In their classical analysis of cleavage structure, party systems and voter alignments, Lipset & Rokkan (1967) suggest that the crucial cleavages and their political expressions can be ordered within a two-dimensional space constituted by a territorial dimension, or centre-periphery axis, and a functional dimension, or cross-local axis. In this article we will concentrate on the first of these two dimensions. Thus, the main focus of the article is the relationship between centre and periphery in Norwegian politics.

Centres can be defined as the privileged locations within a territory where key resource-holders most frequently meet. A centre controls the bulk of transactions among resource-holders across territorial boundaries, and it is able to dominate the communication flow through a standard language and a set of institutions for regular consultation and representation. A periphery, by contrast, is dependent, controlling only its own resources at best, and it is isolated from all other regions except the central one (Rokkan & Urwin 1982). In other words, centres are powerful, rich and influential; peripheries are powerless, poor and influenced (Anckar 1991).

In their analysis of centres and peripheries in western Europe, Rokkan * The data used in this article have been made available by the Norwegian Social Science Data Services, the Central Bureau of Statistics of Norway and by Professor Gudmund Hernes. The author expresses his thanks to Anne Liv K. McKenzie for her assistance and to anonymous reviewers for their comments. An earlier version of the article was presented at the XVth World Congress of the International Political Science Association, Buenos Aires, 21-25 July, 1991.

Resumé

The overall impression of Norway given by Rokkan and Urwin in their classification of statebuilding nation-building processes in western Europe is characterized by the domination of a strong centre over weak peripheries. In a comparative perspective this may be the most conspicuous tendency. The influence of the Norwegian periphery, however, is quite substantial. The aim of this article is to contribute to a more nuanced picture of centreperiphery in Norwegian politics. It takes a closer look at the representation of the periphery in the Norwegian parliament, and at the impact of geography on policy-making and other activities of MPs.

& Urwin (1982) have developed a typology of nation-building and statebuilding basedon space identity and strategies of unification. Political movements tend to identify with both a territory and a group. Hence we have, according to Rokkan and Urwin, to deal with two interrelated dimensions - 'membership space' which means membership of a group that possesses some common socio-cultural stigmata, and territorial which means identification with and occupation of a specific geographical area. Nation-building is thus considered as a continuum which may be divided into four categories: (1) territorial space predominant; (2) territorial space dominant, but with strong membership space characteristics; membership space dominant, but with strong territorial space characteristics; and (4) membership space predominant. Unification or state-building may be basedon centralizing strategies or federalizing accommodation. strategies can be dichotomized, giving a fourfold classification: unitary state, the union state, mechanical federalism, and organic federalism.

According to Rokkan and Urwin, Norway is placed at one extreme of both dimensions. Territorial space is predominant. There is an unambiguous point of control with only weak or no regional institutions. Where political protest does arise, it tends not to have any strong positive regionalist or peripheral perspective, since there is little or no distinctive identity other than with the state. Norway is also a unitary state with one unambiguous political centre which enjoys economic dominance and pursues a policy of administrative standardization. All institutions are under direct control of the centre.

Why, then, has Norway not experienced serious regional unrest? The question is raised by Aarebrot (1982) who has analysed the Norwegian case in Rokkan and Urwin's book. He offers three explanations: (1) the regional level has very little institutional infrastructure; (2) cultural identities and aims are not tied to specific regions; and (3) the basis for economic development has emphasized the local establishment of industries rather than regional. In other words, a 'missing regional link' is the reason why 'fully-fledged regional political problems' have not emerged in Norway.

The overall impression of Norway given by Aarebrot, Rokkan and Urwin, in short, is characterized by the domination of a strong centre over weak peripheries. In a comparative perspective this is the most conspicuous tendency. However, this 'centralistic' picture of Norway needs modification. may be argued that local and regional identities do exist, and the influence of the Norwegian periphery is quite substantial. The regional link is not really missing. It is probably more adequate to say that the interests of the Norwegian periphery are integrated into the institutional infrastructure of the unitary state.

