The Michelsen Institute, Bergen

1. Introduction

Great changes occurred in the political culture of Norway during the last decades of the 19th century. Perhaps the most important of these changes was the channeling of voters' interests into political party organizations: political conflict became party conflict. This development began at the parliamentary level, but already before the turn of the century a similar development took place at the local level, in local politics.

This is the development I have chosen to analyze in this paper: the
entrenchment of political parties in Norwegian local politics. I shall be
concerned with three sets of questions:

(a) how did the parties gradually establish themselves in local politics from
the turn of the century up to the present day?

(b) which were the typical conditions for this "politicization" of local government
1 What were the economic and social conditions for the formation
of political parties on the local level?

(c) how does such "politicization" lead to changes in local political culture?
Party politicization has been interpreted as a fourth step in the development

of mass democracy,2

— the first step in this development was the formal Incorporation of the adult
population through the introduction of universal suffrage,

— next came the mobilization of the new groups which had been enfranchised,
- thirdly, the activation of these new groups into political life,
— and finaly the fourth step, party control of local government.

The aim of my study is to add to our understanding of the processes through which political life is changed as a society goes through the phases of the industrial and the democratic revolutions. In this way it is hoped that the analysis will contribute towards a general theory of what we may call the modernization of society.

The societies, or subsystems, I have chosen as the units for this analysis
are the primaerkommuner (communes) — the units into which the fylker
(provinces) are divided in Norway.

The communes were given no place in the Constitution of 1814. Local government was, to an overwhelming extent, in the hands of royal officials. This was particularly true for the rural communes; in the cities there was some approximation to local »elf-government. A group of citizens, some elected, some appointed by the royal officials (selectmen, "eligerede Maend") had controlling influence in budgetary matters. These bodies developed during the 18th century and their activities are described in many of our city histories.³

All this changed very little with the establishment of the domestically independent regime in 1814, The royal officials and magistrates continued to rule with very little interference at the local level. In the cities the "selectmen" could only exert very limited influence. The Storting could, theoretically, use its powers to limit the authority of the officials, but it took considerable time before it took any such action. In the rural communes there were not even any formal organs of representation. The medieval ting, the assemblies of the freehold peasantry, had largely lost importance, and the local magistrates ruled with only a minimum of interference.

All this began to change with the passing of Formannskapslovene, the Alderman Laws, in 1837. In the rural communes these laws had a profound impact: they gave the peasantry an entirely new set of organs of representation and opened up new channels of political pressure. In the cities the Alderman Laws brought about only imperceptible changes in local government in the first decade or two. Citizens in Tronhjem paid little attention to the new law and even in Christiania there were no apparent changes in the character of local power. The magistrate was still the dominant figure in local decision-making.⁵

During the first years after the introduction of formal selfgovernment the scope of local activity was by and large limited to the execution of tasks imposed upon the communes by the central government. This was not an absolute limitation, however: the local authorities were required to solve problems which were delegated to them by the central government, but the Alderman Laws contained no stipulation that would restrain the communes from taking up tasks on their own initiative.

During the decades that followed, the communal authorities gained a steadily increasing influence in economic life. This was not least due to the strong incentive to local initiative implicit in the "Norwegian system" of balanced central-local financing of infrastructure costs (roads, railways, ports ...)'

This was, however, only the first stage in a process of continuous expansion. In the 130 years which have passed since the passage of the Alderman Laws, Norwegian society has gone through the most important phases of the Industrial Revolution: the industrial "take-off" came at the beginning of the twentieth century and was followed by several waves of economic expansion.⁷

Today Norway is one of the modern industrial societies in the western
world, and the communes have extended their activity far beyond the initially
delimited spheres of influence.

But this development has not been uniform. Even if all of the communes have expanded their activity greatly, there are marked discrepancies among individual local communities. While some have been at the center of the industrialization process, others have merely been carried along by "suction". It is here we find the great possibilities for comparative analysis of social modernization at the local level.

An American sociologist Francis X. Sutton has developed a theoretical model
for the study of this process of change, the transformation of primitive primaryproducing
societies to the typical industrial societies.8 These two types of
societies are described in terms of Talcott Parsons' "pattern variables".
These two models are "ideal types", and Sutton assumes that in reality there
will be a variety of mixed cases.

Sutton's model suggests three central trends in the development of modern
political structures:9

— first, family ties and inheritance will play a much smaller role in the recruitment of leaders in business and politics; universalistic achievement norms become dominant; what counts are know-how and associational ties, - second, the local communities tend to lose more and more of their cultural, social and economic autonomy and a steadily growing network of organizations will develop across the older units, the family, the neighbourhood and the work unit,

third, the spread of such universalistic norms leads on to a stronger political mobilization of all members of the community: politics is no longer exclusively an arena for conflicts among leading kin groups; more and more groups are drawn in through extensions of the suffrage, through the emergence of interest organizations and through the formation of mass parties.

Gabriel Almond has criticized this model for its overemphasis on the polar contrasts. In practice there are a great number of gradations from the one pole to the other. This is true for the system as a whole and it is even more true if one compares the local sub-units of each system with each other: what is needed beyond the initial statement of the polar contrasts is a model for the exploration of the processes of change from the one type of system to the other.¹⁰ Later theories of social development and modernization seek to specify such factors of time-space dynamics."

