreover, is unhappiness with government and politics perhaps accompanied by a desire to increase the political power of citizens at the grass- roots level, that is, by a hint of populism?
A series of statements were read to assess levels of political efficacy, trust, cynicism and support for populism. The analysis of these responses by age group appears in Table 6. The results reveal that beliefs vary only slightly across the age groups on these questions and, as such, are unlikely to be linked to generational differences in political attitudes and behaviour.
Nevertheless, the results reveal the pervasiveness of certain negative opin- ions about our political system. The health of a democracy is directly linked to the degree to which its citizens believe they can effect change if they so desire. A majority of Canadians in all age groups agrees that “people like me do not have much say over what the government does.” The legitimacy of a representative system of government also depends on whether citizens believe their concerns are effectively represented within government. A strong majority in all age groups agrees that “those elected to Parliament soon lose touch with the people.” Some variation is recorded on this question: while 81 percent of the 38-47 group agrees that members of Parliament soon lose touch, this share drops to 67 and 71 per- cent among the oldest and youngest age groups respectively.
Interestingly, however, when respondents were asked whether they agree that “generally most people cannot be trusted,” the pattern is reversed. A major- ity in every age group disagrees with the statement although there is some vari- ation across the age groups. Canadians aged 38 to 47 reveal the greatest levels of trust, with only 3 of 10 respondents agreeing with the statement. The level of general cynicism increases in each of the remaining age groups, peaking at 42 percent in the 18-27 age group. Cynical views of politicians are not necessarily matched by a general cynical outlook.13
Several questions address populist sentiment, that is, the support accord- ed to granting a greater measure of political power directly to the “people.” For example, the results reveal that a significant measure of consensus exists on the question of whether “We could probably solve most of our big national problems if we brought them back to the grassroots level.” Between 75 percent and 82 per- cent of respondents in all age groups agree with the statement. On the related question of the degree of confidence respondents have in the ability of citizens to render good and wise judgements on political questions, a significant majority in each age group expresses some confidence in Canadians, a level which varies only marginally across the age groups (from 71 percent to 79 percent). Canadians of all ages overwhelmingly believe that Canadians can and should be trusted to help solve national problems. In spite of this populist bent, however, opinion is evenly split within each age group between those who agree that “The major issues of the day are too complicated for most voters” and those who do not. Canadians of all ages appear to desire increased political power even if they are not convinced that all Canadians can deal with complicated political issues.
Money and Politics
Recently, much attention has been paid to the effectiveness of existing spending controls on campaigns and elections and to the need for tightening them.14 The generational differences seen above might reflect varying beliefs on this score. The results, however, suggest a significant degree of pessimism exists among all age groups regarding the degree to which money provides an unfair advantage in Canadian politics.15
The SCD Survey asked respondents whether they agreed with the state- ments “People with money have a lot of influence over the government” and “The party that spends the most is the most sure to win the election.” Table 7 reports these results for each of the five age groups. There exists little in the way of a clear generational pattern in responses to the first question. The share of respondents agreeing with the statement constitutes a majority of responses in all age groups. This majority is at its lowest, 86 percent, among the youngest respondents and at, or above, 90 percent in each of the remaining age groups, peaking at 96 per- cent in the 48-57 group. Nearly all Canadians believe money unfairly advantages some individuals, though this belief is marginally weaker among younger Canadians.
Canadians appear less convinced that money has an influence over elec- toral results, though a majority in every age group continues to agree that spend- ing determines electoral success. And older Canadians hold only slightly more cynical beliefs regarding elections as fair contests between relatively equal polit- ical parties than do younger Canadians. More than two out of every three respondents over age 47 agree with the statement.
The Electoral System
Respondents were asked for their views on the fairness of the first-past- the-post system, an increasingly contested institution in light of its distortion of vote shares among the electorate into seat shares in the legislature. Here again, beliefs regarding the fairness of this institution might account for differences in willingness to participate in it. Responses to four versions of this question regard- ing the electoral system have been collapsed and appear in Figure 3.16 As shown, however, the results are less than helpful in explaining generational differences in political behaviour.
A significant number of respondents in every age group do not hold an opin- ion on this question, the non-response rate being particularly high among the younger age groups.17 But among those respondents who do offer an opinion, the system is overwhelmingly regarded as unfair or unacceptable, often by a ratio of more than two to one, and in some cases by a ratio of just under three to one. Although the youngest age group seems somewhat less likely to render a negative assessment of the electoral system than the remaining age groups, this difference is marginal. The most critical view of the system comes from those in the 38-47 group, with 74 percent of respon- dents believing the system is unfair/unacceptable. The legitimacy of the electoral sys- tem is undoubtedly in question among many Canadians.
Political parties play a fundamental role in structuring electoral choice and ultimately government policy. Accordingly, respondents were asked to assess the importance of parties within Canadian democracy and the degree to which Canadian parties offer voters a clear choice, attitudes presumably significant to determining individual participation choices.
