Political Cleavages in Indigenous Representation: The Case of the Norwegian and Swedish Sámediggis

Abstract Using Lipset and Rokkan’s cleavage theory in an Indigenous context, the article compares political cleavages in internal Sámi politics in Norway and Sweden. The authors discuss the historical legacies of each country’s policies toward the Sámi and subsequent development of party systems before using survey data to analyze cleavages within the Sámi electorates. The analysis shows a prominent cleavage in Sámi politics in Norway regarding the extent of Sámi self-determination, whereas the main cleavage in Sweden can be found in the category-split between reindeer herders and other Sámi created by state policy. Contemporary cleavages in Indigenous politics may, thus, be deeply rooted in nation-building processes.


Introduction
For several decades, Indigenous peoples have been increasingly claiming their rightsespecially to self-determination, to ownership of their traditional lands, and to their own cultures. 1 These developments have resulted in different institutional solutions. 2 In a state where an Indigenous population is territorially concentrated, such as the Inuit in Greenland, which is part of the Danish state, institutions for regional self-government may be an attractive solution. However, this is not a viable option where Indigenous peoples live territorially scattered. One solution is reserved seats in national parliaments, such as the M aori seats in the parliament of Aotearoa/New Zealand. The Nordic states with Indigenous S ami minorities have chosen a different solution. 3 Finland, Norway, and Sweden have each established a S amediggi (S ami Parliament in Northern S ami) consisting of popularly elected S ami representatives.
In the research on Indigenous representation, we may distinguish between internal and external perspectives. Whereas an internal perspective focuses on the politics within an Indigenous people, an external perspective deals with the relationship between an Indigenous people and the state in which they live. This distinction is rarely clear-cut in practice, as several representative Indigenous institutions, such as the S amediggis, unite the internal and external aspects of political representation.
Internally, representative institutions acknowledge the political diversity within an Indigenous population. Indigenous peoples-just like other peoples-comprise individual political actors who may have diverse ideologies, values, agendas, and strategies. Nevertheless, as Nils Oskal points out, Indigenous peoples often face external expectations of conformity. 4 To be perceived as "authentic," Indigenous individuals are expected to conform to preconceived notions of indigenousness. The establishment of the S amediggis breaks with these conformity expectations and institutionalizes the right to political disagreement. 5 Externally, representative institutions, while acknowledging disagreements, channel these disagreements and produce either a majority position or a compromise. Indigenous peoples have experienced marginalization, discrimination, and oppression and often need to speak in a united voice against the authorities of the state in which they live. This is presumably also an advantage for the states, which can deal with one authorized voice instead of several groups that may have conflicting views.
While much of the literature on Indigenous representation concentrates on external perspectives, we focus on internal aspects in this article: Which political cleavages form the basis of Indigenous politics in Norway and Sweden? Based on unique survey data from the S ami Election Studies that were carried out in both countries in 2017, we compare the political cleavages that structure Indigenous politics in two seemingly similar cases: the S ami in Norway and Sweden. As a people, the S ami have been divided during the course of history by national borders (Norway, Sweden, Finland, and Russia), and Scandinavian neighbors Norway and Sweden are generally regarded as similar countries, at least from a wider international perspective. Likewise, the S amediggis are often conceived of as similar and internationally described as models "for indigenous self-governance and participation in decision-making that could inspire the development of similar institutions elsewhere in the world." 6 In this article, our aim is twofold. First, we aim to contribute to our understanding of internal S ami politics in Norway and Sweden by using Seymour M. Lipset and Stein Rokkan's theory of cleavage structures in Western Europe. 7 Accordingly, our starting point is that the political cleavages structuring S ami politics today are deeply rooted in history and specifically in nation-building processes. Second, we aim to explore the relevance of Lipset and Rokkan's approach in an Indigenous context.
In the next section, we introduce our theoretical framework on political cleavages, which focuses on territoriality and the nation-building process. Thereafter, we present the institutional context-that is, the formal position of the S amediggis and their actual autonomy and influence. Next is the empirical analysis, divided into two parts. The first part discusses the historical legacies set by Norway's and Sweden's policies toward the S ami and the subsequent development of cleavages and party systems. In the second part, we use survey data to analyze the fundamental divisions that characterize the S ami electorates in Norway and Sweden, and we discuss whether they correspond to the legacies of the past. Furthermore, we compare the importance of S ami-specific cleavages and the historically dominant political cleavage within the majority societies in Norway and Sweden-namely, the left-right cleavage.