Stein Rokkan has indicated several factors which he judges to be decisive for the process of mass democratization." He has emphasized the need to assemble data series for a wide range of variables characterizing the "crosslocal transaction flows":

— the monetization of exchanges and the consequent entry into a wider net
work of economic relations;

- the entry into the credit market and the consequent increase in the dependence
of the peripheral units on the central ones;

— the spread of urban commodities, skills and technologies and the consequent
changes in the structure of the local labor force;

- the mobility of workers from the primary sector into the secondary and
the tertiary and from the peripheral areas to the central;

— the development of cross-local contacts through the schools, the armed
forces, the administrative services and the dominant church;
— the growth of a membership market for voluntary associations and the
establishment of local branches of regional and national organizations;
— and finally the entry into wider market of information exchange.

This brief review of views and theories of social development and modernization may serve as a framework for the analysis of the emergence of political parties at the level of local government. The analysis below will hopefully be of some help, if but a modest one, in the understanding of the problems of political development with which the recently founded states of Asia and Africa find themselves confronted.¹¹

2. The Development of Party Politics at the Local Level: A Summary¹⁴

The Election laws of 1896 and 1919 offered important incentives to the
formation of political parties at the local level.

The law of 1896 permitted elections by PR if a designated part of the electorate petitioned for it. The election law of 1919 made PR elections obligatory as long as at lea&t two lists were presented. These two laws were of prime importance for the development of "politicized" local elections. The number of communes using the PR-system steadily increased after the election of 1898. Although pure parly lists were rare in the first PR-elections, the development went steadily in the direction of more party lists and fewer nonparty lists.

The marked regional differences in the progress of politicization seem to reflect the predominant dimensions of conflict in Norwegian society.¹⁵ Originally two lines of conflicts dominated: a territorial one and a cultural one. The peripheral areas were opposed to the capital, the peasants to the administration of the royal officials, and the defenders of the cultural traditions of rural Norway were opposed to the secularism and rationalism of city culture. As a result it took much longer tc integrate the local units into the larger national community in the peripheral areas.

This brought about differences in the politicization rates between the regions of the country, as will be seen in Table 1. Politicization was slow in getting started in Western and Northern Norway and these areas are still the least politicized. There were important differences between the regions in the initial

phase, but the basic process seems to be the same: once local government is taken over by the political parties the same kinds of changes seem to take place in political culture. Regional differences are primarily important for the timing of the processes, not their end states.

Our more detailed study of the process is based on data from one of the
peripheral areas, the West.11

Some of the trends of development toward party-political elections in the communes of Western Norway can be read out of Table 2. There were important differences among the parties and their roles in the process: some were eager establish themselves as distinct units in the local contests, others did so only after decades of competitive pressure.

The Labor party was the leader in the politicization process: in 90 per cent of the communes that passed from the traditional plurality elections to modern partisan contests, the Labor party was the first to establish itself as a distinct unit.

The opponents of the Labor party were essentially faced with three options - they could vote for independent lists, lists without any distinctive ties to the nation-wide parlies operating at the level of the parliamentary elections; — they could join forces and set up joint non-socialist lists at the local level (Borgerlige Felleslister) to compete with the Labor party;

- or they might present their own distinctive party list for the Liberal Party,
the Christian People's Party, the Agrarian (later Center) Party or the Conservative
Party.

Gradually, the independent lists disappeared as the political lists took over
and the elections became contests solely among political parties.

The faster the rate of politicization in a commune the more stable the
formation of parties seems to have been: once a list has been presented the
probability for its reappearance in the subsequent elections is very high.

3. Conditions for the Emergence of Local Parties: A Summary17

The politicization of local government came about through pressures from
"below" as well as from "above":

— there were demands from "below", from voters urging communal activity,
demands which were taken up and channelled by the political parties;
— and there were the needs of the competitors at the "top", the national
political parties, to maximize their vote potential.

The first source of demands increased in importance with the development of the economic system: the change from a primary economy largely at the subsistence level to a modern money economy. We can date the industrial "take-off" in Norway to about the first decade of this century. The urban and the rural working classes which emerged had to contend with a variety of problems both social and economic. In order to solve these problems at least two things were necessary: first, united action within a party of the underprivileged classes; second, a governmental institution through which to work. The urban and the rural working classes had become increasingly class conscious under the impact of the spread of socialist and syndicalist ideas and soon realized that the local governmental system could be used as an important channel of influence. A united front at local elections was necessary in order to fill the local councils and boards with their representatives. This is the essential background for group formations at the local level.

The different alternatives of value allocation which appeared in local politics had their parallels at the national level. Similar ideological views were pitted against each other at both levels. This provided a natural contact point for national and local politics: there were important parallels in the alliance alignments at the two levels.

While the local units found it to their advantage to campaign under the banner of the national parties, the national parties also found it to their advantage to establish themselves at the local level. This is the other basic reason why the parties established themselves at the local level. By establishing local organization machinery it seemed possible to use the voter potential to better advantage in Storting elections. At the beginning, this motive proved strongest within the Labor Party, but little by little the other parties also began to give high priority to representation at the local level.

