Northern political culture?: Political behaviour in NunavutUne culture politique du Nord?: le comportement politique au Nunavut

  • Ailsa Henderson

…plus d’informations

  • Ailsa Henderson
    Political Scioence,
    Wilfrid Laurier University,
    Ontario, Canada,
    N2L 3C5.

Corps de l’article


Presently, Canadians are bombarded with news that their cynicism in politics is rising and that their participation in traditional avenues of politics is declining. Voter turnout is decreasing as is trust, confidence and satisfaction in politics. The failure of Meech and Charlottetown constitutional accords suggested that constitutional peace proved elusive. The latter also pointed to a considerable schism between the arrangements reached by elites and the attitudes of citizens. The news, then, of political attitudes and behaviour, is usually negative. Within this context, the creation of Nunavut in 1999 represented a good news item. Negotiated as part of a land claim settlement with the Inuit of the Eastern Arctic, the creation of a public government provided an example of successful constitutional change. The establishment of a new territory created a new polity, but also created institutions that would distinguish that polity from any other within Canada. The new institutions sought to redefine the relationship between citizens and the State. At a time when there was little other than negative news about perceptions of politics and political behaviour, here was an experiment in institutional design that addressed that news and sought to alter the political culture. The following article examines the context of political participation in Nunavut with reference to how, if at all, the avenues of political behaviour provide for different levels than in the rest of Canada. The analysis demonstrates that the opportunity for democratic participation is proportionately greater in Nunavut than in other jurisdictions, but that the context of political life in Nunavut has the capacity to promote considerable civic fatigue among the electorate.

Literature and context

Early works in political culture were primarily concerned with the impact of attitudes on the functioning of institutions. Given the unsuccessful projects of institutional change, researchers sought to determine the extent to which norms of attitudes and behaviour within countries produced varying levels of democratic stability. Works by Almond, Verba and Pye in the 1960s examined the German experience with fascism and the unsuccessful grafting of otherwise successful constitutions onto Third World politics (Almond and Verba 1963; Pye and Verba 1965). Thus, political culture research has been, from the very beginning, both concerned with political participation and with institutional change. Times of institutional change provide a particularly fruitful opportunity to examine political culture as they allow for a before-and-after vision of the dominant political attitudes.  Research from Tocqueville to Putnam suggests that the dominant political attitudes and behaviours affect the choices made by institutional designers and structure political life for years to come. The opportunity to exert such influence renders these attitudes and behaviours more influential. Canadian examples often focus on the role of the United Empire Loyalists in providing Canada with a Tory-tinged liberalism. The establishment of Nunavut has also been coloured by the dominant values of the day. Growing dissatisfaction with government encouraged advocates of change to not only establish a legislature, but to address the relationship between citizens and the state (Dahl 1997; Flumian 1999; Gombay 2000; Henderson 2001; Kusugak 2000). In particular, the Nunavut Implementation Commission (NIC) explicitly addressed the need to alter patterns of participation (NIC 1995). For this reason, trends of participation, before and after the establishment of Nunavut, are worth studying. This paper examines the dominant modes of political participation in Nunavut and the ways in which the establishment of the legislature altered the opportunities for political participation. Such tasks are essential first steps in any research that might seek to measure the impact of the legislature on political life in the eastern Arctic.

Existing research on political culture in Canada tells us two things. First, that recently, Canadians display declining trust, satisfaction and confidence in their government. In his investigation of the World Values Survey research from Canada, Nevitte (1996) demonstrates that the myth of Canada as a country of deference should be re-evaluated. In their attitudes, Canadians are becoming less supportive of elites and are showing a growing interest in quality of life or post-material values. In short, their values are changing (Adams 2003). In their behaviour, Canadians also display a greater support for unconventional modes of participation such as petitions or boycotts (Dalton 2002; Nevitte 1996). Political attitudes and behaviour are changing in Canada, and they are not uniform within the country. Simeon and Elkins (1974) indicated in the 1970s and 1980s that provincial boundaries represent significant cultural boundaries and produce differing patterns of efficacy, trust and involvement. Recent research continues to suggest that there may be provincial sub-cultures operating within Canada (Henderson 2003). Such research typically avoids a discussion of northern Canada, in part because of the paucity of available data. Existing research suggests, however, that the context of political life in Canada’s territories is different from that in the provinces. Sizeable Aboriginal populations, economies of scale, vast geographic distances, small legislatures, and adapted Westminster systems are all credited with creating territorial political systems noticeably different from their southern provincial counterparts (Cameron and White 1995; Dickerson 1992; White 1991).

Articles on Canadian political attitudes and behaviours almost exclusively rely on the analysis of quantitative data from large surveys, data that is not readily available for the territories. Only the 1997 election study contained respondents from the three territories, and the sample of under 200 is insufficient, not least because it does not distinguish between residents of the eastern and western portions of the Northwest Territories (NWT). The 1997 Canadian Election Survey (CES) dataset contains 90 respondents from the Yukon and 97 from the then-unitary Northwest Territories. It is impossible to distinguish which respondents are from territorial capitals or, in the case of the NWT, whether respondents are from Nunavut or the west. Of the 90 Yukon respondents, eight describe their ethnic background as Native or Inuit, while 22 of the 97 NWT residents indicated the same. The proportions, which represent 8.9% and 22.7% of the Yukon and NWT, fall well below the 21% and 52% recorded in the general population. In short, the dataset that Canadian political scientists often use to analyse political attitudes and behaviour is of little use for researchers of the North. This certainly helps to explain the lack of quantitative research on political culture in the North. Yet, we have reasons to believe that in terms of political participation, residents of Nunavut display different patterns of activity.