As shown in Table 8, a significant majority in each age group agrees that polit- ical parties are a key component of democracy, with over 70 percent of all respon- dents agreeing with the statement, “Without political parties there cannot be true democracy.” When respondents are asked whether they agree or disagree with the statement “All federal parties are basically the same; there isn’t really a choice,” opin- ion is more divided. A majority of respondents in every age group except the youngest agrees that political parties do not offer an effective choice; just under half of the 18-27 group concur. It seems younger Canadians are as committed to the idea that political parties play a key role in democracy as older Canadians, and they are more likely to believe political parties offer effective choices during federal elections.
It is commonly accepted that one legacy of the Charlottetown Accord is a weakened willingness among Canadians to concede final decision-making authority over constitutional, and perhaps additional political questions, to elect- ed officials alone. That younger Canadians have come of age and been political- ly socialized during such a heightened period of participatory politics may have translated into greater expectations regarding referendums as tools of democrat- ic decision-making. Younger Canadians may be more prone to believes it is their right to have a say in political questions, at least through referendums.
As shown in Table 9, however, when asked if they think referendums are “good things, bad things or that they don’t make much difference” a majority in every age group considers referendums to be a good thing. Opinion is only slightly less positive among the oldest age group, with 51 percent providing a positive evaluation of referendums, although more than one in three respondents in this age group believe referendums do not make much difference.
This pattern varies somewhat in the responses given to four separate sur- vey questions regarding the use of referendums on specific policy issues. Respondents were asked whether referendums would be appropriate for deci- sions on constitutional changes, tax increases, for cuts to social spending and on land claims agreements with Aboriginal peoples (Table 10).18 The trend across these questions is for younger age groups to be more supportive of the use of ref- erendums than those in older age groups. If Charlottetown has a legacy, it lingers in younger Canadians’ greater demand to be included in decision-making.
When respondents are asked about referendums on moral issues, however, the pattern of support changes. Respondents in the survey were randomly assigned one of two questions dealing with referendums on the moral issues of abortion and capital punishment. Table 10 provides the breakdown of responses to these two questions by age group. The pattern of weaker support for referendums among older age groups is not repeated when the issues in question are moral ones.
A majority of Canadians in all age groups supports the use of referendums on the issue of capital punishment. This support is strongest among the three middle age groups (respondents aged 28 through 57) ranging from 74 to 78 per- cent. The lowest level of support on this issue appears in the 18-27 group, with only 62 percent of respondents believing referendums should always or some- times be employed. On the question of employing referendums to decide abor- tion policy, opinion is more divided. Among the oldest age group, only 30 per- cent of respondents believe referendums should always or sometimes be used on the issue.19 Forty-six percent of respondents in the youngest age group give a similar response. The use of referendums for abortion policy receives majority support, however, in the three middle age groups. Thus, younger Canadians are more willing to employ referendums for policy-making except when the issues in question are moral ones.
There has been some media discussion of “mirror” legislative representa- tion, given Jean Chrétien’s decisions in 1993 and 1997 to directly nominate sev- eral women candidates, thereby overriding the authority of local constituency associations in an effort to increase female representation among MPs.20 The legit- imacy of such acts rests squarely on the perceived need to alleviate the weak political representation of certain groups. To gauge opinion on this issue, half of the respondents to the survey were randomly selected and asked about the under-representation of women in the House of Commons; the remaining respondents were asked their opinion on the lack of visible minorities. Table 11 reports these opinions.
An examination of opinion on women’s under-representation provides ini- tial support for the existence of generational differences in support of this “post- materialist” issue. A minority in every age group responds that the low number of women MPs is a very serious or serious problem. Equally clear, however, is that the post-materialist impulse is not strongest among younger Canadians. While 44 percent of respondents in the 48-57 group considers the lack of women in the House of Commons a problem, this share falls to 28 percent among the 18-27 group. The pattern, then, is that the youngest and oldest age groups are least concerned with the legislative under-representation of women.
On the question of the under-representation of visible minorities, howev- er, the pattern changes. Although a majority in each age group remains uncon- cerned, the two youngest age groups do reveal higher levels of concern. Just 28 percent of respondents in the youngest age group suggested that the under-rep- resentation of women was a very serious or serious problem; this figure jumps to 46 percent when the under-represented group is visible minorities. If younger Canadians are more likely to consider the political “gender battle” won, they are much more likely to consider the “visible minority war” as just beginning.
A significant amount of attention has recently been directed at multination- al corporations (MNCs), given their privileged position within the global economy and changing world economic structures. Coverage of recent protests, including APEC, World Trade Organization and FTAA demonstrations, might give the impression that vocal, organized protest against such organizations is dominated by younger rather than older citizens. As shown in Figure 4, however, Canadians of all ages agree overwhelmingly that MNCs are too powerful. When read the follow- ing statement – “Multinational corporations have become so powerful that the gov- ernment no longer has very much control over how things are done in Canada” – between 73 percent and 81 percent of respondents in every age group agree.