The Labor Party established itself first as the spokesman for "the underprivileged"

and in answer to this, the non-socialist parties organized themselves
as defenders of "the Establishment".18

It is not a sufficient condition for the poiiticization of local government that the national political parties feel the need for local party organizations. By and large it seems that a necessary prerequisite for poiiticization is some degree of economic modernization: the economic structure must be differentiated beyond some minimum level, communication among the inhabitants of the community must be adequate, and the local unit must be of a minimum size.

Tables 3 and 4 illustrate these relationships between poiiticization and type
of economy, urbanization and size.

It seems, however, as if the weight of these causal factors has undergone some change from the turn of the century until today. There is a tendency for local electoral institutions to change their character with poiiticization. Local issues are pushed into the background and the local elections tend to approximate pure referenda among the party alternatives: these make it possible for the political parties to measure their strength at some point between the decisive contests, the national elections. We shall take a closer look at this in the next section where we shall discuss the changes which have occurred in the local political culture after the establishment of political parties.

4. Local Political Culture Change trough Politicization

We have so far treated politicization as a dependent variable: we asked ourselves what structural changes in the community led to the politicization of local politics. We shall now examine politicization as an independent explanatory variable: what happens to a community when political parties establish themselves in local politics? We shall examine questions along these lines:

— Does the local election become less a contest among individuals and more
a contest of principles when the local politics have been politicized?
— Does one find a clearer accentuation of the conflicts when the political
parties become opponents in local politics?

— Does this in turn increase interest and support at election time; is there an
increase in electoral participation?

— Do the political parties represent an activist policy which leads to an
expansion of communal activity?

— Finally, will the recruitment of local leadership come to reflect more directly the functional divisions in the community when the parties take over local government; do we find increased activity from femal voters and from voters in secondary and tertiary occupational groups?

In order to establish causal relationships, we must be able to bring in control variables. This has only been possible to a limited degree since the total number of units (communes) used for this analysis is small. This in turn has to some extent limited the possibilities of interpretation.

a. From elections among individual candidates to elections among parties

The typical rural society in Norway in the 19th century was a family society in which the household was an economic unit and where the individual ties to this unit were extensive. The family structure was paternalistic. The head of the household had absolute authority within the family and took care of the family's external affairs.

In industrial society the family does not play as dominant a role in the social system. One of the characteristics of industrialization is that the individual is "drawn out of" the family. An economic community is created between individuals who have similar roles in industrial production and further between all who share similar stakes in the* industrialized national economy. Industrial society is much less paternalistic than the old agricultural society.

The differences in social structure between industrial society and agricultural
society must also necessarily find expression in basically different electoral
situations:

- In the typical pre-industrial community conflict issues at the election of local representatives are anchored at the micro-level. The voters differ very little on the issues facing the representatives to be elected. There is little or no conflict over the allocation of values because most voters tend to accept

the Establishment. Family and neighbor ties and personal sympathies and
antipathies tend to be the deciding factors for the voters in their evaluation of
candidates.

— In the typical industrialized community, on the other hand, conflict issues are to be found at the macro-level: competing groups constitute component parts of society. Alternativcis of allocation are pitted against each other and each group will attempt to require control over the local political machinery. Personal modesty is no longer considered important. There is no reluctance to accept candidates who vote for themselves: they no longer represent themselves as persons but abstract alternatives of value allocation. The vote is no longer only for "the best men".

When the Labor Party presented its program for the local election in Oslo in 1890, C. Jeppesen, chairman for the Executive Committee declared: "Once again, it is not the change in personnel that is important and — I am convinced — every worker in Christiania desires the execution of the present program"."

Such a statement is typical for the Labor Party. However, it was not the first party to put up a program for the local elections in Oslo. The Left (Venstre) had been first in the attempt to integrate the growing working class in its ranks. But it soon became clear that this policy of the Left would not be successful both in the capital and in the provinces. The Labor Party was the first party to become consistently program-oriented. The necessity of solidarity and discipline was stressed much more distinctly in the labor movement and the high degree of partisan activation also helped to check the growth of personal leadership.²⁰

We shall now concern ourselves with two specific questions:

(a) To what degree does party politicization lead to party orientation among
the voters, and

(b) is this party orientation a characteristic of the most "politicized" parties,
the parties within the labor movement?

To study variation in party orientation I made use of electoral records; these offer information not only about the different lists presented but also about the votes for each candidate. The Electoral Law permits a voter to cumulate votes for an individual candidate: he can write in the name of his candidate one more time than it appears on the list. The name of another candidate must be crossed off the list for the cumulation to be valid. The total number of cumulations gives an index of the voters' preference for individual candidates over the partynominated list.

---

One statistical parameter of distribution of "list-votes" for candidates through cumulation may be the standard deviation. This is, however, a function of the median, m. We must use the quotient s/m in order to he able to compare cumulations on lists with different medians. In this case the minimum for s/m will be equal to 0 (no candidates cumulated or crossed off the list),while the maximum v/ill be equal to 1 (half of the candidates have been crossed off of the list and half have been cumulated).²¹) In this way we can rank the different lists along a continuous scale from 0 to 100 (when we multiply «/m by 100).