Politics in Nunavut operate according to the basic rules of representative democracy in Canada. Political parties contest seats in the federal election, there is a territorial legislature and local councils for the communities. These three levels mirror those operating in the rest of the country and each level provides opportunities for voting and for standing as candidates. There are three principal differences, however, between political life in Nunavut and political life in the rest of the country. These differences are well known but are worth revisiting as they exert an important influence on political culture in Nunavut. First, political parties do not contest seats in the territorial elections but instead candidates campaign as independents. This is similar to municipal elections in many Canadian cities. Political parties operate federally and thus contest the single constituency that Nunavut represents. Political parties are not prohibited in Nunavut. Their absence is a holdover from the pre-division Northwest Territories where candidates have contested seats as independents since the establishment of a fully-elected territorial legislature in 1975. The government of the Northwest Territories began as an appointed council and underwent a slow transformation as the number of appointed members fell and the number of elected members rose. Given that the existing appointed members sat as independents and that political parties were virtually non-existent in the NWT at the time, partisan contests for the remaining seats were highly unlikely. Devolution did not alter this feature of northern politics. The Nunavut Implementation Commission did not see it in their remit to recommend the creation of party politics nor did public consultations held before 1999 provide evidence that the electorate welcomed any change. In their research note on the first election NIC indicated: “It is not possible to predict with any confidence whether or not party politics will emerge in Nunavut in the early days of its Legislative Assembly” (NIC 1995). The major parties have chosen not to field candidates in the first territorial elections and for this reason, political parties do not operate at the territorial level in Nunavut.

Political parties are not irrelevant to territorial politics. Many members of the legislature are members of political parties, overwhelmingly the Liberal party. One member of the executive, Finance Minister Kelvin Ng, is a member of the national executive of the Liberal party. Jack Anawak, until recently a minister in the executive, was the Liberal MP for the Nunatsiaq riding from 1988 to 1997. The strength of the Liberal party is likely tied to a combination of factors. First, the Liberal party is dominant in many of the “have not” provinces in eastern Canada. The finances of the eastern Arctic would suggest that it is in an economically subservient position to the centre. Second, the Liberal party is popular among “ethnic” voters within Canada. These two points suggest that the natural party of governance, as it thinks itself, is best able to meet the needs of certain populations. Third, Inuit leaders saw the Liberal party, which had been in government for much of the twentieth century, as the best guarantor of a land claim, and later, Nunavut. When Peter Ittinuar (NDP MP for Nunatsiaq) crossed the floor to sit with the Liberals in 1982, his decision reflected a fundamental belief that one party, and one party only, would be in the position to deliver programs, services and a land claim for the Inuit of the eastern Arctic.

While the complete list of reasons for Liberal hegemony remains beyond the remit of this paper, it is worth noting that the party is more organized and better funded than others in the North, and that it is not just active during federal elections. It is accurate to say that political parties are not completely absent from territorial politics, but they are not the main method of organization in territorial election campaigns. In her contribution to the research conducted for the Royal Commission on Party Financing and Electoral Reform, Valerie Alia (1990) discussed the impact of political campaigning in the north. She noted that large distances make political campaigning difficult both because of the great time required to travel to communities but also because of the cost of travel among communities. Flights between the northern communities in the Baffin and Kitikmeot regions cost thousands rather than hundreds of dollars. Political campaigns are expensive even in concentrated ridings in southern Ontario. In the North, not only are the costs higher but the resources of political parties are not available to candidates. The absence of parties during electoral campaigns is related to a second distinguishing characteristic in Nunavut.

The legislature in Nunavut operates according to a system of consensus government. Members of the legislature elect from among themselves the speaker, the premier and the cabinet ministers. The premier is responsible for assigning portfolios to the ministers. The executive thus does not possess a legislative agenda that is necessarily ideologically consistent. Likewise, regular members do not necessarily possess a common ground from which to critique the government nor do they possess an organization structure that would allow for a concerted activity. The executive also involves a proportionately greater number of the seats in the chamber. Unlike the Canadian House of Commons, where the cabinet represents 12.5% of MPs, the Nunavut executive currently involves 37%[1]. Executive dominance is not uncommon in small legislatures in Canada. What is uncommon, however, is an executive occupying almost half the seats in the legislature but not bound by a partisan label. This form of consensus is similar to that operating in the Northwest Territories. Upon division, the Nunavut Implementation Commission did not recommend an end to consensus government. Public government thus operates according to rules quite different from those operating in the provinces but consistent with institutions governing political life before division.