Modifying Political Processes – Election Timing
A number of questions in the SCD Survey addressed the willingness of respondents to modify traditional political institutions and practices. Here the findings are more consistent with the lower levels of political interest and dem- ocratic participation described above – younger Canadians appear far less com- mitted to existing processes and conventions than older Canadians. One ques- tion addressed the timing of elections. Canadian governments are free to set the date of elections subject only to a maximum government tenure of five years; some have suggested that governments use this flexibility to their advantage by timing elections when their popularity among the public is highest. In some countries, most notably the United States, the timing of elections is fixed. It could be argued that importing the practice to Canada would increase the per- ceived fairness of the process. Respondents were asked “Do you think we should have set dates for elections, or that the government should decide when elections are held, or do you not have an opinion on this?”21 (Figure 5).
Although a majority in every age group prefers fixed dates for elections, sup- port (ranging from 76 to 82 percent) is somewhat higher among the three youngest age groups than it is among the two oldest groups (between 62 percent and 63 per- cent). Older Canadians are less willing to tinker with this long-standing practice.
A second question dealing with changes to political practices asked one half of the respondents whether political parties should be required to increase the number of female candidates they select to run in elections.22 Responses to this question appear in Table 12.
The results for this question are less than straightforward. At least a plu- rality of every group – in three of five groups, a majority – favours a gender par- ity policy for political parties. The highest level of support exists among the youngest age group, which is somewhat counter-intuitive given that this group was not particularly convinced that gender under-representation was a problem.
The remaining respondents were asked if they would support measures requiring parties to select more members of visible minorities as candidates.23 These results reinforce the conclusion that younger Canadians are much more willing to impose candidate selection requirements on parties than are older Canadians. While 68 percent of respondents in the youngest group would sup- port such measures, only 44 percent of respondents in the oldest age group would as well.
When all respondents were asked if they would support the designation of Aboriginal seats in the House of Commons, a similar pattern of response is found across the age groups (see Table 12). Respondents in the two youngest age groups were more likely to support setting aside parliamentary seats for Aboriginal representatives than respondents in the three oldest groups. Although a majority in every group favoured the policy, support ranged from a high of 70 percent among the 18-27 group to a low of 58 percent among the over-57s. In general, then, younger Canadians appear to be more willing to adopt measures to counter representational weaknesses in legislative institutions, especially for visible minorities and Aboriginal peoples.
Repeated calls for the loosening of parliamentary conventions that require legislators to vote along party lines have been heard in recent times. Many Canadians question what appears to be the self-interest of MPs winning out over the interests of their constituents. The SCD Survey asked respondents three dis- tinctly worded versions of a question on free votes in the House of Commons, each to a randomly selected third of the respondents. Responses to these three versions have been merged and the results appear in Figure 6.24
The results suggest that, unlike opinion on some suggested reforms, age does not significantly demarcate opinion on the question of party discipline. Instead, consistently strong support for relaxing party discipline exists across all age groups and appears to vary little with question wording: agreement that laws would be better if party discipline were loosened never falls below 82 percent in any age group.
Respondents who agreed that free votes would lead to better laws then had their opinions challenged by providing them with a counter-argument: “What if this means that after an election it is more difficult for the government to do the things they said they were going to do?” Despite the challenge, agree- ment never falls below 81 percent of respondents in any age group on any of the three questions (results not shown). Canadians of all ages are committed to their dislike of party discipline.
A further set of questions in the SCD Survey dealt with the relative power of institutions within or connected to the Canadian political system. The first address- es the balance of power within the federal system. Respondents were asked whether they believed “the provinces should have more say over the way things are done in Canada.” The results suggest there is relatively little variation across age groups in willingness to side with the provinces in the federalism game (Figure 7). A significant majority in each age group agrees the provinces deserve a greater say.
A second question deals with the division of power between the courts and the legislature. The Charter of Rights affords significant power to the courts, particularly the Supreme Court, in that government legislation can be overturned in court challenges based on the argument that the legislation com- promises the protection of basic rights enshrined in the Charter. Recent Supreme Court decisions have led to criticism of the Supreme Court’s author- ity, since it, unlike Parliament, is a non-elected body that is not directly accountable to anyone. Younger Canadians, however, have come of age at a time when considerable effort has been expended to generate support for the Charter, only recently enshrined in the Constitution Act, 1982. Consequently, they may be more willing to grant the Supreme Court the independence nec- essary to protect Charter rights and less willing to allow governments to over- ride the Court’s decisions.