Political cleavages: Lipset and Rokkan in an Indigenous context
Processes of nation-building-together with conflicts emerging from the Industrial Revolution-form the basis of political cleavage structures in Western Europe, according to Lipset and Rokkan's seminal 1967 work, Cleavage Structure, Party Systems, and Voter Alignments: An Introduction. They underlined the importance of history in contemporary politics. Contemporary party systems are deeply rooted in choices made and alliances formed in the past. During nation-building processes, cleavages evolved from the power struggle between the central authorities and the national government on the one hand and local and regional actors on the other, primarily owing to territorial conflicts (the center-periphery dimension). Another cleavage stemming from this process developed out of the conflict between the (secular) state and the church. From the Industrial Revolution evolved two political cleavages: one between the property-owning elite and the emerging bourgeoisie in the cities (which has since transformed into a cleavage between urban and rural areas) and one between capital and labor. This latter cleavage is the foundation of the ideological left-right dimension that characterized European party politics during the 20 th century.
Lipset and Rokkan's theoretical framework has been challenged in recent decades, especially their conclusion that party systems of the 1960s reflected "the cleavage structures of the 1920s"-in other words, that they represented a "freezing of the major party alternatives in the wake of the extension of the suffrage and the mobilization of major sections of the new reservoirs of potential supporters." 8 Although they might have correctly described development in Western Europe until the 1960s, and the political cleavages that they discerned might still be important in most West European countries, other political cleavages have since then become salient. 9 Moreover, their framework has been criticized for not being applicable outside of Western Europe, for instance, in post-communist Europe or Latin America, as the cleavages may not be the same or the party systems and voter alignments may be different. 10 However, as Peter Mair argues, cleavages may wither, but the party alternatives that were based on these cleavages may persist, and party systems-patterns of party competition-may, in turn, be even more persistent. 11 Furthermore, Lipset and Rokkan's general points on the importance of historical cleavages may be valid outside of a West European context, even though the specific cleavages are different.
Lipset and Rokkan's focus on historical processes, especially conflicts between the nation-building center and a country's periphery, seems to be particularly relevant for understanding Indigenous politics. As dominant political elites attempted to build centralized nation-states at the expense of cultural and linguistic minorities, a cleavage emerged between national elites and their bureaucracies on the one hand and local resistance on the other. This cleavage is an intrinsically territorial conflict revolving "around the dispute for political control over a-peripheral-territory inside the state," a conflict that is emphasized in "countries with distinctive ethnic, cultural or linguistic populations." 12 In Lipset and Rokkan's words, nation building involves "the typical reactions of peripheral regions, linguistic minorities, and culturally threatened populations to the pressures of the centralizing, standardizing and 'rationalizing' machinery of the nation-state." 13 However, such territorial-cultural conflicts do not necessarily lead to demands for secession; they could also "feed into the overall cleavage structure in the national community and help to condition the development [ … ] of the entire system of party oppositions and alignments." 14 In contemporary politics, this center-periphery cleavage is manifested, for instance, in and through regionalist parties. 15 This territorial center-periphery cleavage is fundamental to Indigenous peoples' political mobilization and struggles for self-determination because "Indigenous cultures are intrinsically rooted in their traditional territories and dispossession of their traditional lands throughout colonization has had, and continues to have, 'disastrous effects' on Indigenous peoples," depriving them of their political sovereignty, and it has "contributed to dislocation and loss of cultural integrity, language and cultural connection." 16 The cultural, social, and economic lives of Indigenous communities often depend on, and are conditioned by, the natural resources available, and the traditional livelihoods of Indigenous peoples are commonly linked to land and water. 17 In this way, colonial and nation-building processes have been decisive in structuring the relationship between nation-states and the Indigenous communities living within their territorial jurisdictions. Accordingly, the historical and/or contemporary relationship with the state is most likely to impact contemporary Indigenous politics, not only in relation to the state in which Indigenous peoples live but also historical processes of nation building might be decisive for the political cleavages within an Indigenous people.

Institutional context: The two S amediggis
In the next section, we analyze how differences in the historical legacies of Norway and Sweden have affected the cleavage structures of S ami politics in the two countries. However, before we proceed to this historical analysis, we need to describe some institutional differences between the two countries, such as the design of the two S amediggis. These aspects of our two cases are presented in this section.