Consensus government has long been seen as the defining characteristic of the Northwest Territories, and now, of Nunavut. The importance of the individual in debate is seen as a return to more traditional decision making. While parties were obviously absent in outpost camps, consensus government is a bit of a misnomer, as applied in the territories. Decisions are not dependent on consensus. Votes are carried by majority and unanimity is not required. Cabinet retains an air of solidarity and has pushed for the removal of ministers who break it. In 2003, Jack Anawak was removed from cabinet by a vote of the legislature, because he disagreed with a cabinet decision to decentralize public sector jobs from Rankin Inlet to Baker Lake. Cabinet solidarity is itself anathema to the pure notion of consensus government. Consensus has as much to do with the absence of ideology as a rallying point in debate, and more to do with the importance of discussing things on their own merits. This in turn is fed by the absence of political parties in the North. In short, one cannot evaluate the existence of consensus government without the absence of political parties from territorial politics.

The third important characteristic is the existence of the birthright organizations. Tungavik Federation of Nunavut (TFN) was established in 1982 and following the work of Inuit Tapirisat sought to negotiate the land claim for the eastern Arctic. TFN was reconstituted as the land claim corporate Nunavut Tunngavik Incorporated (NTI) in 1993 after the passage of the Nunavut Land Claim Agreement (NLCA). NTI promotes the rights and manages the responsibilities that Inuit beneficiaries received under the NLCA. The four-member executive of NTI is elected by all Inuit beneficiaries over the age of 16. Terms for the executive are four years in length and staggered so that two positions become available every two years. In the event that members do not complete their terms, for any reason, elections are more frequent. Three regional Inuit Associations—Qikiqtani, Kitikmeot and Kivalliq—also provide positions for which Inuit land claim beneficiaries may vote. The Qikiqtani Inuit Association (QIA), for example, operates in the more populous Baffin region, and offers 16 positions for which voters may stand, including president, vice-president, secretary-treasurer and 13 general board members[2]. The QIA ensures that the principles of the land claim agreement are upheld in the Baffin region. Selected members of QIA sit alongside members from the other regional Inuit associations on the board of NTI. The structure and role of the QIA is similar to those of the other regional associations.

The existence of two political systems, one public, one for Inuit beneficiaries of the land claim, speaks to the dual nature of political life in Nunavut. It was the Nunavut Land Claim Agreement, signed in 1993, that ensured the creation of a public government for all residents of Nunavut. Public government is elected by all within a jurisdiction, regardless of their ethnicity, and passes legislation that provides services for all within that jurisdiction. For Canadian citizens, residency in Nunavut is the only requirement to vote in territorial elections, and the sole requirement to be covered by territorial programs. The land claim, however, creates bodies and services only for Inuit beneficiaries enrolled under the NLCA. Thus, political change as driven by Inuit resulted in a new territory for all within the eastern Arctic, whether Inuit or not. These three factors, the lack of parties in territorial elections, the operation of consensus government, and the existence of a parallel ethnically-based system of governance, influence the tenor of political participation in Nunavut.

Political participation

In his investigation of political participation Lester Milbrath (1965) emphasized a uni-dimensional hierarchy of activity. Individuals could be categorized as apathetic non-voters, as spectators, whose primary activity is voting, or as gladiators, who when most active present themselves as candidates in elections. Critics of Milbrath’s theory argued that it prioritized electoral behaviour and that the hierarchy was cumulative. Such a conception of political behaviour, they argued, did not acknowledge that individuals might specialize in one area of participation and be less active in others (Verba et al. 1978). Membership in community organizations is absent from the original hierarchy. An approach that excludes this form of participation would thus under-represent the participation of women in political life who tend to participate closer to home. Within Nunavut it would exclude participation in the birthright organizations. Subsequent approaches have pointed to the various ways in which citizens may become involved outside of electoral politics. Political participation may be directed towards the electoral system, by voting, contacting politicians, belonging to parties, or may be non-electoral in its focus on community organizations, protest or petitions. Research on social capital also chronicles the beneficial impact on individuals and communities of involvement at a local level (Putnam 1993). While a comprehensive examination of formal and informal participation would provide for a rich understanding of political culture in Nunavut, this paper focuses on the structure of political life in Nunavut and the impact that the campaign for a land claim has had on patterns of political behaviour. In so doing it determines how and why patterns of formal participation are different than in the rest of the country.

We have reason to believe that traditional levels of political engagement are lower in Nunavut. In part, this can be explained by the factors mentioned above. The first two characteristics are obviously related but it is worth distinguishing between the absence of political parties during campaigns and the absence of political parties during the day-to-day operation of the legislature. The existing research on party identification suggests that parties are the main recruiting agent in political life and provide individuals with short cuts and cues to aid them in their own participation. Individuals need not read the fine print of every piece of legislation but rather may examine how their preferred party reacts to policies. The absence of parties not only makes it harder for individuals to identify the goals of candidates, particularly in the context that Alia describes, but also makes it hard for individuals to quickly and continually identify the main issues of any one debate in the legislature. If political parties are a key recruiting agent in politics, then their absence suggests individuals will not be drawn in to participate in a way they might be in other polities. While some would argue that the small population in Nunavut makes it easier for voters to select candidates because such individuals might be known to them, or might be related to them, several complicating factors make this less likely. Many of the ridings encompass more than one community. As a result, electoral contests would pit candidates from communities other than where the voter lives, against candidates who may be known to the voter. We know from the election results that voters do more than just cast a vote for the community candidate. There is a selection process that occurs, that likely takes account of such factors as the issue positions, personality and gender of the candidate. Second, there is a difference between knowing who a person is and understanding how that person feels about a number of issues. The absence of political parties does not remove ideas from political debate in the North. It does, however, make it difficult to discern the views of the individual candidates on the host of issues facing politicians.