It is, of course, not possible to give a comprehensive analysis of this

question within a short article. The aim here, therefore, is to provide some empirical illustrations that may contribute to a more nuanced picture of centre-periphery relations in Norwegian politics. In the first section of the article we will examine the representation of the periphery in the Norwegian parliament, the Storting, at two levels - the county level and the municipal level respectively. In the second section we will take a closer look at how district interests influence committee evaluations, decision-making and questions in parliament, and the MPs' contact with the administration. As we shall see, representation of the periphery in the Storting is strong, and district interests have a substantial impact on policy-making and other activities of the Norwegian parliament.

Representation of the Periphery

The first political parties in Norway, the Liberal Party and the Conservative Party, were established in 1884. They adopted the Norwegian names Venstre and Høyre, which literally mean left and right. But to a large extent the two parties also represented the periphery and the central areas respectively. Roughly speaking, there are two peripheries in Norway, one in the southern and western parts of the country, and one in northern Norway (Rokkan 1967). Traditionally, the Labour Party had a stronghold in the northern periphery, while the so-called middle parties, the Agrarian Party (later called the Centre Party), the Christian People's Party, and the Liberal Party had their strongholds in the southern and western periphery. The Conservative Party had its strongest support in the central parts of eastern Norway, especially in the capital, Oslo. To a certain degree, this is still the case.

We can also see the impact of the centre-periphery dimension on party support at the municipal level. As shown in Table 1, support for the Centre Party and the Christian People's Party decreases as one moves from outlying to central municipalities, while support for the Conservatives and the Progress Party increases.

The County Level

The Norwegian constitution of 1814 established two kinds of constituencies cities and rural districts - within each county. As Rokkan (1967) has pointed out, provisions of the constitution ensured urban overrepresentation.The was enfranchised but kept within bounds: one vote in the cities weighed as much as five votes in the countryside. An increase in the number of registered citizens in a constituency resulted in a proportional increase in its number of representatives. However, the

constitution stated that the ratios were to be changed if the number of city
representatives exceeded one-half of the number of rural representatives.

This provision was not taken very seriously in the first decades after 1814. The gradual increase in the population of the cities produced a more and more marked överrepresentation of their electorates, but action against this injustice was not taken until the peasants begun to mobilize their forces in opposition to the dominance of the cities. A slow process of rural mobilization started in the 1830s. By 1859 the number of representatives from the cities went far beyond the original 1:2 ratio and the peasants finally took action to defend their rights, seeing to it that the 1:2 urbanrural was strictly enforced.

Originally, the 1:2 ratio served to ensure urban överrepresentation. But after 1859 the 'peasant clause', as the rule came to be called, had the effect of defending the rural interests in an increasingly urban society. Yet even in 1885 the cities, with the exception of the capital, were still markedly overrepresented (Rokkan 1967). Only in 1952, when the urban and rural parts of each county were merged into one constituency, was the 1:2 rule abandoned.

At present constituencies for Storting elections are identical with the 19
Norwegian counties (see Figure 1). The number of representatives from
each county roughly corresponds to the number of inhabitants. But the

distribution of seats among the counties is still by no means strictly proportional.Comparedwith number of citizens, central counties are now underrepresented, while sparsely populated and peripheral counties are overrepresented in the Storting. In the 1969 election, for example, there was an average of 14,391 valid votes cast per representative on the national level. In the two centrally located counties of Akershus and Oslo, however,

the number of votes cast per representative amounted to 24,958 and 23,798 respectively. In Finnmark and Nordland in the northern periphery, by comparison, there were only 9096 and 9996 votes cast per representative. If we say that the average voter on the national level has one vote, it means that each Finnmark voter had 1.6 votes, whereas each Akershus voter had 0.6 votes.