To assess opinion on this issue, respondents were randomly divided into two groups and asked one of two related survey questions. Responses to these questions were merged and the results appear in Figure 8.25
Although a plurality in every age group sides with the courts in Charter disputes, a pattern across the age groups nevertheless appears: younger Canadians are more likely to support the courts over the government/legisla- ture, with 14 percentage points separating the youngest and oldest age groups in their support for the judicial branch. It appears that younger Canadians, having grown up with the Charter and more likely to have been formally edu- cated about its benefits, are more willing to cede power to the courts in inter- preting it.
In summary, the results in this first section reveal a mix of generational pat- terns in opinion across a number of issues. On a few questions, such as the use of free votes in the House of Commons and in levels of populist sentiment, little differentiation in opinion exists across age groups. On a number of questions, however, a linear pattern exists, in which support or agreement consistently falls or increases with age.
Younger Canadians are less likely to follow politics, be politically knowl- edgeable or participate politically by either voting or joining a party.
Younger Canadians reveal higher levels of satisfaction with a number of institutions, including democracy and elections generally, and they hold healthier opinions on the federal government, in that they are more like- ly to see it as fair and effective. They also feel somewhat less ineffective in their relationship with government and are less likely to believe money accords certain individuals an unfair political advantage.
Younger Canadians are somewhat less likely to see political parties as essential to democracy but are more likely to believe the existing mix of political parties offers an effective choice to voters.
The proportion that doesn’t hold an opinion on elections generally and the first-past-the-post system specifically is highest among younger Canadians.
Younger Canadians are more willing to change the political system, in that they desire a greater role for referendums (except when the issues in ques- tion are moral ones), are more supportive of fixed election dates, are more willing to impose requirements on parties in the selection of local con- stituency candidates and are more willing to designate House of Commons seats for Aboriginal representatives.
Younger Canadians also are more likely to see the courts as the proper forum for ruling on Charter challenges.
On a few questions, attitudes in the 48-57 group stood out. This group evinced:
the greatest concern for women’s under-representation and the second- highest concern for the representation of visible minorities;
the lowest levels of satisfaction with government, politics and federal elections;
the lowest level of perceived political efficacy;
limited support for the suggestion that government is fair and gets things done effectively; and
greater agreement that money provides an unfair advantage in Canadian politics.
This group came of age during the 1960s and their expectations for the political system might be higher than for other groups. The consequence of high- er expectations is that they are harder to meet, something that very likely influ- ences the distinctive opinions expressed by this group.
Nevertheless, a fairly consistent trend in political opinion exists. Older Canadians seem less satisfied with a range of political institutions, organizations and processes than younger Canadians and they appear to hold stronger views on a number of questions. Yet they are often less likely to want to experiment with traditional practices and conventions. Younger Canadians appear more open to placing restrictions on groups and processes with which they have little involvement, particularly political parties and elections.
If the health of Canadian democracy is in jeopardy, the challenge appears to come predominantly from the political attitudes of older Canadians and from the weak political engagement of those who more recently came of age. In other words, although younger Canadians appear to be less politically engaged, this disengagement appears less a conscious decision to turn away from politics than a failure to see the importance of political participation, combined perhaps with a belief that traditional politics may not be providing effective mechanisms for translating desire into action. Older Canadians, on the other hand, may be par- ticipating at higher levels, but they are doing so in spite of greater dissatisfaction with the system overall.
Life Cycle, Generation and Period Effects
This section attempts to determine to what extent three distinct forces – life cycle effects, generational effects and period effects – may account for the differences in attitudes and behaviours revealed in the first section of this report.
Life cycle effects consist in differences that stem from the process of aging. As individuals age, their values and life experiences lead them to alter their opin- ions and behaviour so that at any given point in time, a cross-sectional survey of attitudes will reveal differences across age groups. If older Canadians are less sat- isfied with democracy it may be due simply to the fact that experience breeds dis- satisfaction.
The second force, generational effects, consists in differences in attitudes and behaviours that are due to the sharing of unique common experiences with- in a generational group (or cohort) during youth and early adulthood. Inglehart has pointed out that post-material (post-war) generations within industrial democracies have not experienced war in the same manner as previous genera- tions, and this has translated into enduring, inter-generational differences in polit- ical values, opinions and behaviour. Differences in attitudes across age groups, then, can be the result of changes in the formative experiences of each generation.
The final force, period effects, identifies the possibility that attitudinal change can occur as a result of shared experiences across all age groups. We might expect that all adults in a population, regardless of age, would experience a similar shift in their attitudes in response to a highly salient political event. In recent Canadian history, the passage of the Goods and Services Tax legislation by the Mulroney government might be one example. The appointment of eight Conservative senators based on an obscure provision in the Constitution ensured the smooth passage of the legislation through the upper house but met with a sig- nificant level of disapproval among the public; the result might have been an increased level of cynicism among Canadians of all ages.