The absence of an organized opposition and executive within the legislature makes it harder for individuals to self-identify with aspects of political debate. In this, the structure of political debate in Nunavut appears similar to that described in early civic studies of the United States that reported low levels of knowledge and engagement (Key 1949). Without political parties, and in consensus forms of government, individuals should be less able to identify issues and candidates, and less able to place political issues on a meaningful spectrum. This is compounded by the absence of local and daily print media in most communities. The two main sources of print news, Nunatsiaq News and News North, are both weekly operations, available each Friday and Monday respectively. Radio and television provide more frequent opportunities to follow political affairs but here too coverage of local issues is minimal. The extent to which these factors affect political participation in Nunavut warrants further investigation.

Data analysis

Examinations of political participation in Nunavut may approach the subject from individual-level data, such as that available in election studies or public opinion polls, and from aggregate or community-level data. To date, only the 1999 and 2001 Nunavut Household Surveys (NuHS) contain individual-level data relevant to the territory. With very few exceptions, however, the survey deals largely with socio-demographic data. Political questions are limited to the assessment of the Nunavut Land Claim and questions probing the perceived impact of Nunavut. Measures of political culture such as trust, efficacy and satisfaction, confidence or post-materialism are not present in the survey, nor are questions probing various measures of political participation or partisan preferences. Given that existing Canada-wide studies exclude the territories from their data collection, and given the demographic nature of the Nunavut data, it is not yet possible to examine political participation fully from individual-level data. What is possible, however, is an examination of the aggregate trends.

The following analysis relies primarily on an examination of a database[3] created by the author that includes data on electoral behaviour and socio-economic characteristics for all communities in the territorial and provincial north in Canada. Electoral data is available for all federal and territorial elections in the 1990s and statistical data draws on the relevant census years of Statistics Canada. The dataset contains information on population change, density, average age, proportion of women, and information on employment, education, language and ethnicity for each community in Nunavut. In addition, it includes information on turnout, rejected ballots, margins of victory, the presence of incumbents, number of candidates, number of female candidates and other electoral information for the 1995 and 1999 territorial elections and the 1997 and 2000 federal elections. This information is used in conjunction with available data on territorial trends in voter turnout and more “gladiatorial” activities to provide for the first examination of formal political behaviour in Nunavut.

Electoral turnout

Turnout levels in Nunavut are lower than in the country as a whole, although they are currently higher than in the Northwest Territories. In federal elections, voters in Nunavut possess turnout rates approximately 10% below the national average. This has not always been the case. Since the creation of the riding in 1976, turnout rates have only been slightly lower than the Canadian average, as Table 1 indicates. The table also demonstrates that turnout rates are falling in Canada, and the North is no exception. This trend is most noticeable since 1993. Between 1993 and 2000, the Canadian turnout rate decreased 9% from 70% to 61%. Over the same period, turnout decreased from 68% to 54% in Nunavut. Turnout in Nunavut is lower than in the rest of the Canada, and is falling as it is in Canada, but it is falling faster than in the country as a whole.

Table 1

Turnout in federal elections, 1979-2000

Turnout in federal elections, 1979-2000

-> Voir la liste des tableaux

Trends in territorial elections are harder to measure accurately. Turnout for the 1999 election, the first for the newly-established legislature, was 89%, higher than turnout in other provincial elections. Within the territory, turnout levels range from 68% in the northern constituency of Quttiktuq to 115% in Cambridge Bay. In contrast, turnout for the post-division NWT was 71%, ranging from 55% in Frame Lake to 91% in Hay River North. Table 2 summarizes the turnout levels in the 24 communities in Nunavut for the four most recent elections, the 1995 NWT election, the 1997 Canadian federal election, the 1999 Nunavut election and the most recent Canadian federal election. The table demonstrates that turnout is higher for territorial contests than for federal elections. It also shows that there is considerable variation among communities. The standard deviation for community turnout levels is similar for the 1997, 1999, 2000 elections and appears to have decreased since the previous NWT election, which produced widely varying levels of turnout. In general, turnout is higher for territorial elections.

Table 2

Turnout in Nunavut communities, 1995-2000

Turnout in Nunavut communities, 1995-2000

* This is the turnout rate for Nunavut constituencies within the NWT. In this election the western constituencies had a turnout rate of 72.49, producing an overall turnout rate of 76.2.

-> Voir la liste des tableaux

In Canada, some provinces, such as Prince Edward Island, display consistently higher levels of turnout in sub-state elections while other provinces, such as Ontario or Alberta, have poorer levels of turnout in provincial elections. Regions with higher sub-state electoral turnout tend to be smaller, and tend to have clear divisions among the dominant political parties. Obviously, the role of parties must be discounted in Nunavut at the territorial level but it is relevant for the lower levels of federal turnout. The hegemony of the Liberal party, credited with declining turnout rates in Canada as a whole, is also likely responsible for lower turnout in federal elections in Nunavut. In addition, the stakes of federal elections, where there is only one MP to select, could appear low to voters. Last, in terms of service provision, the territorial government likely seems more proximate, both physically and psychologically, than the federal government, with whom most voters have likely only ever had contact through its unelected bureaucracy.