Most Norwegian politicians have accepted the principle that long distances the political centre in the Norwegian capital Oslo should be compensated by a certain överrepresentation of the periphery. But over time, due to population growth in the central areas, the degree of underrepresentation the centres became unacceptable. Thus, in 1973 the total number of seats in parliament was increased from 150 to 155, providing Akershus and Oslo with three and two new mandates respectively.

Because of continuing population growth, Akershus was subsequently granted two more seats in the 1985 election. At this point there were an average of 16,572 valid votes cast per representative on the national level. In Akershus, however, there were 20,669 votes per representative, while the corresponding number in Finnmark was 10,578. In view of this situation,

we may say that each Akershus voter had 0.8 votes in this election, while
each voter from Finnmark still had 1.6 votes.

Regional differences between the parties further contribute to the unproportional of mandates between the counties. To make the distribution between the parties somewhat more proportional, it was decided the 1989 election that eight additional seats should be elected on a nation-wide basis. Since the parties entitled to the additional seats tend to gain most of their 'extra' votes in densely populated counties, this arrangement also has an impact on the geographical distribution of the mandates. In 1989 Akershus got three of the at-large seats, Rogaland 2, Oslo 1, Østfold 1, and Hordaland 1.

The effect is shown in Table 2. In the first column we find the actual number of votes per representative in each county. In the second column we have estimates based on the hypothetical situation where no additional at-large seats were elected, and the third column provides an index of under-/overrepresentation for individual districts based on the situation with additional seats.

As we see, the national average in 1989 was 16,046 votes per representative. the arrangement with additional seats, 17,569 votes were cast per representative from Akershus in 1989, whereas had there been no additional seats, the number of votes per representative from Akershus would have been 21,961. In Finnmark, by comparison, which did not get any additional seats, the number of votes per representative was 10,261. Thus, each voter from Akershus counted for only 0.9 votes, while the voters in Finnmark had 1.6 votes each. It should be noted, however, that in the 1989 election Buskerud, a county to the southwest of Oslo in which Drammen, Norway's sixth largest city is located, was the most underrepresented county since in this county each vote counted for only 0.8 of a vote (20,514 votes per representative) as compared to the national average.

Clearly, the distribution of the additional seats has an impact on the geographical distribution of the mandates in the Storting. But the additional seats are only distributed to parties which, after the distribution of the county seats, are underrepresented in view of their overall support on the national level. Consequently, the geographical effect will vary from one election to another. Moreover, even though the underrepresentation of the central counties has been reduced over the years, Table 2 indicates that the peripheral counties are still overrepresented. All of the seven counties at the bottom of the list (i.e. those which are most overrepresented) are the most peripheral counties in southern, western and northern Norway.

In the 1989 election the periphery was not only overrepresented in
relation to the size of the population, the utmost periphery was also
represented directly. In this election, for the first time since 1918, a

non-partisan candidate, Anders Aune, was elected to the Storting. Aune represents the 'Popular Movement for the Future of Finnmark', a constellationformed order to draw attention to the serious problems facing the coastal areas of northern Norway. Anders Aune was also an MP in 1969-73, representing the Labour Party, but he subsequently left the party in protest against government policy regarding Finnmark.

The Municipal Level

The Norwegian constitution originally stated that candidates for parliamentary had to be residents of the constituency in which they were nominated. This requirement was abolished in 1952, but still most candidates have their residence in the county they represent. In 1985 only one percent of the candidates lived outside the county in which they were nominated (Valen 1988). In 1989 the strong tradition for nominating candidates who live within the district even caused problems for a national party leader. Erik Solheim, the leader of the Socialist Left Party, lives in Oslo. But the party's 'secure' Oslo seat was already occupied by a former party leader. Hence, Solheim ran for nomination in Sør-Trøndelag. He succeeded, but many of the local politicians protested strongly.