In order to distinguish between these three forces, data from two separate periods are required. The SCD Survey repeated several questions from earlier surveys to allow for just such a comparison. Most are from the 1990 Lortie sur- vey,26 undertaken for the Royal Commission on Electoral Reform and Party Financing, although an additional item comes from the 1993 Canadian Election Study. Attitudinal and behavioural change over the ten-year period was exam- ined in an effort to disentangle life cycle, generational and period effects. The results for those questions that revealed significant attitudinal or behavioural change within birth cohorts over the period appear below.
In each case the distribution of responses within a birth cohort is com- pared between the SCD Survey and the earlier survey. Although the labels in the tables for the 2000 data have changed, the categories are nevertheless compara- ble; where the first section broke the 2000 sample into age groups, this section reports the same groups by their years of birth, that is, by birth cohort. Respondents from previous surveys are broken down into the same birth cohorts. The 1990 and 2000 survey responses within each birth cohort are then compared, thus revealing any changes in opinion that may have occurred over the ten-year period.27 If significant differences in attitudes within each cohort are apparent between the two surveys, the possibility of period and/or life cycle effects exists. On the other hand, if there is stability within the cohorts, genera- tional effects may be at play.28
Political Interest and Participation
Table 13 allows for the assessment of the relative influence of life cycle, generational and period effects on political interest, voting and party member- ship. On the first question, political interest, there exists relatively little change at the aggregate level between 1990 and 2000: roughly half of all respondents in both surveys reveal some level of political interest. An assessment of change within each birth cohort suggests, however, that life cycle effects are at play on the question: each cohort reveals an increase in political interest over the 10-year period. As Canadians age, they become more politically interested.
The SCD Survey revealed that younger Canadians were much less likely to vote in federal elections than older Canadians. An examination of responses from the 1990 survey suggests the same pattern in voting behaviour across age groups ten years earlier. But something has clearly been changing, as there was a seven per- centage point aggregate drop in reported vote, to 81 percent in 2000 from 88 per- cent in 1990. Closer inspection suggests that this drop is largely the result of gen- erational effects. Every birth cohort but one – 1963-1972 – reveals relative stabili- ty in their voting behaviour. The significant drop in reported voting among the youngest cohort in 1990 (from 74 percent in 1990 to 69 percent in 2000), com- bined with a drop in the level of voting reported among the incoming generation (74 percent in 1990 compared to only 66 percent in 2000), suggests that the weak- er voter turnout among younger Canadians is not likely to change as they age.
It was also true that younger Canadians in the SCD Survey were less like- ly to have ever joined a political party. Comparing 1990 and 2000 party mem- bership rates reveals a similar trend existed in 1990. The examination of birth cohorts, however, points to considerable increases in the share having been a party member only among the two oldest generations (1943-1952, and before 1943). The incidence of party membership among those born before 1943, for example, increases by 10 percentage points over the ten-year period. There are, then, life cycle effects at play; as Canadians age, particularly beyond a certain age, they are increasingly likely to have joined a political party. But there is some evi- dence that party membership figures will nevertheless drop over time: the lack of intracohort movement among the youngest cohorts, combined with a sizable drop in party membership rates in the 2000 data among the entering cohort (1973-1982) suggests that membership numbers will not completely recover over time. The effects are less apparent now, given the offsetting effects of increased participation among the oldest birth cohort. Thus, although more Canadians are likely to join political parties as they age, the overall rate of par- ticipation may nevertheless decrease over time.
Political Attitudes and Beliefs
The comparison of surveys also allows for an assessment of generational, life cycle and period effects on a number of political opinions and beliefs held by Canadians. Table 14 compares satisfaction with democracy expressed in the 1993 Canadian Election Survey with that expressed in the SCD Survey. This table differs from the previous table in that birth cohorts appear in seven-year rather than ten-year intervals in order to allow for direct birth cohort comparisons between the two surveys.
An examination of attitudes in the 1993 study reveals little difference in satisfaction with democracy across the cohorts. As such, the existence of an increased level of satisfaction within almost every birth cohort over the seven- year period, especially within the younger cohorts, suggests a period effect on this question. Canadians at large have become more satisfied with democracy over the past seven years. Additionally, generational effects may account for part of the pattern revealed in Table 14: members of the older cohorts reveal less in the way of attitudinal change than do younger cohorts. Moreover, the incoming generation of Canadians (1976-1982) in 2000 are significantly more satisfied with democracy than the incoming generation (1969-1975) was in 1993.