One statistical parameter of distribution of "list-votes" for candidates through cumulation may be the standard deviation. This is, however, a function of the median, m. We must use the quotient s/m in order to he able to compare cumulations on lists with different medians. In this case the minimum for s/m will be equal to 0 (no candidates cumulated or crossed off the list),while the maximum v/ill be equal to 1 (half of the candidates have been crossed off of the list and half have been cumulated).²¹) In this way we can rank the different lists along a continuous scale from 0 to 100 (when we multiply «/m by 100).

One statistical parameter of distribution of "list-votes" for candidates through cumulation may be the standard deviation. This is, however, a function of the median, m. We must use the quotient s/m in order to he able to compare cumulations on lists with different medians. In this case the minimum for s/m will be equal to 0 (no candidates cumulated or crossed off the list),while the maximum v/ill be equal to 1 (half of the candidates have been crossed off of the list and half have been cumulated).²¹) In this way we can rank the different lists along a continuous scale from 0 to 100 (when we multiply «/m by 100).

Two control variables have been used in Tables 5 and 6: the size of the commune, as measured by the totals registered, and economic structure. I have not been able to control for both variables simultaneously because of the small number of communes for which we have data (only for communes in the Hordaland and Sogn of Fjordane provinces).²²

Table 5 uses size as a criterion: it was a hypothesis that the smaller the electorate the larger the possibility that the voters would know the candidates personally. For example Modalen in the province of Hordaland had 235 registered voters in 1963 while Oslo had 345,468. It is not only personal acquaintance with a candidate which determines cumulation and the deletion of names from the pre-established list. Candidates may be known through mass media because of earlier political activity or because of the activities in other fields. On the other hand mass media play a much smaller role in local elections than in parliamentary elections.

It is also clear that the importance of deletion is evaluated differently by voters in small and large communes. A voter in a small commune has a greater possibility of influencing the election of an individual candidate because, relatively speaking, his vote is worth more than one in a large electorate. There is also a greater tendency for the parties to "pre-cumulate" at the nomination conventions in larger communes: they will list the top candidates several times to increase their chances of election.

A possible explanation might be that because most voters do not know the candidates they will refrain from cumulation and deletion. This means that the few voters who do cumulate in reality decide the composition of the list. In order to prevent this the parties cumulate beforehand, thereby removing the possibility of such minority influence on the outcome.

Table 5 shows the steadily decreasing cumulation frequency from low to
high politicization and from small to large electorates.

The trends are clear for the politicized communes: the larger the commune and the greater the number of party lists the smaller the incidence of candidate cumulation. This comes out even clearer if we single out the four largest communes, all highly politicized. For these four the average value of the

cumulation ratio is only 4 (ELS against 31 for the smallest of the communes with
one to three party lists).

Size clearly makes a difference in the communes where at least one national party presents a list. The result is markedly different for the communes where the competing lists are all local and non-party: the level of candidate cumulation is low and the size differential hardly significant.

Table 6 controls for the economic structure of the communes and the character of the lists presented. The structure of the economy clearly counts more than politicization: tho levels of candidate cumulation were consistently highest in the primary-economy communes. The differences between Socialist, non-Socialist and non-partisan lists were also very small in the primary-economy communes while quite marked in the industrialized communes. In these, Socialist voters hardly ever tamper with the lists presented by their party (most of the lists analyzed in the table are Labor lists, only very few are CP lists) while the cumulation ratio is significantly higher for non-Socialist and non-party lists (there is an exception to this regularity in the most politicized secondary-economy communes in 1931, but the number of cases is very low).

But what about the low level of candidate cumulation in the "prepolitical"
communes: in contests among non-partisan lists?

The simplest explanation would seem to be this: The non-partisan lists tend to be purely geographical and to represent candidates from a single, often very small, district within the commune. Nominations for such district lists are decided on in open meetings, and there is accordingly less incentive to express disagreement with the decisions of the meeting through cumulations and deletions. By contrast, nominations for party lists are decided at meetings of party members only and for the entire commune and this may much more often encourage voters to express their district loyalties through cumulations of their own candidates and deletions of others. A typical example of highly territorial politics is the election at the coastal commune of Fitjar south of Bergen. In 1959 altogether nine local lists, all non-political, were presented and received only 1069 valid votes in all, an average of roughly 120 each.

Our analyses suggest an intriguing interaction process. As long as local politics is strictly territorial, conflicts of representation will be reflected in the multiplication of district lists and quarrels over candidacies will tend to be settled at the open nomination sessions. As soon as the national political parties establish themselves locally and present commune-wide lists, territorial politics tend to be expressed either through continued support for dissident nonpartisan lists or through cumulations of district candidates on party lists. In the densely populated, economically advanced communities there is much less incentive to carry on with territorial politics of this type. The Labor Party, representing a markedly different dimension of cleavage, tends to be the first party to be able to presuade the bulk of its voters to abandon territorial and personal candidate preferences and to vote the straight party ticket.

b. Politicization and polarization

How does this process of local politicization affect the fate of the parties
at the parliamentary elections?