The peripheral component in Norwegian politics is also evident in the nomination process which, in the Norwegian parties, is highly decentralized. party leaders are very reluctant to interfere with this process. Several criteria for group representation are involved. Among the most important ones are district representation, representation of women, of youth, and of different occupations (Valen 1988). Probably the most important one when the parties compose their lists of candidates, however, is geographical balance. Most often the top candidates of a party live in different municipalities within the county. Even so, the concentration of voters in county centres tends to contribute to the nomination of at least one candidate from the central areas of the county on each party list. Since some of the smaller parties will win only one seat in the election, överrepresentation of the central municipalities could quite naturally be the result.

The map in Figure 2 shows the actual distribution of MPs according to the municipality in which they lived at the time of the 1989 election. Three or more MPs were resident in the municipalities with the darkest colour. No MPs lived in the municipalities shown in white. With a total of only 165 MPs and 448 municipalities, it is obviously impossible for all municipalities to be represented in the Storting. But as we can see, municipalities with one or two representatives are scattered all over the country. And insofar as there are municipalities with three or more representatives, these are indeed most often the largest towns in each county.

Yet this situation does not automatically imply an överrepresentation of the centre. The distribution of representatives and population according to the centrality of the municipality is shown in Table 3. As we see from the table, the suggested överrepresentation of the centrally located municipalitiesis

palitiesisby no means confirmed. On the contrary, the most centrally located municipalities are the ones most likely to be underrepresented. The main conclusion to be drawn, however, is that the distribution of MPs and that of the population as a whole are almost identical. The situation in 1989, we may add, is no coincidence. Hellevik (1969) found the same pattern when analysing the period from 1945 to 1969.

Geography and Policy-Making

As we have seen, the geographical background of the candidates plays an important role in the nomination process. The reason, of course, is the supposed effect it has on policy-making. Candidates are expected to promote interests of their own districts. Experience from local politics is probably one of the factors strengthening the connection between geographical and policy-making.

Most of the elected representatives in the Norwegian Storting have experience from local politics. Of the MPs elected in 1989 23 percent have been either mayor or vice mayor in their own municipality or county. A total of 85 percent have served in some kind of local office (Kommunal Rapport, no. 15, 1989). These figures are quite representative for the postwar as a whole, even though the extent of local political experience has been slightly reduced. According to Eliassen (1985), only 12 percent of all MPs elected during the period from 1945 till 1985 had no experience from local offices. Of these MPs 30 percent had been mayors in their own municipality, and an additional 58 percent had been members of the municipal council, while 25 percent had been members of the county council.

The importance of the geographical background of Norwegian MPs has its symbolic expression in the plenary meeting room in the Storting. Goodsell has argued that parliamentary buildings preserve cultural values of the polity, articulate contemporaneous political attitudes, and contribute to the formation of political culture. He quotes Winston Churchill's saying, We shape our buildings, and afterwards our buildings shape us', and points out that the Westminster-style opposition seating, across two sword lines woven into the carpet of the floor of the House of Commons, must invite an atmosphere of confrontation between government and opposition. An alternative semicircular arrangement can also facilitate ideological debate, provided that seats are ordered in a partisan array.

None of these arrangements are found in the Norwegian parliament. The geography of the plenary chamber of the Storting corresponds to the geography of the country. MPs take their seats on county benches, on which they are seated according to the distribution of party mandates. As a result, one finds political opponents from the Labour Party and the Conservative Party sitting together on the Akershus bench. Likewise, on the bench for Troms in northern Norway, the right-wing MP from the Progress Party sits next to the MP from the Socialist Left Party (see Figure 3).

The importance of the geographical background of the MPs is also illustrated by the distribution of the additional seats in the plenary chamber. This distribution caused some problems in the Storting. MPs holding the additional seats are primarily supposed to represent their parties, not the county. The logical, and most convenient, solution would have been to place the additional seats on a bench separate from the county benches. However, parliament chose to place the MPs holding one of the additional seats on the benches of the counties they come from. This is done in spite of the inconveniences it causes. Since the geographical distribution of the additional seats will vary from one election to the next, signs on the chairs showing the number and county of the representatives must according to this rule be moved around after each election.