It was reported earlier that some differentiation appears across age groups in reported levels of political efficacy, cynicism and populism. Table 15 compares survey data from 1990 and 2000 on these three attitudinal dimensions. The com- parison of levels of efficacy exhibited at the aggregate level in 1990 and 2000 (the first two columns) suggests that Canadians’ sense of powerlessness vis-à-vis the government has increased. The comparison across birth cohorts in each period suggests that older Canadians feel less efficacious than younger Canadians, although the trend is not strong and not consistent. An examination of intra- cohort change over the ten-year period reveals that three cohorts possessed sig- nificantly lower levels of political efficacy in 2000 than in 1990, that is, a greater share of respondents agreed in 2000 with the statement “People like me do not have much say over what the government does.” Such intracohort change is con- sistent with life cycle effects – as respondents age they are less convinced they can effect change in government policies – but the magnitude of change is greater than would be expected for life cycle effects alone. Hence, period effects – a slight decline in efficacy across all cohorts – also are likely at play.
The SCD Survey suggests that higher levels of cynicism exist among mid- dle-aged rather than among older or younger Canadians (see column 4 in Table 15). Such differentiation does not, however, appear in the 1990 survey data (col- umn 3). A comparison of intracohort change suggests that minor period effects likely account for the aggregate change in attitudes recorded between the two surveys. Most cohorts report a significant drop in the share agreeing with the statement “Those elected to Parliament soon lose touch with the people,” coun- tering the idea that cynicism increases between youth and middle age and then decreases. Canadians appear to have become less cynical about politicians over the ten-year period.
The last four columns in Table 15 compare populist attitudes expressed in 1990 with those in 2000. On the first question, there is only a relatively minor change at the aggregate level in the share of respondents agreeing with the state- ment, “The major issues of the day are too complicated for most voters.” Relatively consistent attitudes within each cohort suggest little change between the two surveys except within the oldest cohort: while 64 percent of respondents agreed with the statement in 1990, only 51 percent did so in 2000. It seems older Canadians have become more populist. Intracohort movement comparisons between the two surveys on the second populism question reveals a similar result: Canadians in every cohort reveal relatively consistent populist attitudes, except for a modest increase in sentiment within one cohort (1953-1962).
Evidence from the SCD Survey suggested a decreasing level of satisfaction with elections among older age groups, a trend repeated in the 1990 Lortie data (Table 16). An examination of intracohort change suggests, however, that more is going on than simple life cycle effects: every cohort exhibits an increase in satisfac- tion with federal elections, particularly the 1963-1972 and pre-1943 cohorts. Thus, while Canadians do appear to become less satisfied with elections over their life cycle, this effect was offset by an overall increase in Canadians’ satisfaction over the ten-year period. This period effect is evident in the nine-percentage-point jump in satisfaction recorded at the aggregate level between 1990 and 2000.
Asking Canadians to render an opinion on the first-past-the-post electoral system, however, presents somewhat more troubling results. There are few age-relat- ed differences on this item in either the 1990 or the 2000 survey. But an examina- tion of intracohort change suggests the existence of a strong period effect: each birth cohort shows a significant drop in the share of respondents agreeing that the elec- toral system’s distortion of the seats-to-votes ratio is acceptable. This result is rein- forced by the shift in opinion among the incoming generation in each survey: in 1990, 43 percent of the youngest generation saw the system as acceptable; in 2000, only 35 percent did. Thus, Canadians have become more satisfied with elections overall, but they have become less convinced of the electoral system’s legitimacy.
Both the belief that those with money unfairly influence government and the belief that winning elections is partly determined by the amount of money a party spends were found to increase slightly with age in the SCD Survey. Table 17 compares these survey results with data from the 1990 Lortie survey. The first two columns in the table report results for attitudes on the influence of money over government; the last two columns report results for the influence of money over elections. On the first question, there is little in the way of age differences in the 1990 data. But an examination of intracohort change reveals the existence of a slight period effect: between 1990 and 2000, an increase is recorded within each cohort in the share of respondents agreeing with the statement, “People with money have a lot of influence over the government,” although the size of the changes varies significantly. Canadians of all ages have become increasingly cyni- cal about the unfair advantage accorded those with money in Canadian politics.
A much stronger period effect is suggested by intracohort change on the second question. The comparison suggests that while minor life cycle effects exist on this question in both surveys (agreement increases slightly with age group), these are swamped by the substantially increased share of respondents within each cohort agreeing that campaign spending provides an unfair advantage to some parties. Thus, although Canadians become somewhat more cynical on this element of democracy over the life cycle, events between these two surveys appear to have led all Canadians to become significantly more skeptical.
Table 18 reports on changes in attitudes toward federal parties over the ten- year period. An examination of responses in each survey reveals a slight increase in agreement with the statement, “All parties are basically the same; there really isn’t a choice,” across older age groups. Intracohort comparison suggests that life cycle effects do indeed exist: every cohort reveals an increase in the share agreeing with the statement over the ten-year period, a change that brings each cohort closer in attitude to the equivalent age group ten years earlier. Yet the changes in attitudes within each cohort are much larger than would be predicted by life cycle effects alone. While only 10 percentage points separate the attitudes of the youngest and oldest birth cohorts in 2000, the 1963-1972 cohort alone experiences a 17-per- centage-point increase in the share agreeing with the statement. Thus, as they age, Canadians appear to become more cynical about the effective choice afforded by political parties, and Canadians of all ages are becoming more cynical.