In the model suggested by Stein Rokkan²³ the three "middle" parties in Norway tend to express a territorial-cultural dimension of conflict in the political system while the Socialists and Conservatives in most contexts will express a functional-economic dimension. Now it is clear from our analysis that politicization, as we have defined the concept, implies a reduction in the salience of territorial cleavages and an increase in the emphasis on functional-economic conflict. Does this mean that politicization will generally heip to increase the parliamentary strength of the parties at the poles of the functional-economic axis, the Socialists and the Conservatives?

This is obviously not a question of unidirectional causation. In most cases there will be a process of cyclical reinforcement: a success at a parliamentary election may have offered the "take-off" incentive for the formal organization of a local party branch, and the success of a party list at a local election may have attracted a new body of voters at the next parliamentary election. Whatever correlation we find between the two processes may simply result from parrallel changes brought about through still unexplored cycles of interaction.

Table 7 classifies the communes of Western Norway by their economic structure and by their level of politicization in 1937 and in 1959: the cell entries give the Socialist and Conservative share of the total votes cast at the nearest parliamentary election. There is a clear relationship between

the two measures: the lower the level of politicization, the less the polarization. This holds for both pairs of elections and for communes at all levels of economic growth (the one exception is a curiosity: an exceedingly small island commune still sticking to the plurality system at the local election despite the high level of polarization).

The "territorial-cultural" offshoots of the old Left have clearly retained their strength most successfully in the "pre-political" and the least politicized communes. This comes cut with particular clarity in the "transitional" communes on the verge of industrialization. The community conditions which have helped to maintain the middle partes in power in the less industrialized communes of the West have also kept back the drive towards full politicization. Much more detailed analyses of sequences of change commune by commune will be required to determine the typical paths of causation. It seems clear that the initial "disequilibrating" shocks tend to be of an economic nature but it is not clear how far the politicization of local elections helps to accelerate the polarization of votes between the Socialists and the Conservatives.

c. Politicization as a mobilization factor

In analyzing the consequences of the formation of parties at the local level,
we asked the question: to what extent do these new organizations encourage
increased participation in elections?

A variety of generalizations have been suggested to account for variations in levels of electoral participation.^{2 *} A key concept in the analysis of electoral participation is "the expected differential": how do voters perceive the consequences of choosing one alternative or the other? Our suggestion is that turnout will tend to be higher the more clear-cut the expected "payoff" of each mobilized vote and that "the politicization of local elections helps to increase the value of the vote.²⁵

— In the first place, the entry of political parties on the local arena will tend
to accentuate the differences in the alternatives between which the voters
must choose and consequently increase the importance of the choice.
— Secondly, votes cast for a political party tend to serve wider functions than
those cast for non-party lists: in addition to contributing toward the selection
of local councils, they also influence the larger contest of strength between
the parties, a contest which may have consequences for the policies the
parties pursue at the national level.

We may assume that the first condition is generally satisfied wherever political parties participate in local politics; the second is more dependent upon the current national political climate and the balance of forces in the province.

Our other assumption was that participation in local elections would increase when the parties established themselves locally because the local elections also would be interpreted as a contest of strength between national parties. The local elections of 1963 are an eminent example of a "nationalized" local election: they served as an occasion for expressing support of, or opposition to the Labor party in a situation which had placed it in opposition after a vote of non-confidence in parliament.^{2 *}

It is possible to trace similar tendencies in Sweden and Denmark. There is a serious debate taking place in Sweden abtout the extent to which local elections should be allowed to have national political consequences.²⁷ A similar question was raised in Denmark after the losses of the Social Democrates in the local elections of 1966.

Table 9 presents data for the elections of 1955, 1959, and 1963. There was very little movement in the average turnout from 1955 to 1959 and no clear differences by levels of politicizations from one level of politicization to another. If, on the other h&.nd, we look at the movement from 1959 to 1963, we find a clear trend: the increased mobilization of the electorate is a function of the number of party lists presented at the election.

This confirms our hypothe'Sis that a vote cast for a party list has a function beyond that of electing local officials: a vote cast for a party list may also have consequences for that party's policies in the Storting. There is no doubt that it was the results of the election in 1963 which were the immediate cause of the Labor Party's demand to resume the reins of government after the four weeks in opposition.

d. Politicization and local activity

We have earlier stated that the politicization of local government may have
its origin in articulated demands on the part of the electorate for the building
of social institutions, communications etc.

The conflicts which arise out of the demand for increased public activity tend to be made more salient by the political parties. This depends, in part, on the degree to which the parties desire to maximize their votes potential by engaging themselves in local politics.

We may at once assert with certainty that the communes have expanded their area of competence and activity in last hundred years: "The result is that at present one cannot draw any boundaries in principle between the areas of competence for the state and the commune".²⁸ We shall here attempt to illustrate the relationship between the politicization of local politics and communal activity: is local public activity to any extent dependent upon the level of politicization of the communal government?

It is theoretically possible to imagine another causal relationship than that

expected above. We have earlier divided communal obligations into two categories: those which are delegated to the communes by the state and those which are a result of local initiative. Mandates from the state may in themselves create different alternatives of allocation and this may provide the incentive for the formation of political parties in the communes. This is only conceivable when the delegated duties allow some leeway for alternative solutions. If, for example, the national government orders the local authorities to provide certain social services without specifying the extent of these services, conflicts may arise and thus encourage parties to organize themselves. In reality, the communes have very little freedom of action when it comes to the execution of legislative orders. Conflicts arising in the execution of such orders will only rarely be a contributing factor in the politicization of local politics.