It is quite possible that the seating arrangement in the Norwegian Storting has contributed to the widespread consensus which has characterized Norwegian politics in much of the post-war period.¹ When political opponents share a two-seated chair during debates, they have considerable time for informal communication, reducing the political tension between them. This communication may reduce the conflicts between the parties, and strengthen the cooperation between county representatives across party cleavages. The combined effect of this, and of the heavy emphasis placed on the candidates' experience from local politics in the nomination process, may represent a strengthening of the political impact inherent in the geographical background of the MPs.

Standing Committees

Standing committees play an important role in the policy-making process in the Norwegian parliament. All issues to be decided by the Storting are first submitted to the relevant standing committee. The committee makes a recommendation on which the debate in the plenary assembly is based. As Hernes has pointed out, the committees are more than just the first stage in the decision-making process:

they are, as a point of fact, the only stage at which issues can be scrutinized and worked over in detail; where new information can be brought to bear and where it can have an effect; where alternatives can be considered and their consequences evaluated; where outside groups can be heard and their opinions examined; and where changes in a bill or modifications in proposed appropriations can be made to exaggerate the point - one could say that not only is the committee the first stage in decision-making in the Parliament: for most practical purposes it is also the last (Hernes 1971, 68-69).

On several occasions, Gudmund Hernes has asked the MPs to evaluate
the different committees. Three questions are of special interest in this

context: (1) Of which committees do they prefer to be a member? (2) Which committees are most important for carrying through the party programme? and (3) Which are most important for the promotion of district interests? Each MP was asked to name three committees on each question. The results from 1977 are presented in Table 4.

Not surprisingly, the committee for transportation gets the highest ranking the question concerning district interests. It is, by contrast, ranked as number seven concerning the importance for the fulfilment of the party programme. The committee for municipal affairs and environment policy is ranked second on district importance and third on party programme importance, while the committee on industrial affairs has the opposite ranking. Most important for carrying through the party programme is the finance committee. On the district issue, the finance committee is ranked in the fifth place only.

Committee ranking according to which one the MPs prefer to be a member of seems to lie somewhere in between the district and party programme rankings. The committees for municipal affairs and for industrial are most preferred, followed by the committee for transportation and the finance committee.

In Table 5 we have compared the different committee rankings. The table shows the rank correlations, rho, and the average differences between the percentages. The correlations between preferred membership and importance to party programme and district interests respectively are practically identical. This suggests that party programme and district interestshave

estshaveapproximately the same impact on the MPs' preferences as to committee membership. If we have to distinguish between the two, however, district interests seem to be somewhat more important to membershippreference party programmes, as indicated by the average percentage differences.

Voting

Usually it is possible to predict the outcome of parliamentary debates and voting on the basis of the standpoints taken by the parties in committee recommendations. Normally, the parties act very cohesively when voting in the Storting. Votes in the Norwegian parliament have not been thoroughly analysed. However, Shaffer's data from 1985-86 are probably quite representative. 84 percent of the votes in this period every single representative with his or her party (Shaffer 1991).

If a party's vote is split in the Storting, the issue is likely to contain some kind of geographical conflict. If the cabinet's survival is not at stake, the MPs will normally be free to vote according to the interests of their constituencies, irrespective of their party's official standpoint. And even if the survival of the cabinet or a single minister is threatened, it is not always easy to get every MP to vote together with the rest of his party if important district interests are involved. This fact may be illustrated by two noteworthy

In 1970, the non-socialist coalition government under the premiership of Per Borten proposed a rise in the kilometre-based taxation on transport. A conservative MP from the county Sogn og Fjordane and a Centre Party representative from Hordaland declared that they would vote against the proposal made by their own government. The cabinet would not survive a defeat, but the two MPs represented counties with very long transportation distances. Important district interests were involved, and the two MPs did not give in until, after the debate had started, they had been assured that the interests of the periphery would be taken care of in other ways.