The S amediggi in Norway was established in 1989, while the Swedish S amediggi followed in 1993, in many ways following the blueprint of its counterpart in Norway. 18 The two S amediggis are institutions for non-territorial autonomy, although property rights and decision-making power over natural resources within the traditional S ami settlement area, S apmi, are topical in the S ami political struggle. Thus, the right to vote is based on ethnicity, not geography, 19 and the S amediggis shall represent all S ami within each country including those who live outside of S apmi. Translating "S amediggi" as "S ami Parliament" can be misleading. The S amediggis have no legislative authority and no independent financial resources, for instance, through taxation. However, although they are state-based and their (delimited) power is delegated from national parliaments, the S amediggis are Indigenous-controlled and enable an amount of Indigenous autonomy (decision making by Indigenous bodies) and Indigenous participation (in decision-making bodies that are not exclusively Indigenous, e.g., consultations with the Norwegian state). 20 Although similar in many respects, there are several important differences between the two S amediggis. 21 First, their formal legal status and position differ. Both parliaments find themselves in an awkward position in combining the roles of a democratically elected assembly and a government agency, but they differ in how these roles are balanced. The S amediggi in Sweden is, to a greater extent, an administrative authority subordinate to the Swedish government, and its legal basis gives it a more limited mandate. The first paragraph of the Swedish S ami Parliament Act states, for instance, that "[in] this Act provisions are made for a special government agency-the S ami Parliament" and its general mission is to "monitor issues related to S ami culture in Sweden." 22 Its Norwegian counterpart has a more autonomous position and a wider mandate: "The business of the S amediggi is any matter that in the view of the S amediggi particularly affects the S ami people." 23 Second, and even more importantly, the S amediggi in Norway has gradually increased its autonomy and influence. 24 It has, for instance, become a mandatory consultative body for the Norwegian state in matters concerning the S ami. The right to consultations does not guarantee any actual influence, and the impact of the consultations varies between issue areas, but consultations have given the S amediggi in Norway at least some influence. 25 The legal basis for consultations is the International Labor Organization (ILO) 169 convention, which Norway-but not Sweden-has ratified. The convention states that Indigenous peoples are entitled to be consulted on matters that affect them, and this has given the Norwegian S ami a platform in international law from which to make demands and claims to the Norwegian state. In Sweden, both the ratification of ILO 169 and right to consultations have been controversial, and a ministerial proposal on a S ami right to consultation presented in 2017 was severely criticized by the S ami and has yet to lead to any legislative changes. 26 Third, in both countries, the national governments have delegated administrative tasks to the S amediggi, but the delegated tasks differ. 27 In the Norwegian case, these tasks include language initiatives, funding to S ami industry, cultural heritage management, and support for sectors such as S ami culture and education, but not issues concerning reindeer husbandry-a traditional S ami livelihood, which is of great importance to the S ami community and identity. In contrast, administrative tasks related to reindeer husbandry are among the most important delegated tasks for the S amediggi in Sweden besides issues concerning language, education, and culture. It is worth emphasizing that the S amediggi's administrative duties concerning reindeer were extended in 2007duties transferred from other government agenciesdespite an earlier decision by the S amediggi plenary not to accept more duties relating to reindeer herding without also having final decision-making power concerning reindeer-herding issues. 28 Finally, two other institutional factors affect S ami politics. For S amediggi elections, unlike other elections in Norway and Sweden, voter registration on a S ami electoral roll is required. This comes in addition to other voting rights: Registered S ami voters retain their right to participate in general elections. Registration on the electoral roll also entails the right to run as candidates for S amediggi elections. As neither state undertakes official registration of S ami ethnicity, potential voters must take the initiative to register. The S ami electoral rolls of Norway and Sweden are based on the same principles. 29 Persons older than 18 can register as voters if they fulfill two criteria. First, registration is a matter of self-identification: Voters must declare that they identify as S ami. Second, there is an objective, language-based criterion: voters or one of their parents or grandparents (in Norway also great-grandparents) must have used S ami as a home language or, alternatively, one parent must be (or have been) registered on the electoral roll.
The electoral systems of both S amediggis are based on proportional representation, but there are some important differences. For instance, elections to the S amediggi and the national parliament in Norway are held on the same day and at the same polling stations. In Sweden, however, these elections are separated, taking place in different years. Partly due to this, turnout is lower in Sweden. 30 In Swedish S amediggi elections, the entire country constitutes a single constituency (with 31 seats), while Norway's S amediggi (with 39 seats) is elected from seven multi-member constituencies. Moreover, voting for individual candidates has no effect in Norway, whereas personal votes are important for electing candidates in Sweden.