Our empirical problem is how we can define local economic activity. Official statistics are not detailed enough and budgets for the individual communes are only available for a few years before World War 11. Secondly, there is a question as to what degree the given figures are an expression of local initiative: how much of the total expenditures reflect decision taken at the national level and how much difference is there between the communes on this score? Even if there should be differences between the communes on this point, it is fair to assume that they are not of such a size that they can disturb the entire model to any notable degree. Table 10 uses the following measure of communal activity: "Budgetary expenditures according to the accounts", per capita, for the fiscal year 1937—38. The table controls for economic structure in order to separate communes according to their degree of economic modernization.

The table clearly supports the hypotheses that the entry of the parties into local government leads to increased communal activity. The trends are especially strong in communes which have already experienced economic modernization, especially the industrialized communes.

The trend is apparently in the opposite direction in fishing communes. There
are only two such communes which are highly politicized and these do not
seem representative for the group as a whole.

Table 11 is based on total expenditures per capita in 1962. The trends are the same as in 1937: total expenditures per inhabitant increase with increasing politicization. The trends are clear even if we control for urbanization and economic structure. Communes with only non-party lists prove an exception to this rule, however.

A possible explanation for this may be that in communes with only nonparty lists there is a contest between the different geographical areas for funds for infra-structure investment. These local differences are large enough to prevent the organization of parties. Francesco Kjellberg describes a special situation in a commune in Northern Norway which strengthens this hypothesis.²⁹ In his "Northern commune" only one party, the Labor Party participated at the election, but in return this party presented many different lists. Several districts within the commune had their own party list. Local differences were so strong that it was impossible for e.\\ party sympathizers to agree on one list for the whole commune.

More detailed analyses are necessary before one can clarify the situation of
the non-party lists and the conflict situations behind them.

e. Changes in the recruitment to local councils as a result of politicization
on the local level

Individual achievement became an important factor in the definition of
social roles in the modern social structure which developed during the
industrial revolution. Individual ascriptive attributes have lost in importance.30

This development can probably be traced back to the fact that the demands made by functional interest groups have become more and more accepted as legitimate in political life. The mobilization of the peasantry in the nineteenth century led to the acceptance of their economic interests as "political"

interests. A similar acceptance has appeared in the twentieth century for the
interests of the labor movement.

The mobilization of the peasantry led to the formation of political parties on the national level in 1884. It was the labor movement and the mobilization of the working class which above all generated party formations at the local level.

We may consequently expect that in addition to personal achievement a sense of solidarity towards functional interests has become an important criterion on the selection of local leadership in politicized communes. This was illustrated above where we showed that when local government becomes politicized the elections tended to take the form of contests among parties rather than among individual candidates. Further we may expect to find a steadily decreasing number of representatives in the local councils who belong to what we might call the "traditional" occupations, notably freehold farmers, public officials such as teachers, clergymen and lensmenn (appointed local sheriffs). As the quardians of property, knowledge, religious values, and national authority these officials were able to satisfy the demands for local government services in the "pre-political" community. Many of the officials had close personal ties to the local peasantry and this helped to maintain them in power even after the initial changes in economic and political structure. But with continuing modernization we may expect a change. Secondary and tertiary occupational groups will take an increasingly greater importance in the recruitment of personnel to the local councils as political parties enter the political scene in the communes.

With this restructuring of occupational recruitment we may also expect a gradual change in the recruitment by sex. One characteristic of modern industrialized society is a functionalization of roles which leads to an increase in the number of women employed outside the home. In this way women increasingly become distinct units of the economic system. As they become economically active they will be motivated to take care of their own interests. They will no longer be satisfied with the status of subordination to the head of the household.

Thus women will tend to constitute a new pressure group in an industrializing
society and will have better chances of support the more political life
becomes dominated by competition among distinct parties.

We may expect that women have greater chances of election as local representatives in politicized communes than in communes without competition among political parties. From other studies we know that this is the case at the national level at the nomination process for Storting election.³¹ Table 12 shows the recruitment to the local councils from different occupational groups in communes at different levels of politicization: there is a control for type of economy.

The figures for 1931 show clearly that recruitment from the occupational
groups typical of industrial society, that is workers and white-collar personnel,

increases with the level of politicization. The trend is the opposite for primary occupations. There are fewer and fewer fishermen and farmers on the lists as the level of politicization becomes higher. This trend is most prevalent in secondary-economy communes. The figures for 1959 show a similar trend for the category "urbanized secondary" and to a certain degree for the category "rural mixed", but the figures for the group "rural primary" do not fit this pattern.

A possible explanation for this exception may be that politicization was differently motivated during the first waves of politicization than it was in the latter. In our introductory presentation we traced politicization back to two main sources: (1) to changes in economic structure which were a result of the industrial revolution, and (2) to efforts of the national parties to exploit their voter potential and build up networks of organizations. It looks as if this last factor was the most important in the politicization of local politics in the last 20—30 years.