More recently, the Labour government headed by Gro Harlem Brundtland severe problems when the Storting was to decide the localization of a new main airport in the Oslo region. This has been a difficult issue for decades, and still is. In 1988 the government's choice was Gardermoen in Akershus county. The other alternative was Hurum in Buskerud county. Gardermoen is located to the northeast of Oslo and Hurum to the southwest. Several parties split their votes. A majority of 81 MPs voted for Hurum, while 76 supported the Gardermoen alternative proposed by the government. Fourteen Labour MPs voted for Hurum and thus became responsible for their own government's defeat. They represented the counties Vestfold, Østfold and Telemark, all of which are located to the south and west of Oslo. (Later this decision was changed, but that is another story.)

The political consequences of this outcome for the government were severe. The Minister of Transport had to resign. As one of the cabinet's 'heavy-weights', however, he was appointed as new Minister of Municipal Affairs. Several other changes in the cabinet, which were planned to take place somewhat later, were effected at the same time. Afterwards, some of the Labour MPs who voted against the government proposal said they did not know that the proposal was of such great importance to the government. They thought they had their 'normal' freedom to vote in favour of their district's interests. The Prime Minister had not explicitly declared the consequences of a defeat. Other Labour MPs claimed that everybody ought to have known. However, the fact that the Prime Minister and the parliamentary group of the Labour Party did not 'order' the MPs to vote in favour of the cabinet's proposal, illustrates the importance of the district interests. Even when district interests clash with party interests, the party leaders are very reluctant to instruct their MPs to vote against the interests of their constituency.

As we have seen, the parliamentary committee for industrial affairs is also one of the most important ones for promoting district interests. Since Hernes interviewed the MPs in 1977, this committee has become even more important. Norway has become a major oil and gas producer, and is now among the half dozen largest oil exporters in the world, by far the largest outside OPEC. At the same time, Norway is a major supplier of natural gas to Great Britain, Germany and France (Bergesen & Sydnes 1990). As a result of these developments, the committee, which is also responsible for petroleum policy issues, has changed its name to the committee for energy and industry.

Otterå (1985) has examined cleavages in Norwegian petroleum politics. As the basis of his analysis, Otterå used dissenting remarks in petroleumrelatedrecommendations various parliamentary standing committees. It is a declared aim of Norwegian petroleum policy to let the whole country

benefit from petroleum resources. From 1973 until 1985 Otterå found geographical conflicts between centre and periphery, as well as conflicts concerning localization of the petroleum activities, in 15 to 20 percent of the dissenting remarks. The left-right conflict between public and private engagement, by comparison, was manifested in 31-41 percent of the dissentingremarks.

As Schattschneider (1975) has pointed out, conflicts compete with each other. One kind of conflict can become dominant only if another one is subordinated. People must, in short, choose among conflicts. Under normal circumstances, the left-right conflict is the dominant one in Norwegian politics. When this conflict is subordinated, however, some kind of centreperiphery is likely to be involved. An example from petroleum politics is illustrative in this respect.

When the Norwegian state-owned oil company Statoil was established in 1972, there was a potential for both kinds of conflict. State ownership in industry has been a conflictual issue in Norwegian politics for many years. In this case, however, there were no protests against the establishment of a new state-owned company. The only dissenting remarks in the recommendation by the committee on industrial affairs concerned the geographical location of the oil company and the new Petroleum Directorate, was also established at that time. Three towns were proposed: Oslo, Stavanger and Trondheim. Several parties experienced a split vote when the majority decided that Stavanger, which is located nearest to the oil fields in the North Sea, was to be the 'oil capital' of Norway. The leftright in other words, was suppressed by the geographical conflict.

In more recent years, geographical conflicts have been most conspicuous when the Storting made its decisions on the location of supply bases and operational organizations for the different oil and gas fields. On several occasions the MPs have voted according to their geographical background, and not according to their party affiliation (Rommetvedt 1992a).