Political cleavages in the S ami context
In the first part of our analysis, which is based on secondary material and previous research, we apply Lipset and Rokkan's theoretical framework-in particular, the center-periphery cleavage-to the S ami case. Both similarities and differences emerge when comparing the historical legacies of Norway's and Sweden's policies toward their Indigenous minority populations as part of their respective nation-building processes. In both states, the policy toward the S ami adopted in the second half of the 19 th century was based on the alleged racial and cultural superiority of Norwegians and Swedes. Two ideas were taken for granted: that the S ami neither had any right to selfdetermination nor ownership rights to land, water, and other natural resources. 31 Although the two countries shared this hierarchical ideology, their S ami policies developed quite differently. In Norway, the hierarchical worldview justified a harsh assimilation policy from the 1850s to the end of WWII. 32 The alleged racial superiority of the Norwegians was combined with the strong nationalism of a young state. In Sweden, however, this worldview led to a dual policy of both segregation and assimilation. The S ami were defined as a "tribe" or "race" in need of protection by the Swedish state and-most importantly-were only to have a chance of surviving as a people as nomadic reindeer herders. This is known as the "lapp-shall-remain-lapp" policy, where S ami reindeer herders should be segregated from Swedish society and "civilization." 33 Thus, the S ami in any other livelihood, such as farming or fishing, were to be assimilated. This dual policy of segregation and assimilation created a clear distinction-a "category-split"-between reindeer-herding S ami and other S ami institutionalized in and through legislation. 34 In both countries, the official policy toward the S ami slowly changed after WWII, and the legacy of race biology and cultural hierarchies could no longer justify public policy. This gradual policy change was in many ways influenced by the growing mobilization of the S ami themselves. 35 In Norway, the conflict around the damming of the Alta River in the 1970s and early 1980s put S ami rights on the national political agenda. In Sweden, S ami rights became topical after some reindeer-herding communities and individual S ami sued the Swedish state and claimed ownership of reindeer grazing areas in the Taxed Mountains (Skattefj€ allen). 36 In the early 1980s, S ami political mobilization in these two cases led to the appointment of a S ami Rights Commission in each country and, eventually, to the establishment of the two S amediggis. Today, both countries have recognized the S ami as a people with a right to self-determination, and they both voted for the 2007 UN Declaration on the Rights of Indigenous Peoples (UNDRIP). However, while Norway was the first country to ratify the 1989 ILO Convention Concerning Indigenous and Tribal Peoples in Independent Countries (No. 169), ratification of the convention still seems far away in Sweden. 37 In recent years, we have seen a S ami nation-building process recognized by the Norwegian and Swedish states. A separate S ami political system-with political institutions and a nation-building elite of its own-has been established in both countries. Moreover, a collective S ami identity has been more firmly articulated across the borders of the four states that divide the S ami ancestral lands, in parallel to and influenced by the mobilization of Indigenous peoples around the world. A concrete expression of this nation building is a set of cross-border S ami national symbols-a S ami flag, a National Day, and a national anthem-and work for a Nordic S ami Convention. 38 Thus, the S ami political mobilization manifests a distinct and profound cleavage in relation to the states in which they live, fulfilling all three elements of a political cleavage according to Stefano Bartolini and Peter Mair's definition. 39 First, there ought to be an actual social difference between groups of people (an empirical element). Second, these groups should have a sense of belonging and a shared value community, that is, a collective identity, based on their awareness that other groups have different and often conflicting interests (a normative element). Third, this collective identity should also lead to political mobilization and organization into political parties, interest organizations, and lobby groups within the established political system or in various social movements that actively challenge the current political order (an organizational/behavioral element). The S ami demands for self-determination and land rights are clearly based on a collective identity consisting of both an empirical and a normative element, where the states represent opposing interests. Moreover, this cleavage has been politically institutionalized in movements, organizations, and S amediggi parties.
Such developments, however, may cause a counter-reaction. The extent of S ami selfdetermination can be a contested issue among the S ami themselves, not only among the Norwegian and Swedish majorities. The S ami people are territorially scattered. For some, the historical experience of oppression and assimilation policies, and the lasting damages inflicted on S ami self-determination, property rights, language, and culture by these policies, as well as present-day racism, are of extreme importance. From this perspective, reclaiming S ami rights, culture, and identity becomes crucial. For other S ami, the S ami identity can be less important than the local or regional identity that is shared with one's non-S ami neighbors. From this perspective, self-determination may be less important or even seen as creating unnecessary and problematic ethnic divides. Today's S ami politics may, therefore, be described as an intersection of two processes of nation building: the older Norwegian/Swedish and the newer S ami.
Thus, self-determination could be a dividing issue within the S ami electorate in both countries. Another dividing line within the S ami society in Sweden-also stemming from the historical legacy of state policy-is the category-split and creation of two categories of S ami with different rights, which is still evident within S ami politics and society, as discussed below.