The communes in the &roup "rural primary" were the last to become
politicized. The party list constellations in these communes were to a lesser

degree a reflection of socio-economic conflict. Politicization was mainly a
result of efforts of the parties to maximize their voter potential. This might
explain the very small changes in the recruitment pattern for these communes.

With regard to the group "public officials" our hypothesis was that this group would gradually disappear from the lists as local government became politicized. We were mainly concerned about the traditional public service groups, i.e., teachers, clergymen and lensmenn who, because of their position in the community as agents of knowledge, religion and state authority automatically and necessarily became key figures in local administration in the first decades of local government in Norway.³² Table 12, however, also includes groups under the category "officials" which do not play similar roles in the community. There are other sources which at least in part seem to bear out our hypothesis. There were 104 elementary school teachers who help the positions of mayor in 1931, while the figure in 1961 had decreased to 30, even though the total members of mayors remained relatively constant.³³

The category "independents" also played a dominant role in the pre-industrialized community. The members of this group often were the sole economic point of contact with the outer world (the extreme case were the vaereiere in Northern Norway, the monopolistic ruler of the fishing Communities who controlled the fates of the fishermen through ownership of industries, stores, houses and through credits for fishing boats and expeditions). Table 12 shows that "independents" gradually disappear from the local councils as local government becomes dominated by party politics.

Table 13 shows that as the national parties become engaged in local politics, economic/functional interests are more clearly accentuated in the local councils. A conscious effort is made to recruit representatives from different interest groups to the councils. This also confirms the trends in TabJe 12.

We hypothesized that women take a more active part in local government with increasing politicization. In 1913 there was only one female representative in 120 local councils in Wostern Norway and only seven alternates. By 1931 the number of female representatives in 201 local councils had increased to nine with 40 female alternates.

It was not until 1937 that the number of women in local councils became so
high that it makes sense to calculate the number in percent of the total
number of representatives (Table 14).

This table shows a clear trend which strengthens our hypothesis: more women are recruited to the local councils as local government becomes politicized. The figures for female alternates are larger than the figure for female representatives. This makes it clear that even if women are nominated to lists the voters do not accept female candidates as readily as their male counterparts when it comes to the administration and governing of a commune. They may be placed on the list to attract votes but they will tend to be deleted with such frequency that they do not get elected as regular representatives.

Table 15 gives differences in the recruitment of women to local councils by type of list. The political party lists tend to recruit women more frequently than the non-party lists; the one exception is the Center Party. This party is dominated almost completely by farmers and smallholders, groups which more than others have preserved the originally patriarchal culture. This may explain the differences between this party and the others.

It has been established in earlier analyses that the mobilization of women for roles in the political system first got under way in the established classes.³⁴ This is also demonstrated by the relatively high percentages for the Conservative Party in Table 15. The Socialist parties have followed up in this mobilization process in spite of a handicap of lower access to higher education among the classes which support these parties.

5. Conclusion

I have in this article described the increasing engagement of the political parties in local government in Norway and have suggested some of the conditions for this development. Politicization of local government in Norway began at the turn of the century. The process was a central factor in the mobilization of the underprivileged classes brought into the political system through the introduction of universal suffrage. The political parties, especially the Socialist parties, have been the main agents of this mobilization. The development was most rapid and went furthest in those parts of the country where the industrial revolution brought about the highest degree of economic modernization, in Eastern Norway and in Trøndelag.

In more detail I have described the changes which have taken place in what I have called the "local political culture," a vague and allinclusive term which obviously needs to be broken into several facets. As a result of the entry of the parties into local goverment:

— voters tend to become more party-oriented and less candidate-oriented;
— conflict issues tend to become more accentuated than earlier, politics more
polarized;

— increasing importance tends to be placed on elections because conflict
issues are more clearly articulated by the parties, and this in turn increases
participation;

— the parties tend to increase the extent of communal activity, they represent
an "activist" policy;

— the recruitment basis for the local council tends to change as the partier take over the elections: the parties crystallize the differences between different interest groups and the recruitment patterns tend to reflect the internal group structure of the community.

The basic model of analysis can be summarized in this scheme of causal
chains:

I believe I have demonstrated through these initial analyses that the trends of change have been in the direction of the arrows. However, I have not been able to establish the weight of these influences and to compare them with other sources of change over time. To do this, additional data will have to be gathered and more sophisticated techniques of analysis will have to be applied.

---

NOTES

1 I shall use "politics" in two distinct senses in this paper. "Politicization" means "party politicization". This is an operational definition: a much more general concept of "politics" is implicit in the use of the term "local politics".

2 S. Rokkan: "Electoral Mobilization, Party Participation and National Integration", in J. Lapalombara and M. Weiner eds., Political Parties and Political Development. Princeton, N. J., Princeton University Press, 1966, pp. 241—265.

3 E. g. Knut Mykland: "Fra Søgaden til Strandgaten 1807-1880", Vol. 11l of Trondheim bys historie. Trondheim, F. Bruns Bokhandels Forlag, 1955, and S. C. Hammer in Vol. IV of Kristianias historie. Kristiania, Cappelen, 1923.