Comparing changes in behaviour and opinion between 1990 and 2000 suggests a mix of effects working on the subset of attitudinal and behavioural measures avail- able for comparison. Among the findings, the more marked include the following:
Over the life cycle, Canadians develop a greater interest in politics and become increasingly likely to participate in political parties, but this is accompanied by a host of increasingly negative political evaluations, including a greater sense of political inefficacy, increased cynicism regard- ing political parties and the role of money in elections, and less satisfac- tion with democracy, elections and the electoral system in particular.
Substantial period effects between 1990 and 2000 include an overall increase in satisfaction with democracy, federal elections and in populist sentiment, a decrease in political cynicism and political efficacy, a decreased willingness to accept the electoral system’s distorting effects, increased cyn- icism regarding the influence of money in politics, and heightened dissatis- faction with the effective choice provided by existing political parties.
Generational effects are evident, suggesting that today’s younger Canadians are participating at lower levels than previously recorded both in elections and with political parties.
What do these results portend for the future of Canadian democracy? The most worrying trend is the increased political disengagement apparent among younger Canadians, a trend unlikely to be overcome as they age because it is rooted in generational effects. If the legitimacy of a democratic political system rests on the number of citizens who participate in its elections, then Canadian democracy is in some danger. And although less central perhaps to the success of the democratic enterprise, decreasing involvement with political parties among younger Canadians should similarly raise concerns. There is some evi- dence, however, that younger Canadians are turning instead to alternative polit- ical groups and organizations, which cushions these conclusions somewhat. While we lack a 1990-2000 comparison on this measure, we do know from the SCD Survey that the relative propensity to participate in interest groups versus political parties is much higher among younger Canadians.
Evidence that the more negative opinions expressed by older Canadians on a number of issues is due partly to life cycle effects suggests a natural cycle of opinion that Canadians of all generations move through as they age. But these findings are greatly overshadowed by the findings of changing opinion and behaviour among all Canadians over the period, evidence that points to increased cynicism on several political questions. Most notable here are attitudes on the electoral system, political parties and the influence of money in politics.
Canadian governments would do well to address these findings, both by increasing dialogue with Canadians generally and by implementing measures designed to increase political participation among Canadian youth. The latter might prove more difficult, inasmuch as withdrawal from politics among younger Canadians seems to originate more in political apathy than in political disaffection; it will require the federal and provincial governments to make a con- certed effort to convince young Canadians of the importance and the necessity of their participation in the democratic process. It is a significant undertaking, but one that is crucial to maintaining the vitality of Canadian democracy.
See David Foot, Boom, Bust and Echo: How to Profit from the Coming Demographic Shift (Toronto: Macfarlane Walter & Ross, 1996) and Michael Adams, Sex in the Snow: Canadian Social Values at the End of the Millennium (Toronto: Penguin Group, 1997).
See Ronald Ingelhart, Culture Shift in Advanced Industrial Society (Princeton: Princeton University Press, 1990) and more recently, Neil Nevitte, The Decline of Deference (Peterborough: Broadview Press, 1996).
See Robert D. Putnam, Bowling Alone: The Collapse and Revival of American Community (New York: Simon and Schuster, 2000).
The response rate for the survey was 60 percent. Using standard methods of calcu- lation, results for the full sample of 1,278 respondents are considered accurate with- in 2.7 percent nineteen times out of twen- ty; for sub-samples, the margin of error is greater. Full technical details can be found in Appendix A of Paul Howe and David Northrup, “Strengthening Canadian Democracy: The Views of Canadians,” Policy Matters, Vol. 1, no. 5 (July 2000).
Among marketers, for example, the birth years that make up Baby Boomers, Generation X and Y groups differ depending on the source.
The minimum age for respondents in the SCD Survey is 18 (born before 1982) given that 18 is the minimum age for voting.
The sample size for each age group is: 18- 27: N=273, 28-37: N=275, 38-47: N=285, 48-57: N=228, and over 57: N=212. Group sample sizes vary slightly by ques- tion due to missing values; tables will indi- cate when this deviation in substantial.