The organizational element of political cleavages: a "double" and a "single" party system Bartolini and Mair's third criterion-the organizational element of a cleavage-necessitates further discussion. 40 Here, the question is whether the party systems of the two S amediggis-which emerged when these institutions were established in the late 1980s and early 1990s, respectively-reflect the frames set by historical developments in each country.
It should be noted that not only the historical context but also the choices made by S ami political actors were different in the two countries. The outcome was two very different party systems. In Norway, S ami politics involves a "double party system." First, several S ami parties and candidate lists participate. Some are national S ami organizations that field candidates in all or most constituencies. Others run for election in a single constituency. Second, traditional Norwegian parties also participate in S amediggi elections in Norway. Norwegian parties were established political actors in the traditional S ami settlement area long before the S amediggi was founded, with S ami members and S ami municipal councilors. When this new political arena was established, participation in S amediggi elections was a small step.
Two main competitors have dominated Norwegian S ami politics since the S amediggi was established. One is a Norwegian party, the Labor Party, and the other a S ami organization, the NSR (the Norwegian S ami Association, Norgga S amiid Riikkasearvi). The NSR had the majority of seats in the S amediggi until 1997, but neither party has had a parliamentary majority since. The two main competitors, thus, must seek support from smaller parties and lists to form a parliamentary majority. Lina Gaski highlights that Labor and the NSR have had very different histories. 41 The Labor Party governed Norway for much of the postwar era and was responsible for implementing the assimilation policy toward the S ami. The NSR was originally a S ami cultural and political organization working for S ami rights and for the revitalization of a S ami identity long before the S amediggi was founded, often with the Labor Party as its main opponent, for instance, during the conflict over the damming of the Alta River around 1980. Historically, therefore, the NSR and the Labor Party were antagonists regarding the selfdetermination cleavage. In present S amediggi politics, this is no longer the case, but earlier voter studies nevertheless described differences between the two parties' voters on this issue. 42 Besides the Labor Party, three other parties from the national Norwegian party system participate regularly in S amediggi elections: the Progress Party, the Conservative Party, and the Center Party. While Labor, the Conservatives, and the Center Party belong to the mainstream of S ami politics in Norway, the right-wing populist Progress Party's S ami policy differs radically from those of all other parties and candidate lists. According to the Progress Party, the S amediggi is a case of ethnic discrimination, and its position is that the S amediggi should be abolished. Besides the NSR, the largest S ami parties are Arja (the Northern S ami word for "commitment") and Nordkalottfolket (People of the Northern Cap). While Arja's policies are quite similar to those of the NSR, Nordkalottfolket aims to represent the mixture of people-S ami, Norwegians, and Kvens (descendants of Finnish immigrants)-who inhabit Northern Norway. Furthermore, the electoral system enables representation of candidate lists with a limited geographical focus, only participating in a single or a few constituencies. For example, there are lists that represent minorities within the S ami society, such as the Southern S ami community, or some specific interests within a region (e.g., reindeer herders).
A potential cleavage in Norwegian S ami politics is the urban-rural divide. However, this issue has not become party politicized, as no major party unambiguously prioritizes either rural S ami communities or urban S ami. 43 In elections to the S amediggi in Sweden, only S ami parties field candidates. The traditional Swedish parties stay out of S ami elections, although no legal barriers prevent them from participating. There is, however, a tradition within the S ami community in Sweden-dating back to the S ami National Conference in 1918-to adopt a neutral attitude to Swedish politics to avoid internal division. Thus, historically, the S ami have not tried to influence Swedish politics through the traditional national political parties. 44 The S ami parties in Sweden represent many different interests within the S ami community. Some of them were founded on the basis of S ami civil society associations, such as organizations for reindeer herders, non-reindeer herders, or youth organizations. 45 The party structure has been quite stable over the years, although the 31 seats in parliament have been distributed among 6 to 11 parties. Seven of the nine parties represented in the S amediggi since the last election in 2017 have been represented in parliament since the first election in 1993. 46 The main division within the S amediggi in Sweden separates parties representing the reindeer herders from those representing other S ami interests. This division reflects the state policy established in the early 20 th century (see above), when a system of S ami rights was constructed around reindeer husbandry excluding all S ami involved in other S ami livelihoods, like fishing, hunting, and handicraft. Thus, the state policy created two categories of S ami with different standings in relation to the system of rights, including the right to hunt and fish on "Crown land"-members of reindeer-herding communities and S ami outside of reindeer herding-which is upheld in contemporary legislation. 47 This division could be described as a conflict between institutions: which political or organizational body should represent S ami interests vis-a-vis the state on issues concerning reindeer husbandry? Should, for instance, the right to predator compensation for lost reindeer be decided and dealt with by the S amediggi or by the reindeer-herding communities directly representing the reindeer owners?