Questions and Contact with the Administration

Normally, the cabinet initiates decision-making processes in the Norwegian parliament. In most cases the Storting and its committees respond to bills and reports submitted by the government. However, MPs may also play a more active role, especially in addressing questions to cabinet ministers. Over the last few years the number of questions asked in parliament has strongly increased. In 1985-89, for example, there was a total of 755 questions per year, i.e. a doubling compared with the election period leading up to 1973 (Rommetvedt 1992b).

Kuhnle & Svåsand (1984) have examined the questions addressed to
some of the ministers during the period 1977-81. Even though they found

that the number of district-oriented questions was somewhat smaller than expected, quite a substantial number were focused on district-related problems.More 56 percent of the questions addressed to the Minister of Transport were related to the districts. Almost 33 percent of the questions to the Minister of Municipal Affairs and Employment concerned district interests, while 30 and 26 percent of the questions to the Ministers of Education and Social Affairs respectively were district-oriented.

Blichner & Olsen (1986) have similarly analysed petroleum-related questions in the Storting from 1960 until 1985. They found that most of the questions on petroleum policy were raised by MPs living in the coastal counties near the offshore oil and gas fields. Forty percent of the petroleumrelated concerned problems in a specific county or district. This share has increased since 1973. In 1981-85 more than half of the petroleumrelated were district-oriented. Eighty-seven percent of the questions the development of petroleum fields, 70 percent of the questions concerning delivery of goods and services, and 40 percent of the questions on education and research were related to a specific district. The proportion of questions referring to a specific constituency or county has risen considerably.

In a parliamentary system like the Norwegian one, contact between the Storting and the ministries is supposed to be taken care of by the ministers. Nevertheless, a considerable amount of informal contact between individual and civil servants in the ministries has been registered. When MPs were interviewed by Hernes in 1977, they were asked what were the most important and second most important reasons for their contacts with the ministries. Seventy-two percent of the MPs referred to committee work, while 69 percent said that contact was made on initiatives from their constituency. Only 23 percent said that personal interest in the issue was among the most important reasons for contact with the ministry (Hernes & Nergaard 1989).

The MPs were also asked how they could best promote the interests of their districts. Seventy-two percent of the representatives said that their contact with the ministries was the most important, or second most important Next came committee work, which was mentioned by 30 percent of the MPs. Twenty-three percent of the MPs said that cooperation MPs from their own constituency representing other parties, was the most important or second most important factor for the promotion of district interests.

Conclusion

In this article we have examined the representation of the periphery in the

Norwegian parliament and the impact of district interests on policy-making in the Storting. We have tried to give a more nuanced picture than the 'centralistic' impression given by Rokkan & Urwin (1982) in their comparativetypology state-building and nation-building processes.

At the constituency or county level, it is evident that the periphery has been overrepresented in the Norwegian Storting during this century, even though the underrepresentation of the central counties has been somewhat reduced over the years. At the municipal level, there is at present an almost identical percentage distribution of population and of representatives, depending on how central the residential municipality is. Yet to be more specific, even using this measure the most centrally located municipalities are slightly underrepresented.

Policy-making in the Norwegian parliament is marked by conflicts between political parties. The left-right dimension is the dominant one in Norwegian politics and policy-making. We have, however, seen several examples where the left-right cleavage has been suppressed by other conflicts. When party votes are split, the issue is likely to contain a conflict between different geographical areas. Even the survival of the cabinet or that of a single minister may be in danger if centre-periphery conflicts occur. And MPs expend a considerable amount of energy on the promotion of district interests.

The Norwegian parliament first and foremost constitutes an important part of the political centre of Norway. As we have seen, however, regional and district interests are well integrated into the Storting. To a certain degree, we may say that the Storting constitutes a regional link, a link which is missing according to Aarebrot (1982). Hence, the Storting is not only the political centre of a unitary state, it is also the central assembly of the Norwegian periphery.