There are several reasons to treat these numbers cautiously. First, turnout in territorial elections may be recorded above 100% if individuals who were not originally on the electoral list appear on voting day with proof of eligibility. When turnout is calculated, the former denominator is used, which would elevate the reported level of turnout. Second, there is a significant gap between the number of registered voters in Nunavut and the number of eligible adults. This also suggests that the denominator is smaller than it should be, resulting in an over-representation of turnout at the territorial level.

The existence of birthright organizations provides additional opportunities for Inuit to vote. All Inuit beneficiaries may cast votes for board positions of NTI and one of three regional Inuit associations (Qikiqtani, Kitimeot, Kivalliq). Turnout for these organizations is generally lower although it is worth noting that in this case the electorate includes 16 and 17 year olds. We know from existing research that younger members of the electorate are less likely to vote than those in their 40s. This suggests that the inclusion of 16 and 17 year olds in the voting population would produce lower levels of turnout if voting patterns are consistent with those in other parts of Canada. Nonetheless, since the signing of the land claim in 1993, turnout for birthright elections has ranged from approximately 30% to 60%. Turnout for the most recent NTI election in 2001 was 45%, 13% higher than in the previous election. Turnout for the regional associations tends to be lower still. Only 33% voted in the 2002 QIA election. These rates obviously have more in common with turnout rates for municipal elections across Canada than for federal or territorial/provincial elections.

Table 3 summarizes the various turnout levels for the range of voting opportunities other than territorial and federal elections. In addition to recent turnout information for the birthright organizations, it tracks the decreasing turnout for the various plebiscites associated with the progress of the land claim and the political accord. Turnout for the land claim and the capital plebiscite has more in common with recent electoral behaviour in the territorial elections than in the federal election. The 1997, gender parity plebiscite, in which voters ultimately rejected the proposal to have one male and one female representative from each constituency, is the obvious exception. While a comprehensive analysis of the turnout for that plebiscite is beyond the scope of this paper, it is worth noting that the Inuit leadership involved in the negotiation of the land claim and political accord was relatively united on the issue of the boundary division and the land claim, but largely silent on the location of the capital. This was not the case for the gender parity proposal, which was more divisive than the other issues (Dahl 1997; Gombay 2000). The inconsistent message coming from negotiators, and the general unwillingness of the electorate to embrace a fundamental reform of their institutions, as seen in the public consultations of the NIC, could account for the lower turnout in that plebiscite.

Table 3

Turnout for recent public plebiscites, Inuit-specific votes and elections to birthright organizations

Turnout for recent public plebiscites, Inuit-specific votes and elections to birthright organizations

-> Voir la liste des tableaux

Given what we know of turnout in Nunavut, how can we account for the different trends, and specifically for the general decline in turnout levels? Part of the answer lies in the known predictors of turnout. Those likely to increase turnout include both individual factors, such as age and university education and systemic factors such as a proportional election system and the perception of a meaningful electoral contest. Factors likely to decrease turnout include the frequency elections and a geographically dispersed electorate. If we examine each of these factors from an aggregate level, we learn that as a territory Nunavut possesses several factors that would predict lower levels of turnout.

Predictors of turnout

Nunavut has the youngest electorate in Canada. The median age of Nunavummiut is 22.1 years. The median age in Canada, by contrast, is 37.6 years. Nunavut also has a small proportion of the electorate with university degrees, 8.8% rather than almost 17% across Canada. Both of these effects are exacerbated among Inuit. The age profile for Inuit suggests that their median age is even younger, at 19.1 years, and as a group the proportion of university-educated Inuit is smaller. Only 1.7% of Inuit have university degrees. If these two features are predictors of political behaviour, we would expect lower levels of turnout than in the rest of the country.

In terms of systemic effects, we know that none of the federal, territorial, municipal or birthright elections operate according to proportional electoral systems. All employ a majoritarian first-past-the-post method. Furthermore, the electorate is geographically-dispersed. None of the communities in Nunavut is joined to any other by roads. While two communities, Iqaluit, the capital, and Rankin Inlet contain more than one territorial constituency within their boundaries, many of the 19 constituencies contain communities separated by hundreds of kilometres. Air travel is thus a necessary and expensive feature of electoral campaigning. These two features are often associated with lower electoral turnout.

An additional factor is the perception of a meaningful electoral battle. We can assume in this case that an electoral battle is meaningful if it is close, or if it is fought between diametrically opposing views. At the territorial level this is hampered by a lack of political parties. This is not to say that electoral contests are not tightly-fought races but the absence of polling and local daily media makes it less likely that individuals will know if there’s a tight race. The absence of political parties would make it hard for individuals who pay only cursory attention to political campaigns to detect, at a glance, whether the views of candidates are similar or remarkably different. The sheer size of the constituencies inhibits all-candidate debates that might expose these differences. In short, it is not that electoral battles in Nunavut are not tightly-contested or home to radically different visions of political life, but that it is hard to tell whether they are.