Since the first S ami parliamentary election in 1993 up until 2009, parties representing the reindeer herders' interests were the majority in the S amediggi. Dominant among these parties-especially during the first three parliamentary terms-was S amid Riikkabellodat, a party that developed out of S amiid Riikkasearvi, the first national S ami organization established in 1950 with its base in the reindeer-herding communities. Yet this party has never had a majority of seats and has relied on coalitions with other parties to form a majority, such as Guovssonn asti, with its origin in the reindeer owners' organization (Ren€ agarf€ orbundet), and Vuovdega-Skogssamerna, representing the forest reindeer-herding S ami. However, the largest party in the S amediggi during the most recent terms has been Jakt-och Fiskesamerna, primarily representing the interests of the S ami not engaged in reindeer herding but in hunting and fishing. This party works closely with Landspartiet Svenska Samer, originating from the largest organization representing S ami outside of reindeer herding, and Albmut. A previous study showed that there is great affinity between the voters of these two blocs. 48 In between these two party blocs, two parties have been in coalition with both sides over the years to form a majority: Min Geaidnu and S amit/Samerna. In the following, we turn to survey data to analyze three potential political cleavages within S ami politics in Norway and Sweden. Based on our above discussion on processes of nation building and, more specifically, the states' policies toward the S ami, we explore two S ami-specific potential cleavages: self-determination and (in Sweden) the category split. However, it is also possible that the dominant cleavages in the majority societies in both Norway and Sweden have left their mark on S ami politics-especially in Norway, where traditional Norwegian parties participate. To explore this possibility, we include the traditional left-right dimension, which originates in the cleavage between labor and capital. As described by Lipset and Rokkan,49 the conflict between capital and labor developed during the Industrial Revolution and left its mark on all West European countries. Traditionally, this has been the major political cleavage in both Norwegian and Swedish politics. 50

Survey data
Using data from the 2009 S ami Election Study in Norway and the 2013 S ami Election Studies in both countries, earlier studies suggested that the cleavages discussed above are also reflected in voter attitudes. 51 In this article, we use data from the 2017 S ami Election Studies that were carried out in Norway and Sweden. Both election studies are based on random samples of registered voters drawn from the electoral roll. Permission for this sampling was granted by both S amediggis, and the studies were subject to review by the ethical review boards in both countries.
The study in Sweden was conducted through a combination of postal and web questionnaires in three languages (Northern and Southern S ami and Swedish) with four reminders: two via postcard and two via telephone message. Enk€ atfabriken, a polling agency specialized in surveys in cooperation with scholars, conducted the fieldwork. The response rate was 43%. The study in Norway was conducted through a combination of postal and web questionnaires as well as a follow-up phone call to nonrespondents, who were given the option of responding over the phone. Questionnaires were available in four languages-the same as in Sweden and additionally in Lule S ami. A commercial polling agency called Kantar conducted the fieldwork. Despite having the additional option of replying to the questionnaire over the phone, or perhaps because fewer reminders were used, the response rate was only 21.3%. The main reason for the difference in response rates between the two countries is probably that people are used to rejecting queries from commercial polling agencies. These agencies often get even lower response rates than those obtained in the Norwegian S ami Election Study. We use demographic weights to compensate for the effects that this may have had on the final sample. 52 The results for two minor parties in Sweden and several minor parties in Norway are not presented because these parties had too few respondents for statistical analyses.

Analysis of the survey data
In this section, we use survey data to analyze and compare three potential political cleavages in S ami politics in Norway and Sweden: self-determination, the category split, and the left-right scale.

Self-determination
As we have previously shown, 53 S ami voters in Norway seem to have quite coherent attitudes toward the issue of S ami self-determination, and a factor analysis of the 2017 data confirms this. We use six survey items to create an index of people's attitudes toward self-determination. 54 The items have a Cronbach's alpha of .81, indicating a high degree of coherence across items, making them suitable for index construction. We standardize the index from 0 to 10, with the lowest values indicating strong support for the S amediggi specifically and for self-determination generally. Figure 1 depicts the overall distribution of voter attitudes in the Norwegian case as well as the average values of each party's/list's voters. The overall mean for all voters is 3.3. Most voters lean toward supporting increased self-determination, as can also be seen from the distribution in Figure 1. This is not surprising, as it is the position of the vast majority of parties represented in the S amediggi. There is nevertheless quite a bit of variation among voters of the different parties in this regard, which indicates that self-determination is a salient cleavage for voters in Norway's S amediggi elections. Those voting for either the NSR or Arj a strongly support self-determination, followed by the Labor Party and the Center Party. A difference between the two traditional opponents-the NSR and Labor-is, thus, still visible, but today, the difference is only one of degree. The Conservative Party and Nordkalottfolket are closer to the center of the scale (value 5). This is to be expected in the case of Nordkalottfolket, which aims to represent the mixture of people in Northern Norway rather than specifically S ami interests. The Progress Party stands out as the only party where voters are rather skeptical about self-determination. This is not surprising, as the party favors disbanding the S amediggi. The Progress Party is relatively small, with 7.5% of the votes and just one seat in 2017, but has established itself as a permanent opposition to mainstream S ami politics in Norway.