---

NOTES

1. In the period from 1945 till 1973, 84 percent of the recommendations made by the Storting committees were unanimous (budgetary recommendations not included). Dissenting remarks occurred in only 16 percent of the recommendations. Since the beginning of the 19705, however, the conflict level in the Norwegian parliament has increased. In 1985-89 the proportion of recommendations with dissenting remarks more than doubled. In this period, there were dissents in 35 percent of the committee recommendations (Rommetvedt 1992b).

REFERENCES

Aarebrot, F. 1982. 'Norway: Centre and Periphery in a Peripheral State', in Rokkan, S. &
Urwin, D. W. eds., The Politics of Territorial Identity. London: Sage.

Anckar, D. 1991. 'Parliaments, Centers, Peripheries'. Paper presented at the XVth World
Congress of the International Political Science Association, Buenos Aires, 21-25 July, 1991.

Bergesen, H. O. & Sydnes, A. K., eds., 1990. Naive Newcomer or Shrewd Salesman?
Norway - a Major Oil and Gas exporter. Lysaker: The Fridtjof Nansen Institute.

Blichner, L. Chr. & Olsen, J. P. 1986. Spørsmål i Stortinget. Bergen: Universitetsforlaget.

Eliassen, K. A. 1985, 'Rekrutteringen til Stortinget og regjeringen 1945-1985', in Nordby, T.,
ed., Storting og regjering 1945-1985. Institusjoner - rekruttering. Oslo: Kunnskapsforlaget.

Goodsell, C. T. 1988. 'The Architecture of Parliaments: Legislative Houses and Political
Culture', British Journal of Political Science 18, 287-302.

Hellevik, O. 1969. Stortinget - en sosial elite? Oslo: Pax.

Hernes, G. 1971. Interest, Influence, and Cooperation: A Study of the Norwegian Parliament.
Ann Arbor, Michigan: University Microfilms.

Hernes, G. & Nergaard, K. 1989. Oss i mellom. Konstitusjonelle former og uformelle kontakter
Storting- Regjering. Oslo: FAFO.

Kommunal Rapport, no. 15, 1989.

Kuhnle, S. & Svåsand, L. 1984. 'Spørreordningene og politiske profiler i Stortinget 1977-1981',
Berg, O. & Underdal, A., eds., Fra valg til vedtak. Oslo: Aschehoug.

Lipset, S. M. & Rokkan, S. 1967. 'Cleavage Structures, Party Systems, and Voter Alignments:
An Introduction', in Lipset, S. M. & Rokkan, S., eds. Party Systems and Voter Alignments.
New York: The Free Press.

Otterå, M. 1985. Petroleumsvirksomhet og konfliktlinjer. Bergen: Institutt for administrasjon
og organisasjonsvitenskap.

Rokkan, S. 1967. 'Geography, Religion and Social Class: Crosscutting Cleavages in Norwegian
Polities', in Lipset, S. M. & Rokkan, S., eds. Party Systems and Voter Alignments. New
York: The Free Press.

Rokkan, S. & Urwin, D. W. 1982. 'Introduction: Centres and Peripheries in Western Europe',
in Rokkan, S. & Urwin, D. W., eds., The Politics of Territorial Identity. London: Sage.

Rommetvedt, H. 1992a. 'Regionoil' Politics and Policy in Norway, Regional Politics & Policy
(forthcoming).

Rommetvedt, H. 1992b. 'Norway: From Consensual Majority Parliamentarism to Dissensual
Minority Parliamentarism', in Damgaard, E., ed. Parliamentary Change in the Nordic
Countries. Oslo: Norwegian University Press (forthcoming).

Schattschneider, E. E. 1975. The Semisovereign People. Hinsdale, Illinois: The Dryden Press.

Schaffer, W. R. 1991. 'Inter-Party Spatial Relationships in Norwegian Storting Roll Call
Votes', Scandinavian Political Studies 14, 59-83.

Valen, H. 1988. 'Norway: Decentralization and Group Representation', in Gallagher, M. &
Marsh, M., eds. Candidate Selection in Comparative Perspective. London: Sage.