At the federal level the pattern of hegemonic support for the Liberal party prevents a meaningful or close electoral battle. The Liberal party won the last election with 70% of the population vote. This party makes money off elections in Nunavut, raising more than it spends. In 2000, it raised $65,000 and spent $35,000. This contrasts with the Progressive Conservatives (PCs), who in 2000 raised and spent about $6,000. In 1997, the PCs raised $4,000 but spent $11,000. The amounts raised by both parties, and the gap between fundraising and spending, point to the dominant position of the Liberal party in Nunavut. The Liberals are able to raise far more money in the territory than their rivals.

In addition, more so than any jurisdiction in the country, partisan contests in Nunavut retain a distinctly pre-1993 flavour where the main contests take place among the Liberals, NDP and Progressive Conservatives. Federal elections in Nunavut involve fewer small parties and fewer candidates than in the rest of the country. Four candidates contested the one riding in 2000, the three pre-1993 official parties and the Green Party[4]. In the previous election, the Liberals, New Democratic Party (NDP), PCs and Reform Party each presented candidates. There were no independent candidates. For these, and for the 1993 election, the Nunavut contest contained fewer candidates than other ridings in Canada, where on average six candidates contest seats. This is consistent with existing research that suggests urban electoral contests field larger and more diverse lists of candidates.

A final predictor of turnout is the frequency of elections. A greater number of elections is credited with decreased turnout. This helps to explain turnout levels in Switzerland and the United States, where elections occur more frequently than in other polities. If trips to the polls dampen turnout, then here too do we have reason to expect lower levels of turnout in Nunavut. Since 1992, voters in Nunavut have been to the polls at least 10 times, not including municipal elections. This works out approximately to one election per year. In addition to the federal and territorial elections, there were plebiscites on the boundaries of the new territory, the location of the capital, and gender parity. Turnout for these plebiscites ranges from a high of 80% in the 1992 boundary plebiscite, to below 40% in the gender parity plebiscite. Turnout in the boundary plebiscite would have been elevated by fears that a low rate of turnout among the Inuit would have allowed the majority of Mackenzie Valley voters to dictate developments in the eastern Arctic.

In Nunavut right now, each resident will have a federal MP, elected every four or five years, a territorial legislative member, elected up to every 5 years, a mayor and a host of municipal councillors, approximately nine per community. The terms for these posts are four years in length but the elections are staggered. In addition, if individuals do not come forward as candidates, vacant positions are re-opened for competition the following year. This means that individuals may have an opportunity to vote in community elections every year or two. For Inuit voters, there are additional opportunities to vote in the staggered elections for NTI and the relevant regional Inuit association. The result is that the Inuit in Nunavut are expected at the polling station approximately once a year, often more than that. Indeed, because so few NTI presidents have served a full four-year term, elections have been more frequent than intended. As a result, voters in Nunavut, and particularly Inuit voters, have been to the polls more frequently than those in any other province or territory in the country.

Analysing turnout

Attempting to explain turnout levels in Nunavut, and analyzing the impact of various predictors are two different things. The following analysis employs turnout in four elections, 1995, 1997, 1999 and 2000 as the dependent variable. The independent variables are tied to the predictors mentioned in the discussion above. Population, population change and population density are included as a test of community cohesiveness. If voters are geographically dispersed they are less likely to vote. Similarly, a high degree of voter mobility could indicate lower levels of local political knowledge or engagement. Age and university education are included as both are viewed as predictors of increased turnout. The number of Inuit in a constituency is included as a test of civic fatigue. If Inuit enjoy more frequent opportunities to vote, and trips to the polls are seen as predictors of lower turnout, then we might expect that turnout in communities with a higher proportion of Inuit would be lower. Family income is included as a measure of civic engagement. Whether the incumbent ran in the race can be seen as an imperfect test of whether the electoral campaign was closely fought. The number of candidates is also included as a test of the campaign climate. Table 4 summarizes the results of standard OLS regression results for the 1995 and 1999 territorial elections.

Table 4

Predictors of turnout in territorial elections

Predictors of turnout in territorial elections

Results are standardized beta coefficients from ols regression with standard errors in parentheses.

Dv turnout, n=24, *=p<.05, **=p<.01

-> Voir la liste des tableaux

Table 4 indicates that there are three statistically significant predictors of turnout in 1995 and one significant predictor of turnout in 1999. In both elections, as the proportion of the Inuit population increases in a community, so too does turnout. If Inuit voters have greater opportunities to vote, and thus greater capacity for civic fatigue, their proportion in communities seems to have a surprising impact on turnout. This could have less to do with Inuit voters themselves, and more to do with the non-Inuit population. Non-Inuit residents tend to remain in the territory for shorter periods, arriving for employment opportunities and then returning South. If non-Inuit voters are newer arrivals, or plan to stay for only a few years, then they may not participate in political processes the way they might down South. This would likely be most evident in Iqaluit. In communities with higher proportions of Inuit voters, non-Inuit voters may be longer-term residents, and thus more likely to participate. In addition, in the earlier territorial election, size of community and family income also had a positive impact on turnout. In other words, in the NWT election turnout was higher in larger communities with predominantly Inuit populations and higher family incomes. It was lower in more heterogeneous communities such as Iqaluit and in smaller communities. In 1999, however, both size and income ceased to be significant predictors of turnout.

An identical model cannot be run for the 1997 and 2000 elections. The presence of the incumbent and number of candidates was identical for all communities. As a result it was not possible to include these variables. Run with the remaining variables the model fails to produce any significant indicators, nor does the adjusted R2 suggest that the model is providing a robust and complete understanding of turnout in federal elections[5]. A better understanding of turnout in federal elections can be gleaned from additional data.