On the Swedish side of the border, we find some similarities and some differences. Looking at voter attitudes, a factor analysis reveals a distinct self-determination dimension in Sweden as well. The seven questions that make up this index have a Cronbach's alpha of .86, suggesting that this is a coherent attitudinal dimension. 55 However, the party positions do not vary as much as in the Norwegian case. Figure  2 shows that all parties' voters are at the lower end of the scale and, thus, in support of increased self-determination for Sweden's S ami, and there is no equivalent to the Progress Party. The overall mean for all voters is 2.8. There are nevertheless some minor differences between the parties. Jakt-och Fiskesamerna's voters are somewhat more skeptical about self-determination and the S amediggi than the others, while Min Geaidnu's voters are the strongest supporters of self-determination. The difference between the other parties is small.

The category split
The main difference between voters in Norway and Sweden is that there exists a unique and distinct attitudinal dimension in Sweden, which plays an important political role. Previous analyses showed that the status of reindeer-herding communities constitutes   an important cleavage-and especially the rights to hunt and fish attached to membership in a reindeer-herding community-in line with our discussion of the category split between reindeer herders and other S ami, which originates from the history of Swedish policy toward the S ami. 56 The organizational structures of reindeer husbandry on the Norwegian side of the border play a much less influential role. There is simply no comparable organizational structure on the Norwegian side of the border. As there are no specific rights attached to reindeer herding, there is no comparable issue on the political agenda in Norway. Although conflicts between reindeer herders and other S ami may be found in Norway, they are much less institutionalized than in Sweden. Accordingly, this topic was not addressed in the Norwegian survey.
Conversely, in Sweden, this issue is certainly on the agenda. Some wish to preserve and extend the status and decision-making power of reindeer-herding communities, while others are critical of their unique role. This dimension also appears in a factor analysis. The four questions in the Swedish voter survey about reindeer-herding communities have a Cronbach's alpha of .69, indicating that they are closely related, though not as strongly as the self-determination questions. Thus, we use these four questions to create an index. 57 The distribution and average values of each party's voters are shown in Figure 3. Clearly, voters hold differing opinions in this regard and there is quite a bit of variation among the parties.
The voters of S amid Riikkabellodat and Guovssonn asti stand out as the strongest supporters of the reindeer-herding communities. As this is the policy of these parties, and they grew out of the reindeer-herding communities and reindeer owners' organization, this is no surprise. On the other side of the scale, we find, as expected, the two parties that most strongly promote the interests of the S ami outside of reindeer herding: Jaktoch Fiskesamerna and Landspartiet Svenska Samer. The three remaining parties are located closer to the center of the scale.

The left-right scale
So far, we have considered cleavages based on issues that are unique to the S ami. Now, we investigate a cleavage that dominates the politics of the majority populations of Norway and Sweden: the ideological left-right dimension. We use voter self-placement on a 0-10 scale, from left to right, to assess the impact of this cleavage within Norwegian and Swedish politics.
As Figure 4 shows, in Norway, the average voter placements also vary with respect to this dimension. The parties that participate in Norwegian national politics adopt their well-known positions. The Labor Party is on the left, the Center Party is in the center, whereas the Conservative Party (Høyre) and the Progress Party are on the right. The fact that these national political parties run in S amediggi elections is what makes the left-right dimension relevant. The S ami lists/parties are all positioned on the center-left, but NSR voters are slightly more left-leaning than the others.
The left-right scale plays an insignificant role in S amediggi elections in Sweden ( Figure 5). The average voter has a centrist position and party differences are minimal. Voters from all parties have a center or center-left position, similar to the specifically S ami parties on the Norwegian side of the border.