See A Survey of Attitudes Towards Electoral Reform (1990) [hereinafter Lortie Commission survey]. Principal investiga- tors: André Blais and Elisabeth Gidengil. Funded by Royal Commission on Electoral Reform and Party Financing, carried out by Institute for Social Research. A comprehensive analysis of the results of this survey can be found in André Blais and Elisabeth Gidengil, Making Representative Democracy Work: The Views of Canadians, Vol. 17 of the research studies commissioned by the Royal Commission on Electoral Reform and Party Financing (Toronto: Dundurn Press, 1991). The 1993 Canadian Election Study Team included Richard Johnston, André Blais, Henry Brady, Elisabeth Gidengil and Neil Nevitte. The survey was funded by the Social Sciences and Humanities Research Council of Canada and carried out by the Institute for Social Research, York University. None of the above are responsible for the analy- ses and interpretations presented here.
Change in the turnout rate at the aggre- gate level could nevertheless occur as a result of changes in age distribution resulting from differences in the total population within each cohort over time.
The percentages in every table are based on the sample size for each age group with all Don’t Know and Refused responses excluded. Instances in which significant variation in these two response categories appears across age groups will be identified for the reader.
The Pearson correlation for voting and the importance a respondent assigns to voting is a moderate 0.30.
A similar pattern appears when respon- dents are asked about their satisfaction with “government” and “politics” rather than with “democracy,” although the share expressing some level of satisfac- tion falls in every age category with each subsequent replacement.
The Pearson correlation between the two questions is 0.02.
This issue was addressed by the Lortie Commission in 1990, and the recent leadership contest within the Canadian Alliance has maintained interest in the subject in light of concerns over the bulk purchase and distribution of party memberships and significant differences between candidates’ campaign budgets.
This pessimism is combined with a belief that there is little hope that any- thing can be done about it within elec- tion campaigns. The Pearson correlation between beliefs on campaign spending’s influence and the ability to control par- ties’ expenditures is 0.30.
For this question the SCD Survey employed two question and response wording experiments to assess the effect of wording on opinion for this issue. These changes had relatively little effect on the distribution of responses. Combined with the fact that the already small sample sizes for each of the varia- tions became increasingly small when broken down by age group, the minimal effect of question rewording justified the collapsing of results for the question ver- sions shown in Figure 4. See Paul Howe and David Northrup, “Strengthening Canadian Democracy: The Views of Canadians,” Policy Matters, Vol. 1, no. 5 (July 2000).
The non-response rate appears some- what related to age: the highest non- response rate, 39 percent, appears among the youngest age group and decreases steadily to a low of 15 percent among the 48-57 group. The pattern ends there, however, as the over-57 group records a non-response rate of 27 percent on this question.
A slight age pattern in non-response was evident in responses to the referendums on tax increases, social spending and land claims agreement questions. In each of these cases, the rate of non-response increased slightly with age.
The pattern of non-responses found in the previous referendum question is replicated on moral issues; while only three percent of respondents in the 18-27 group did not respond to the abor- tion question, 14 percent of the over-57 group failed to respond.
Mirror representation exists in the legis- lature if the descriptive characteristics of its members closely match those of the public from which they were drawn. See Anne Phillips, Engendering Democracy (University Park, PA: Pennsylvania State University Press, 1991). For an example of the debate surrounding Chrétien’s decision to appoint women candidates in the 1997 election see Allan Chambers, “Alberta Liberals meet Chrétien’s quota for women,” Edmonton Journal, March 18, 1997, p. A3.
Like the question on general satisfaction with elections, a significant proportion within each age group does not render an opinion on the issue (between 24 and 27 per cent) although there exists little in the way of a generational pattern in non-responses.
Respondents to the under-representation of women question (half the survey sam- ple) were then randomly divided into two groups to receive one of two follow- up questions on this issue. The first group was asked whether they would support or oppose the requirement that parties choose as many female as male candidates during elections. The second group was asked their opinion on a weaker option of requiring parties to choose more female candidates than they do now. Given the small sample sizes for the breakdown by age group, responses to these two questions were combined.
A slight pattern in non-response exists on this question: slightly more than one in ten respondents in the 18-27 group did not offer an opinion on this ques- tion; this share increased steadily with age to almost one in five in the over-57 group.
Merging the responses to the three ques- tions runs the risk of concealing varia- tions in responses within each, but this is more than offset by the statistical con- fidence gained in the results by the increased sample size for each age group.
A significant number in each age group offer no response to this question and there is a pattern of non-response across the groups. While eight percent of respondents in the 18-27 group did not respond, this share increases to 22 per- cent in the over-57 group.
The sample sizes for the birth cohorts in the Lortie data are as follows: 1963-1972 N=523; 1953-1962 N=727; 1943-1952 N=637; Before 1943 N=992. The data are weighted by the National Household Weight included in the data set. Tables will identify when sample sizes deviate significantly from those reported here.
Panel data, in which the same respon- dents are interviewed at different points in time, would be preferable to compar- ing different samples from the same pop- ulation over time. Unfortunately such data are unavailable.
For a discussion of the difficulties in try- ing to disentangle cohort, life cycle and period effects see Norval D. Glenn, Cohort Analysis (Beverly Hills: Sage Publications, 1977).