4 Cf. Arne Bergsgaard: "Formannskapslovene, opphav og utvikling", in Minneskrift til formannskapslovenes 100-års jubileum: 1837-1937. Oslo, Gyldendal 1937, 7-13. 1937, 7-13.

5 Cf. K. Mykland: "Fra Søgaden til Strandgaten 1807-1880", op. cit. Vol. 11l pp. 224—225 and S. C. Hammer, op. cit., p. 227.

6 Jens Arup Seip has analyzi?d this system and stressed the importance of Saint ideas on economic policy in Norway at this time in "Det norske system", Historisk tss. 38(1), 1959: 1-58.

7 Cf. W. W. Rostow: The Stages of Economic Growth. Cambridge, Cambridge Univ. Press, 1960.

8 F. X. Sutton: "Social Theory and Comparative Politics", in H. Eckstein and D. Apter, eds., Comparative Politics. The Press of Glencoe, 111., 1963, pp. 67—81.

9 These formulations were originally given in S. Rokkan: Sammenlignende politisk sosiologi, Bergen, The Michelsen Institute, 1958.

10 Cf. G. A. Almond in Almond and J. S. Coleman, eds.: The politics of the Developing Areas, Princeton, Princeton Univ. Press, 1960, p. 23.

11 E.g. D. Learner: The Passing of Traditional Society, Glencoe, 111., Free Press, 1957 and K. Deutsch: "Social Mobilization and Political Development", Amer. Pol. Sci. Rev. 60 (3), 1961: 493-514.

12 Cf. S. Rokkan: "Electoral Mobilization, Party Competition and National Integration," op. cit.

13 Cf. particularly the concluding chapter of Almond and Coleman, op. cit

---

14 Detailed discussion and documentation will be found in T. Hjellum: Partiene i lokalpolitikken. Oslo, Gyldendal, 1967.

15 See S. Rokkan: "Norway: Numerical Democracy and Corporate Pluralism", pp. 70—115, in R. A. Dahl, ed.: Political Oppositions in Western Democracies. New Haven, Yale University Press, 1966.

16 This choice was partly made on economic grounds, but the choice was also made because data from Western Norway is particularly suited for such an analysis: the slow development of politicization makes it possible to analyze the process in greater detail, to study both the conditions for and the effects of the emergence of parties at the local level.

17 Detailed discussion and documentation in T. Hjellum: Partiene i lokapolitikken, Oslo, Gyldendal, 1967, op. cit.

18 By and large "underprivileged" refers to groups which were enfranchized in 1898. By "The Establishment" we mean the groups which were already an integral part of the political system in 1898. Cf. S. Rokkan, chapter on Norway in R. A. Dahl, ed., op. cit.

19 C. J. Hambro: Oslo Høire gjennom 50 år. Oslo, Johan Grundt Tanum, 1934, p. 137.

20 See J. Valen and D. Katz: Political Parties in Norway. Oslo, Universitetsforlaget, 1964. p. 129-131.

21 It is possible for a candidate to obtain more than 2m list votes if the candidate has also been placed on other lists(slengstemmer). This has little effect since the total number of such votes is usually small in comparison to the median m.

22 Examinations of s/m are complicated and time cosuming even with the help of a computer. It had, therefore, only been possible to use data from two of the provinces and for three elections: 1931, 1959 and 1963.

23 See S. Rokkan: "Norway: Numerical Democracy", op. cit.

24 A useful review of variables affecting levels of political participation is given in A. Campbell: "The Passive Citizen," in S. Rokkan, ed.: Approaches to the Study of Political Participation. Bergen, The Michelsen Institute, 1962, pp. 6—21.

25 Stein Rokkan and Henry Valen have explained the variations in the differences in the level of political participation between local and parliamentary elections through this type of reasoning. See Rokkan and Valen: "The Mobilization of the Periphery," in S. Rokkan, ed., Approaches to the Study of Political Participation, op. cit. pp. 114-158.

26 For details see S. Rokkan: "Norway: Numerical Demoncracy", op. cit.

27 See Författningsjrågan och det kommunala sambandet. Betänkande av länsdemokratiutredningen, Stockholm, Justitiedepartementet, 1965, p. 482.

28 K. Nordanger og A. Engh: Kommunalkunnskap. Oslo, Tiden Norsk Forlag, 1951 p. 20.

29 F. Kjellberg: "Politisk lederskap i en utkantkommune", Tidsskrift for samfunnsforskning, 6 (1), 1965, pp. 74—90.

30

31 Cf. Henry Valen: "The Recruitment of Parliamentary Nominees in Norway", Scandinavian Political Studies, Vol. I (1966), pp. 121—66, especially pp. 130—39.

32 An examination of participation in political life by clergymen and lawyers has been made by Ulf Torgersen: The Political Participation of Norwegian Professionals 1870—1940: Minister and Lawyer. Oslo, Institutt for Samfunnsforskning, 1966.

33 Director of Schools, Svendsen, in a speech held at Trøndelag, 1963, as cited in Dagbladet.

34_See Rokkan and Valen: "The Mobilization of the Periphery" in S. Rokkan, ed: Approaches to the Study of Political Participation, op. cit. p. 114.