A second data source for federal turnout is the Nunavut Household Survey (NuHS). The 2001 survey relied on face-to-face interviews conducted in either English or Inuktitut. The sample for the survey was large, at just over 5,800. The questionnaire asked respondents whether they voted in the previous election. This survey also asked individuals about their perceptions of the land claim and the impact of Nunavut on themselves, their community and Nunavut as a whole. Generalized disaffection with the political system could dampen turnout, as individuals who are unhappy with the current state of the government might choose to express their views by avoiding the ballot box. Attitudes to Nunavut are generally positive at present. Over half of all Nunavummiut feel the land claim has had a positive impact on their lives. Not surprisingly this number is higher for Inuit, of whom 60% believe the land claim has had a positive impact on their lives. Only 37% of non-Inuit feel the land claim has had a positive impact on them personally. This is not unexpected, as the land claim contains few provisions that would have a direct impact on non-Inuit.

The indirect impact on society as a whole should not be ignored. Implementation of the land claim included the creation of the Nunavut government, which significantly increased the number of high-paying civil service positions in the territory, the majority of which have been filled by non-Inuit. It is also worth noting that non-Inuit did not believe that the land claim had had a negative impact, but that it had no discernable impact on them as individuals. The survey also asked respondents about their perceptions of the new territory. The proportion of Inuit who believe that the territory has had a positive impact on all Nunavummiut or on the community tends to be lower than for non-Inuit. For example, only 71% of Inuit believe Nunavut has been good for all, while 88% of non-Inuit hold the same belief. The gap is smaller for impact on the community but still present. Only on the perceived impact of Nunavut on the individual are Inuit assessments more positive than those of non-Inuit. Almost three quarters of Inuit believe that Nunavut has had a positive impact on them, while less than two thirds of non-Inuit believe this. The previous discussion suggests that perceptions of Nunavut are generally positive. We can include measures of support for the land claim and Nunavut alongside predictors that approximate the model for the territorial elections. Age, university and income of respondent are included to determine if they function differently than in territorial elections. Similarly, status as land claim beneficiary has been included as a test of civic fatigue. Whether the respondent lives in a small community, with a population of fewer than 1,000, replaces the population and population density measures included in the territorial model.

Table 5 shows that age and income have the intended positive impact on turnout. Level of education is also a significant predictor of voter behaviour. Thus when examined at the individual level rather than in aggregate-level data, these predictors have the intended impact on voter turnout. As the territorial model suggested, whether a respondent is Inuit or not does have a significant impact on voting. Despite the capacity for civic fatigue among Inuit voters, they tend to vote more frequently than their Qallunaat counterparts. Positive evaluations of the land claim and of Nunavut have a small but significant impact on voter turnout, suggesting that engagement with the political system as a whole is a positive predictor of voting. In short, while the aggregate level results suggested that the proportion of Inuit residents was the only consistent statistically significant predictor of voting, the individual results not only confirm this view but also point to other factors affecting turnout. Two things are worth noting. First, the adjusted R2 for the model is quite low, at .130. This suggests that the model is not providing us with a comprehensive vision of turnout in Nunavut. Second, 68.8% of respondents indicated in the survey that they voted in the previous federal election. This is over 15% higher than the actual turnout rate in Nunavut. With a sample of just under 6,000, it is possible but very unlikely that the NuHS uncovered a pocket of voters and consistently excluded non-voters. The gap, however, between the actual turnout rate and the turnout rate of NuHS respondents inhibits the ability of the model to assess predictors of voting.

Table 5

Predictors of turnout in federal elections

Predictors of turnout in federal elections

Results are standardized beta coefficients from ols regression with standard errors in parentheses.

Dv turnout, n= 5,816, *=p<.05, **=p<.01

-> Voir la liste des tableaux

Political gladiators?

For Milbrath (1965), turnout was the lowest form of political participation. The highest form of political behaviour involved those “political gladiators” who ran for and held elected office. These individuals were at the apex of political involvement, more involved, more informed and certainly more active and influential than other citizens. We know from previous sections that turnout rates for federal elections are lower in Nunavut. In territorial elections, predictors such as age and income have an expected impact on turnout. Indeed given the presence of multiple factors that would drive down political activity it is remarkable that turnout is not lower in the territory. If, at first glance patterns of political behaviour in Nunavut suggest declining voter participation and engagement, an examination of turnout suggests this is not the case. These findings are clearer when studied in light of more “gladiatorial” political activities.

Examinations of electoral office show that there are more elected positions per capita in Nunavut than in other parts of the country. As mentioned previously, there is a wide array of elected positions for which one might stand in Nunavut. Opportunities to vote are accompanied by opportunities to stand as candidates for election. There are 288 elected positions in the territory, not including bodies that elect members from general assemblies or single issue bodies such as hunter and trappers organizations, the district education authorities, the boards of directors of the community housing associations or the Inuit Circumpolar Conference. This number includes all elected posts at the federal, territorial, municipal level and the positions available in the birthright organizations. In other words, at any one time, 1.02% of the Nunavut population can hold elected office. Although some individuals may hold more than one position at any one time, the capacity for involvement is quite high. This compares to a typical municipality in Canada, such as Windsor, which, with a population of 200,000 has elected positions for .0075% of the population. If we assume that these electoral contests are fought by similar numbers of individuals then the proportion of the population competing for electoral seats is 1.6% in Nunavut and .02% in a typical Canadian community. We can also compare the propensity of individuals to stand as candidates at the various levels of office.