Concluding remarks
In this article, we aimed to analyze internal S ami politics using Lipset and Rokkan's cleavage theory. 58 The analysis of historical developments and survey data confirm that the political cleavages that structure S ami politics in Norway and Sweden are very different, both when the party systems are compared and when survey data are used to look at the voter level.
First, the massive support for self-determination among the electorates in both Norway and Sweden confirms that the establishment of S amediggis has not reduced the territorial center-periphery conflict. The S ami-state cleavage is still most vivid. A clear majority of the S ami electorate favors increased self-determination in both countries, but the extent of self-determination stands out as more of a party cleavage in Norway. This is partly a consequence of the participation of traditional Norwegian parties in Norwegian S amediggi politics. Whereas the Labor Party long ago abandoned its traditional skeptic position on self-determination, another Norwegian party-the Progress Party-really stands out with a deviating position regarding both party policy and voter attitudes. As mentioned above, the fact that the Norwegian S amediggi is more influential than its counterpart in Sweden may make self-determination a more contested issue in Norway. However, another explanation may be that the opponents of self-determination are possibly more politically integrated in Norwegian S ami politics: They have chosen to register on the electoral roll and participate in elections. Presumably, the Progress Party has mobilized its supporters to register and vote. Moreover, having elections for the S amediggi and the Norwegian national parliament on the same day may be important because this makes it easier to cast a vote at the same time for both elections, even if one is not interested in, or strongly against, an increase in the S amediggi's power.
However, a previous analysis indicated that there could have been potential for a stronger self-determination cleavage in Swedish S ami politics if the Swedish parties had participated in S amediggi elections. 59 Those S ami who vote for the Sweden Democrats (right-wing populists) in general elections also stand out as somewhat more skeptical about self-determination. Moreover, that self-determination may become a major cleavage structuring S ami politics in Sweden is also indicated in studies on how social integration in S ami society affects voter behavior and attitudes. Of importance in this context is the strong relation between opinions about S ami self-determination and social integration in the S ami society: For every unit increase in social integration in the S ami society, the stronger claims for self-determination become. 60 Second, the category split between reindeer-herding communities and other S ami is the dominant cleavage in Swedish S ami politics, but is almost absent in Norway. This split follows from Sweden's historical legacy, created by the earlier (and contemporary) S ami policy of the Swedish state, and it structures the party system of the S amediggi. From this perspective, the Swedish policy may be described as a divide-and-rule strategy-even if not a conscious tactic-by "fomenting divisions among subjugated groups by sowing mutual mistrust." 61 Third, the left-right dimension appears to be important in Norway, but not in Sweden. This is clearly a result of the participation of traditional Norwegian parties in S ami politics, as the left-right scale does not distinguish between the specifically S ami parties in Norway or Sweden. The left-right dimension may matter for voters: Those with less knowledge of S ami politics can vote for a Norwegian party based on their Norwegian party identification. However, the left-right division is not necessarily important at the party level in practical S amediggi politics. Many issues that divide parties along a left-right axis in national politics, such as taxation and welfare programs, are not on the agenda of the S amediggis. Moreover, as Figure 4 shows, the voters of the two main competitors in Norwegian S ami politics, the NSR and Labor, are close to each other on the left-right scale. When the NSR governing council was brought down by a vote of no confidence in 2016, it was replaced by a coalition of three parties that are widely dispersed on the left-right scale: Labor, the Conservatives, and Arja. Therefore, we regard the left-right dimension in S ami politics in Norway as a reflection of the participation of traditional Norwegian parties rather than a cleavage of importance in practical politics. Accordingly, we conclude that S ami politics in both countries has been shaped mainly by the unique historical experience of the S ami, rather than reflecting cleavages of the majority population.
Our second aim was to explore the relevance of Lipset and Rokkan's approach for our understanding of Indigenous politics today. We believe that our analysis clearly shows the importance of Lipset and Rokkan's perspective, especially their starting point that contemporary political cleavages are deeply rooted in nation-building processes. The historical legacies of states' Indigenous policies seem to be decisive for understanding the contemporary politics of representative Indigenous institutions. This demonstrates the importance of external perspectives when we study the internal politics of Indigenous peoples. In our analysis, self-determination is an obvious divide, with roots in the territorial center-periphery cleavage, as evident in the S ami-state divide. As in Norway, this divide may also become a (party) cleavage within the Indigenous community. Moreover, the differences found in comparing the S ami in Norway and Sweden also indicate that the state might create profound cleavages within an Indigenous society. Such effects have been produced by policy measures-harsh assimilation in Norway against all S ami and the dual policy of assimilation and segregation in Sweden directed toward different groups of S ami-as well as through legislation creating different categories of rightsholders.