Table 6. summarizes the patterns of competition for elected office in Nunavut. For every elected position, it indicates the number of individuals competing for posts. With five candidates vying for one seat at the House of Commons, the federal election witnesses the greatest competition for seats. It also involves far fewer individuals than most other competitions. The municipal elections, for example, included 198 candidates. These numbers mask, to a certain extent, variations within Nunavut. In the municipal elections some communities saw 12 or 13 candidates vying for five positions on council. In other communities, seats went unfilled as candidates failed to materialize. In Kimmirut, for example, only one candidate campaigned for the four seats on council. Variations tend to be regional. The Kivalliq, or Keewatin, region in Nunavut, found above Manitoba on the mainland of Canada, witnesses the greatest competition for seats. For every mayoral seat in the Kivalliq, 2.8 candidates presented themselves. This compares to an average contest of two candidates per seat in the other regions. This pattern holds true for council seats as well. In the Kitikmeot and Baffin regions, three candidates contest every two seats on council. The number is much higher in the Kivalliq, with 2.47 candidates per council seat.

Table 6

Competition for elected office in Nunavut

Competition for elected office in Nunavut

* This column indicates the number of individuals standing as candidates for every elected position.

† Elections for municipal elections are staggered so that not all positions are vacant during any one election. This number, and the number for mayors, indicates the number of vacant positions rather than the total number of councillors or mayors in Nunavut.

-> Voir la liste des tableaux

Further evidence of the opportunity for civic engagement surfaces from an examination of Nunavut’s current MLAs. Of the 19 current representatives, five had held elected office at the territorial level in the Legislature of the Northwest Territory. A further five had served as mayors of their respective communities and one served as a deputy mayor. The remaining eight MLAs include among them an MP, and three councillors. Only four MLAs had not served a term in elected office for a public government before 1999. Two, including the premier, had no previous electoral experience, a third ran the local hunter and trapper organization, while a fourth was an elected board member of QIA, the regional Inuit Association in the Baffin. Devolution may have created the opportunity for new individuals to come forward as political candidates. Those elected, however, represent a seasoned political class. Whereas the 19 new positions could have allowed political neophytes to hold office in Iqaluit, the 1999 election appears to have created space at other points in the political system.

Milbrath (1965) argued that the hierarchy of political participation operated as a pyramid, that most people would vote, still fewer would pay attention to politics consistently and fewer still would hold elected office. This is certainly true in both southern Canadian provinces and in Nunavut. That pyramid, however, would have a slightly different shape in Nunavut, with a narrower base at the bottom and a slightly larger point at the top. If we have reason to believe that political engagement is low in Nunavut, due to the absence of political parties and a daily media, evidence suggests that this is not the case. Whatever dampening effect the absence of political parties might have, this appears to be offset by the greater opportunity for political engagement in the territory. In addition these results suggest that southern notions of political engagement and their dependence on features of southern political life fail to provide much insight into political behaviour in Nunavut. Turnout is higher in the elections without parties than it is in party contests, suggesting that the political parties are insufficiently engaging individuals who would otherwise vote. If age, education and income appear as consistent predictors of engagement, and this confirms existing research on turnout, we are clearly missing part of the picture.

Table 7

Variables included in the regression analysis

Variables included in the regression analysis

-> Voir la liste des tableaux


The institutional architects of Nunavut sought to create a political system that provided a “made in Nunavut” solution while maintaining a remarkable degree of institutional continuity with the previous territorial legislature. By bringing government physically closer to the people, in part by re-shaping the boundaries and in part by pursuing a policy of administrative decentralization, devolution in Nunavut was seen as a way to draw residents back into political life, and to create legislation that better reflected the needs of northern residents. By standard predictors of political behaviour, political engagement should be low in Nunavut. A young population with a small proportion of university degree, the presence of majoritarian elections, continual trips to the polls could all dampen interest and participation. In addition, the perception of a meaningful contest is questionable at the territorial level, without the political cues provided for voters during campaign.

At the federal level, a meaningful contest is hampered by Liberal hegemony. Turnout is declining quickly at that level, and while the model described here suggests that the usual predictors can account for turnout, there is little in the NuHS that would help to probe the full impact of Liberal hegemony. Furthermore, perceptions of the relative importance of municipal, birthright, territorial and federal elections and administrations would help to tease out the various motivations for voting and standing for election that might operate in different ways in elections. Future research on political behaviour in Nunavut would certainly benefit from additional individual-level data that would allow us to compare the predictors of territorial and federal elections. In addition, participation in voluntary activities and other measures of social capital might help to tease out the relationship between social participation, political participation and personal characteristics. At present, however, the greater capacity for “gladiatorial” engagement in Nunavut does not appear to be creating a sense of civic fatigue but rather is providing Nunavummiut with unparalleled opportunities for political participation in Canada. An analysis of patterns of participation in such activities also points to regional variations across the territory, a topic that would also benefit from additional research.

Parties annexes